Matchpoints

Second Edition

Kit Woolsey

B RIDGE W INNERS P RESS

© Kit Woolsey 2015
All rights reserved.

Second edition
Edited by Adam Parrish
Printed in the United States of America
May 2015

Cover design by Greg Humphreys

Cover art © Baron Barclay. Used by permission of Baron Barclay.

ISBN: 978-0-9905229-6-6
Library of Congress Control Number: 2015937220

First edition ©1982 Devyn Press

Edited by Bernie Yomtov

Bridge Winners Press

[Link]
 FOREWORD

Not too long ago, I had the pleasure of playing a week of bridge with
Kit Woolsey.
Now anyone who has played with a variety of partners knows that
their ability and experience do not necessarily contribute to being a
good partner. In Kit’s case, however, we had no such problems and
our short-lived partnership was a pleasure.
But there has always been a question in my mind about Kit. Every
now and then there would be a lull in the action and I would look up to
find Kit in some sort of trance, his head gently nodding, gently bobbing.
Eventually, he would come out of it with some play or bid. In spite of
the fact that he usually came up with the winning decision, I have often
wondered what he was thinking about.
Now I know.
I’ve certainly played my share of matchpoints and I feel secure that
I am familiar with most strategies. Nonetheless, as I read Matchpoints I
discovered a number of concepts which I had not fully appreciated and I
found another group of concepts which had not really been crystallized
in my mind.
One topic which had particular appeal to me was a section on slam
bidding. Kit offers a number of pertinent points, three of which stick
out for special comment.

1. Even at matchpoints, it is not necessary to pig out in 6NT. Quite

often six of a minor will prove to be a top or near so. If 6 ♣ is
cold and 6NT is only so-so, you will win whenever 6NT fails or
when the field gets to game only.

2. If your sequence tells you that 3NT is the best contract and you
are already at, say, 5 ♣ before making this discovery, go ahead
and bid 6 ♣. If the field is making +630 or +660, your +620 will
be as worthless as −100. Bid 6 ♣ and hope. Not good bridge, but
practical matchpoints.
 MATCHPOINTS

3. In this day of science and more science, I am quite pleased to

see someone recommending a throwback to the days of brute
blasting.

N-S vul., South holds ♠ AQ9764 r A6 q 6 ♣ AJ62

West North East South

— — — 1♠
2q 3♠ 1 4q ?
1. Limit raise

Kit suggests that you simply bid 6 ♠. Perhaps you will make it. Perhaps
the opponents will save.
Quite crude. I like it.
Other topics range from a discussion of the peculiar problems
of matchpoints to ways of creating additional problems for your
opponents.
Later, Kit discusses whether to bid game or a partscore and then
which game or partscore to bid. Should you play in an obvious 4–4 fit
or should you reject it for 3NT? When should you attempt a 4–3 fit?
Why should your decision differ at matchpoints as opposed to IMPs?
And finally, having negotiated the auction, how should you play it
and defend it? In other words, should you go for the overtrick or play
safe for the contract? Should you risk an extra undertrick or should you
quietly accept down 1? Alternatively, should you go all out to defeat
their contract or is it sufficient to hold them to the contract or simply
limit the overtricks?
All in all, Kit has covered or touched on an enormous number of
significant ideas; you will not find such a summary between the covers
of any other book.
A word of warning. This is not a book for the casual reader. There
are many interwoven concepts, and less than a careful reading will
not give the reader the full benefit. In fact, a cursory reading may be
dangerous. Handle it with care, however, and it will be a rewarding
effort.

Michael Lawrence
1982

iv
 FOREWORD TO THE
SECOND EDITION

There are important milestones in every bridge player’s development—

moments when the light turns on and you realize something about the
game you hadn’t grasped before. Your first finesse, your first squeeze,
the first time you’re able to work out declarer’s shape. Each time you
reach one you discover a whole new level of bridge, on which the
experts have apparently been operating all along, though you were until
now totally unaware of its existence.
For me, the most profound of these moments came a year or so after
I had started playing bridge seriously. I was in my 20s and had maybe
300 masterpoints, most of them earned within the last year. I was a Life
Master now and thought I had this game pretty much figured out. Then
someone gave me a book with a bright yellow cover called Matchpoints.
Of course I knew how matchpoints worked, and that my score on a
hand was based on the results at other tables. But it had never occurred
to me that I could anticipate likely outcomes at those tables and evaluate
my decisions based on how my results would matchpoint against them.
A concept like taking a safety play because your contract is so much
better or worse than the contract the field is likely to reach was totally
foreign to me—It’s matchpoints: you go for all the overtricks you can
get! But it became obvious as soon as I read it. How had I not realized
this before? I’m sure I had heard the term “matchpoint double” at
that point, but I didn’t really understand the concept; it clicked when
I started thinking on an expected matchpoint gain/loss basis.
Over the last few months I have had the privilege of diving even
deeper into Matchpoints as I worked with Kit on the second edition. I
still marvel at the clarity of Kit’s writing and the quality of his analysis,
and I’m sure my game is better for having studied this book so carefully.
Look out Platinum Pairs, here I come!
A few things struck me as we worked on putting the new edition
together. The first was how much had changed in “standard” bidding
 MATCHPOINTS

since Matchpoints was originally published in 1982. At least half of

the auctions in the book had some element that felt outdated to me.
We changed some things around to make the 2/1 bids game-forcing and
took out some conditional language like “If you use Negative Doubles.”
We left other example auctions and hands alone, even if they seemed a
tad unusual, because it is the underlying concepts that are important. I
was often asking Kit things like, “This bid seems odd to me. Should we
change it, or at least explain it?” His answer was usually, “Who cares.
People do strange things sometimes. The bidding isn’t the lesson, the
concept is.”
The second thing I noticed was how little of the actual analysis
we had to change. A few examples got cut or changed around to
make the bidding make sense in a modern context, but there were no
concepts in the original book that aren’t true any more or have changed
dramatically. The theory here is really solid and timeless.
My favorite observation is that today’s average opponent is a much
better matchpoint player than the average opponent 33 years ago—
largely thanks to this book! Statements like “It is likely that other
tables will not receive as much competition and will buy the hand
for 2 r” may still be true, but to a much lesser degree. Even if
you haven’t read Matchpoints, the strategies it teaches—aggressive
overcalls and preempts, matchpoint doubles, competing aggressively
using the Law of Total Tricks—have become so mainstream that you
probably consider them basic bridge knowledge.
There have been a lot of bidding innovations over the last 30 years,
and at several points in the process I asked Kit how he felt about
some newfangled treatment that I thought would solve the problem
being presented. His responses, though not always mainstream, were
instructive and insightful, though sadly they fell outside the purview
of this book, and much as I would have liked to have included them,
such digressions only would have muddied the waters. I think I can
summarize Kit’s general thinking in one sentence, though: he likes
conventions that establish a fit and put pressure on the opponents, and
he dislikes conventions that give the opponents anything (bidding space,
opportunity for a lead-directing double, a convenient description of
their hand). The two he most adamantly despised were constructive
raises (you give the opponents the whole 2-level and hide your fit from

vi
partner!) and mixed raises (you take up less space and provide an
opportunity for a lead-directing double [assuming the raise is artificial],
and the more precise definition of your strength is often more helpful
to the opponents than to partner—a jump-raise to 3 ♠ is much harder
to deal with as an opponent when it could be a 2-count or a 9-count).
Frequency is the name of the game at matchpoints, so missing the odd
game is made up for by consistently putting pressure on your opponents
and forcing them to guess at the 3- and 4-level. Ultimately, it’s not
gadgets that win matchpoint events, it’s judgment, and that’s where the
focus of this incredible book remains.
It has been a privilege to work on such a classic, and on behalf of
the entire bridge community I’d like to thank Kit for taking the time to
update and republish it. I hope you enjoy reading it as much as I have
enjoyed working on it, and that it does for your game what it did for
mine all those years ago.

Adam Parrish
2015

vii
 Table of Contents

I General Concepts

1 The Cost of Being Wrong . . . . . . . . . . . . . 3

2 The Mathematics of Matchpoints . . . . . . . . . . 9

3 Two Ways to Win . . . . . . . . . . . . . . . . 19

4 Loading the Dice . . . . . . . . . . . . . . . . . 27

Review Problems . . . . . . . . . . . . . . . . 33

II Constructive Bidding

5 The Best Game . . . . . . . . . . . . . . . . . 57

6 Game or Partscore . . . . . . . . . . . . . . . . 79

7 The Best Partscore . . . . . . . . . . . . . . . . 95

8 Paving the Way . . . . . . . . . . . . . . . . . 103

9 Slam Bidding . . . . . . . . . . . . . . . . . . 117

10 Low-Level Interference . . . . . . . . . . . . . . 127

11 High-Level Interference . . . . . . . . . . . . . . 143

Review Problems . . . . . . . . . . . . . . . . 155

III Competitive Bidding

12 The Fighting Ground . . . . . . . . . . . . . . . 177

13 The Law of Total Tricks . . . . . . . . . . . . . . 187

14 Matchpoint Doubles . . . . . . . . . . . . . . . 201

15 Nuisance Bids . . . . . . . . . . . . . . . . . . 215

16 Balancing . . . . . . . . . . . . . . . . . . . . 223

17 The Last Guess . . . . . . . . . . . . . . . . . 237

18 High-Level Decisions . . . . . . . . . . . . . . . 247

Review Problems . . . . . . . . . . . . . . . . 259

IV Defensive Bidding

19 Entering The Auction . . . . . . . . . . . . . . . 281

20 Preempts . . . . . . . . . . . . . . . . . . . . 297

21 Penalty Doubles . . . . . . . . . . . . . . . . . 311

22 Sacrifice Bidding . . . . . . . . . . . . . . . . . 325

Review Problems . . . . . . . . . . . . . . . . 333

V The Play

23 Opening Leads . . . . . . . . . . . . . . . . . . 355

24 Declarer Play . . . . . . . . . . . . . . . . . . 367

25 Defense . . . . . . . . . . . . . . . . . . . . . 385

Review Problems . . . . . . . . . . . . . . . . 397

 INTRODUCTION

Many events at bridge tournaments are played at matchpoint scoring.

Matchpoints is popular because it is only necessary to get a partner
rather than a team, and there is enough luck so that a weaker player has
a chance to win on a good day.
The average player who plays in a matchpoint event usually
understands the basic differences between matchpoints and IMPs,
such as the importance of overtricks, the −200 kiss-of-death score,
and the necessity for competing in low-level partscore battles. This
is satisfactory for getting reasonable results, but more is necessary
to achieve maximum scores from the bidding and play. Winning
matchpoint tactics are quite unique, and those players who understand
these tactics have a distinct advantage over other players regardless of
playing ability.
This book explains these winning tactics in terms the average player
can understand. Expert matchpoint players have been using them for
years, even though many of these experts don’t fully understand the
reasons why these actions are successful. Any player who understands
and successfully applies the concepts in this book will significantly
improve his results in pair games.
The book is divided into five sections. The first section is on
general concepts, and explains the overall goals that a matchpoint
player is trying to achieve with his choices of bids and plays. It
forms the basis from which the rest of the book is built. The next
three sections are on constructive bidding, competitive bidding, and
defensive bidding. There is a fair amount of overlap between these
sections; for example, a constructive auction may suddenly become
competitive. The competitive bidding section is the most important, for
as we shall see, it is in competitive auctions that most of the matchpoints
are won and lost. The final section is on the play of the hand, and
explains the thinking process declarer and defenders must go through
at matchpoints. Throughout these chapters, the important concepts are
explained and then illustrated with several examples, with each example
 MATCHPOINTS

fully analyzed. In this way, the reader is shown how to approach a

bidding or play problem at matchpoints.
This book is not a presentation of a bidding system or philosophy.
While some parts of the book may not be applicable to certain bidding
methods, for the most part the concepts can be used with any bidding
system or style. Bidding systems are used to gather the information
with which to make the final decision. The emphasis in this book is
on making that final decision correctly and, equally important, making
it difficult for your opponents to make the correct decision. No doubt
several of my personal preferences will slip into the discussion, but the
major points are valid regardless of what bidding system or style is in
use.
The concepts presented here are the result of years of analysis,
and they have survived the test of many sessions of tournament play.
In very bad fields, such as some club games, or in unusually strong
fields, such as the finals of the Blue Ribbon Pairs, some of the ideas
may need modification, but most of them are valid for any pair game.
Nothing in this book is gospel, and several of the recommended bids
are based on my assessment of a hand, which may be quite different
from yours. This is particularly true of the review problems at the end
of each section. There is not necessarily a right or wrong answer to
any bidding situation. Top players whose bidding judgment I have the
greatest respect for disagree with me on several of the answers. Our
thinking process for analyzing the situations is the same, but a different
evaluation of the hand can lead to a different conclusion. Many of the
decisions are quite close. It is not so much the answer you arrive at that
is important; how you came to your final conclusion is what matters.
Learning to think along the suggested lines will improve the results of
any player, regardless of his skill level.

xii
 INTRODUCTION

THE SECOND EDITION

There have been many changes in bridge since Matchpoints was

originally published in 1982, but the basic strategies outlined here
are still valid. The new book has been updated to take into account
modern bidding theory, but the concepts remain the same regardless
of system. Unless otherwise specified, bidding sequences assume a
modern Standard American system, with 5-card majors, 15–17 HCP
1NT openings, and game-forcing 2/1 responses. Any unusual bids are
denoted and explained below the auction. We have indicated common
conventions with an asterisk (*).

Kit Woolsey
2015

xiii
 Part I

General Concepts
 THE COST OF BEING WRONG

Bridge is a game of decisions. It differs from a game such as chess in

that the element of chance affects most of these decisions. In chess, the
consequences of any decision can, in principle, be calculated to the last
detail; there is no factor of the game that is hidden from the player.
In bridge, on the other hand, the location of the unseen cards is an
unknown factor. Unless a player can find a line of play or defense that
guarantees that the maximum number of tricks available will be taken
regardless of the lie of the unseen cards, he has a probability of getting
less than the best available score regardless of which play he chooses.
When a player makes a bid or play that produces less than the
optimal result on the hand, he is said to have made a wrong decision.
This does not mean the player actually made a mistake; only that his
decision turned out wrong. If you take a finesse with QJ109 opposite
Axxx you are certainly making the correct play, but if you lose to a
singleton king you have made the wrong decision, for the alternative
play of going up ace does a trick better. If you open 1NT with ♠ J1074
r QJ7 q AQ5 ♣ KQJ and play it there going down 1 trick when 2 ♠
in the 4–4 fit is cold you made a wrong decision with your choice of
opening bid, even though it is certainly the correct call. The alternative
of opening one of a suit would have been the winning action this time.
When we talk about a wrong decision, we mean a decision different
from what you would have done had you been able to see all 52 cards.
The nature of bridge is such that in a given session any player
will make several wrong decisions, simply because he can’t see his
opponents’ or his partner’s cards. It would seem that the goal would
be to make as few wrong decisions as possible. This would indicate
that, for any decision, the bid or play that has the greatest chance of
being right should be chosen. However, this is not always correct. The
reason is that some wrong decisions are more costly than others.
Let’s look at a simple illustration of this principle from rubber
bridge:
 MATCHPOINTS

♠ 654
r A8
Both vul.
q AK85
♣ Q842

♠ A 10 8
r KJ4
q Q74
♣ AK75

West North East South

— — — 1NT
Pass 3NT Pass Pass
Pass

Opening lead: ♠ K.

You duck the first trick, but win the spade continuation as East shows
out, pitching a heart. When you try the minor suits they both spit
badly: West started with four diamonds and East with four clubs. West
discards spades on the second and third rounds of clubs, and East
pitches a heart on the third round of diamonds. Now you lead a heart
to the ace and a heart from dummy. Do you finesse or not? This is a
silly question. Nobody in their right mind would risk a game contract
at rubber bridge for an overtrick. A count of West’s hand shows he has
6 spades, 4 diamonds, 1 club, and therefore 2 hearts. If you finesse and
lose to his doubleton r Q you are down, while the r K is your ninth
trick.
The obvious play of guaranteeing the contract will be wrong 75%
of the time, while the heart finesse will be wrong only 25% of the time.
Since East started with 6 hearts and East only 2, East is a 3-to-1 favorite
to hold the r Q. So why do we all choose the play that is likely to be
wrong? Because the cost of being wrong by failing to take the finesse
is a mere overtrick, while the cost of being wrong by losing to the
doubleton queen is a game contract.
What we are trying to do when we play bridge (or any game of
imperfect knowledge) is to minimize the expected or average cost of
being wrong, rather than the likelihood of being wrong. The cost of

4
 THE COST OF BEING WRONG

being wrong is a function of both the probability of being wrong and

the difference between being wrong and being right. To calculate this
cost for a particular action, calculate the difference in the score you
would receive for taking action if it proves to be wrong compared to the
optimal score on the hand—the score you would have received had you
taken the winning action. Then multiply this difference in score by the
probability of being wrong.
Because it measures how an action will affect your score, the cost of
being wrong depends on the scoring of the game, which explains why
one bid or play might be correct at matchpoints while an alternative
action is superior at IMPs or rubber bridge. In the previous example
we lose an overtrick worth 30 points if we play the king and are wrong,
while we lose the contract, game bonus, and a penalty for a total loss of
about 700 points if we finesse and are wrong. So the average costs of
the two plays are:

The king: .75 X 30 points = 221⁄2 points.

The jack: .25 X 700 points = 175 points.

This calculation shows what we all know—that when maximizing your

score is your primary concern, as it is at IMPs, the king is clearly the
better play, even though it is more likely to be wrong.
This type of reasoning can be applied to everyday decisions in life;
people go through this analysis on a regular basis, even if they don’t
realize they are doing so. One good example is the decision of whether
or not to buy health insurance. Obviously the decision not to buy is
more likely to be correct than the decision to buy, since buying is right
only if you have a serious illness that year. The cost of being wrong
by not buying may be so high, even including possible bankruptcy, that
most people choose to pay the cost of buying even though they know
that they are probably wrong. Another real-life decision is whether or
not to exceed the speed limit. Since you are wrong in speeding only
when you are caught, which is not likely, and since the cost is only
moderate, say $150, many people choose to speed rather than pay the
small cost of not speeding (being a little late or driving more slowly
than they like). If these same people were driving through a known
speed trap, with increased likelihood of getting caught, or if the fine for
speeding were very high, say $5,000, the increase in expected cost of

5
 MATCHPOINTS

being wrong by speeding would cause speeders to stay under the speed
limit.
This approach of estimating the expected cost can be very helpful
when you are trying to find the best percentage play on a hand. For
example:

♠ Q73
r K2
E-W vul.
q AQ4
♣ AK754

♠ AKJ42
r QJ6
q J83
♣ 93

West North East South

— — — 1♠
Pass 2♣ Pass 2♠
Pass 4NT ∗ Pass 5r ∗
Pass 6♠ Pass Pass
Pass

Opening lead: r A.

West leads the r A and shifts to a diamond. This is clearly his best
defense, for you are forced to decide whether or not to take the diamond
finesse before testing the clubs to see if they split 3–3. The probability
of a 3–3 club split is 36%. We analyze the costs of the alternative plays
as follows: If you take the diamond finesse, you will be wrong when
East has the q K. If this occurs you will cost the contract 36% of the
time; namely when the clubs split 3–3. If you rise with the q A you
will be wrong when West has the q K. If this occurs you will cost the
contract 64% of the time; when clubs aren’t 3–3. Consequently, we
must ask the question: Is East a 64-to-36 or better favorite to hold the
q K? If so, we play for the 3–3 club split, if not, we take the diamond
finesse.

6
 THE COST OF BEING WRONG

This question is, unfortunately, not too easy to answer. West has an
obvious diamond shift if he doesn’t hold the q K, but if he does he might
be wary of the shift for fear that South holds ♠ AKJxx r QJx q Jx
♣ xxx and only the diamond shift allows the contract to be made, for
declarer would have no choice but to finesse and later pitch his losing
club on the good diamond. Your decision would be based on your table
feel and your judgment of West as a tricky or straightforward player, but
at least we know what probability figure to use after this judgment has
been made. Note that if dummy has the q 10 instead of a small diamond
the diamond shift away from the q K would be quite safe, so the finesse
should be taken unless we judged that this particular opponent could
not work out that the shift is safe.

♠ rq ♣ ♠ rq ♣

Throughout the rest of this book, as we examine different situations

and problems, keep in mind that our goal is to minimize the cost of
being wrong. The payment is in matchpoints, so the question to be
asked about at a possible bid or play is: What is the probability of
this bid or play being wrong, and how many matchpoints will it cost
(compared to the winning action) if it is wrong? The product of these
figures is the expected or average cost of being wrong for a given play
or bid, and we want to choose the action that has the smallest expected
cost. The determination of these figures can be quite difficult. Proper
hand evaluation is needed to estimate the likelihood of being wrong,
and understanding matchpoint scoring and what the rest of the field is
likely to do is necessary to determine the cost of the action if it is wrong.
No matter what type of decision is under consideration, minimizing this
expected cost of being wrong is our ultimate goal, so this chapter really
forms the basis for the rest of the book.

7
 THE MATHEMATICS OF
MATCHPOINTS

Matchpoint scoring is very simple: a pair scores one matchpoint for

every other pair holding the same cards whom they outscore on a board,
and 1⁄2 matchpoint for every pair they tie. Every tournament player is
familiar with the scoring. In the heat of battle, however, it is easy to
forget what the goal is—to garner as many matchpoints as possible on
each hand.
Let’s see what this means when we have a decision to make. Since
the scoring is in matchpoints, the cost of making a losing decision is
also measured in matchpoints. Consequently, the expected cost of being
wrong is the probability that the action is wrong times the number of
matchpoints the decision costs when it is, in fact, the wrong decision.
The matchpoint cost is the difference between the matchpoint score you
will receive for your wrong decision and the score you would have
received for the winning decision. This expected cost is what we are
trying to minimize.
It is apparent that some wrong decisions are far more costly than
others, and the cost is not necessarily related to total points scored
on the board. If your opponents step out of line and go for a number
on a mundane partscore hand and you subsequently misdefend, setting
then only 800 instead of 1100, you have lost 300 points but probably
0 matchpoints, since +800 is likely to be a top. On the other hand,
if you misdefend a 3NT contract and allow your opponents to score
630 instead of 600 it is only 30 extra points, but it will cost several
matchpoints if most of the field is in 4 ♠ for +620. One of the primary
goals of this book is to enable the reader to recognize which decisions
will cost a larger number of matchpoints when wrong, and which will
cost only a few matchpoints.
How does one estimate the cost in matchpoints of a losing action? It
isn’t easy, but it can usually be done fairly accurately. Let’s suppose that
we are playing on a 12 top, which is common (a larger top will make
 MATCHPOINTS

the numbers bigger, but it won’t affect the results). This means that the
board is played at 12 other tables against which we will be comparing
our score. Suppose that on a hand we have a choice of two actions, A or
B, and we choose A when B was the winner—i.e., B would have gotten
us a better score. Our cost in matchpoints is as follows:

• For each table where the result falls in between the score
from action A and the score from action B we cost ourselves
1 matchpoint, since we go from beating their score to losing to
their score.

• For each table where the result is exactly either the score
from action A or the score from action B we cost ourselves
1⁄ matchpoint, since we have a tie instead of a win or a loss
2

instead of a tie.

• For each table where the result falls outside the scores from
actions A and B we cost ourselves 0 matchpoints, since either
both actions beat the other score or both actions lose to the other
score. In this case, these tables are no longer in competition with
us for this hand, as far as our choice of actions is concerned.

10
 THE MATHEMATICS OF MATCHPOINTS

Let’s look at several hands that show how this reasoning is put into
practice:

♠ Q8632
r A64
N-S vul.
q K5
♣ KQ4

♠ A J 10 9 5
r 83
q A6
♣ AJ72

West North East South

— — — 1♠
Pass 2NT 1 Pass 3♠ 2
Pass 4♣ Pass 4q
Pass 4r Pass 5♣
Pass 5q Pass 5♠
Pass Pass Pass
1. Forcing raise
2. Good balanced minimum

Opening lead: r K.

A tough bidding decision by North at his final turn, but when South
couldn’t bid 5 r, North decided they were probably off a heart trick
and possibly a spade trick as well, since the ace and king of trumps in
addition to the controls already shown in the minors would give South
a very good hand. Figuring slam was at best on a finesse, he chose to
pass.
His analysis is correct; it is exactly on a finesse. It is not an
unreasonable slam, and several other pairs figure to bid it. Let’s estimate
that at 6 of the other 12 tables the slam will be bid. How should we
play the spade suit, assuming that East is a good enough player to
duck smoothly with ♠ Kx? Many players will say to themselves: “I
will have a bad board if the ♠ K is onside, since other pairs will be in
slam. Therefore, I must hope that the ♠ K is offside, so I will play the
♠ A in case it is offside singleton.”

11
 MATCHPOINTS

Let’s try our matchpoint cost analysis and see what we conclude.
Our possible plays are the finesse or the drop, and the possible scores for
these plays are +450 if wrong and +480 if right. It is clear that the pairs
in slam will score either +980 or −50, which are well out of our range,
so the cost of choosing the wrong play compared to them is zero. What
about the other pairs who stop in game? They, too, will score +450
or +480. Since these match our “right” and “wrong” scores, we lose
1⁄ matchpoint to each of them if we take the losing action, whichever
2

one we take. Consequently, the matchpoint cost of both the finesse and
the drop if wrong is the same—3 matchpoints. Therefore, we should
take the play that is least likely to be wrong.
A full matchpoint chart might help emphasize this point. Suppose,
again, that half of the field bids the slam, and suppose further that the
whole field takes the finesse. Our matchpoints are as follows:

Expected Matchpoints for Finesse vs. Drop

Take the finesse Play for drop
Finesse works 3 0
Singleton king offside 9 12

This table confirms the previous analysis—we gain 3 matchpoints by

playing for the drop and being right, but we lose 3 matchpoints by
playing for the drop and being wrong. Thus, the correct play is to make
the percentage play in spades, which is the finesse.

E-W vul., South holds ♠ 10 r J53 q AQ8743 ♣ KJ9

West North East South

1♠ Pass 2♠ 3q
3♠ 4q 4♠ ?

Should you save? It would be hard to imagine partner’s having a 4 q bid

that won’t provide 8 tricks opposite this hand, so 5 q doubled won’t go
for more than the value of their game. The key question is: Will they
make 4 ♠? This is not so clear. Most players will simply try to judge if
4 ♠ will make, and if they think it is more likely than not to make they

12
 THE MATHEMATICS OF MATCHPOINTS

will save. However, a matchpoint cost analysis shows that this is not the
correct approach. It would be fine if we could be sure that every E-W
pair will bid to 4 ♠, but this is not likely to be the case. Our opponents
were sort of pushed to game (remember that everybody won’t risk an
overcall at the 3-level on q AQxxxx), so it seems reasonable to assume
that at half the other tables game will not be bid—E-W will stop in a
spade partscore or possibly sell to 4 q . We can assume that 5 q doubled
will go for −300 or −500, so our matchpoint costs are as follows:

• If we pass and are wrong we cost ourselves 1⁄2 matchpoint for

each of the 6 tables at which game is bid, since the score there
is −620 (or −300 or −500 if the save is taken there). We cost
nothing against the tables at which game isn’t bid, since −170 is
outside our possible range. Total cost: 3 matchpoints.

• If we save and are wrong we again lose 1⁄2 matchpoint for each
table at which game is bid. In addition we now lose one full
matchpoint for each table at which game is not bid, since −140
is between −300 (or −500) and +100, which are our potential
scores. Total cost: 9 matchpoints.

Therefore, we are risking 9 matchpoints by saving, but only risking

3 matchpoints by passing. The save doesn’t seem justified, for there
must be better than one chance in four that they will go down. A typical
layout might be:

♠ 72
r A64
E-W vul.
q KJ62
♣ 7653
♠ AKJ865 ♠ Q943
N
r K92 r Q 10 8 7
W E
q 10 q 95
S
♣ Q 10 8 ♣ A42

♠ 10
r J53
q AQ8743
♣ KJ9

13
 MATCHPOINTS

West must guess the r J to make 4 ♠, while South will probably go

down 3 in 5 q doubled. A further argument against saving is that your
overcall, which won’t be made at all tables, will get partner off to what
is likely to be the best lead for the defense. Consequently, even if 4 ♠
is cold, you have a chance to gain points from pairs who get off to an
inferior lead that gives away an overtrick. You may be well ahead of
the field simply by having gotten away with the overcall.

Both vul., South holds ♠ K4 r AQ82 q QJ74 ♣ J109

West North East South

Pass Pass Pass 1q
1♠ Double ∗ 2♣ 2r
Pass Pass 2♠ Pass
3♣ Pass Pass ?

Now what? The obvious choices are 3 r, double, and pass. Apparently
the opponents don’t have an 8-card spade fit because they elected not to
play in 2 ♠, and since both opponents are passed hands it is likely that
other tables will not receive as much competition and will buy the hand
for 2 r. A reasonable assessment of both contracts is that hearts will
make 8 or 9 tricks, and 3 ♣ will make or go down 1, with the probability
of each result difficult to estimate accurately. If 3 ♣ is going down,
double is obviously the winning action, while if 3 ♣ is making either
pass or 3 r is better. So what is right?
Let’s compare with the most likely result at other tables, 2 r making
either 2 or 3. It is clear that pass is never the best action available: if
3 ♣ goes down double is the winner, while if 3 ♣ makes 3 r should be
tried. This does not necessarily mean that pass is not the percentage bid;
all it means is that pass is never the double-dummy action. Let’s first
compare double with 3 r. Double is wrong when 3 ♣ makes. In this case
it costs 1⁄2 matchpoint when 3 r also makes, but nothing when 3 r fails,
since +110 is unattainable (remember, we are comparing with a table
that bought the hand for 2 r, the most likely result). The 3 r bid is wrong
when 3 ♣ is going down 2. If this is the case it costs 1⁄2 matchpoint when
3 r is making (we tie the +140 instead of beating it with +200), but
costs a full matchpoint when 3 r is going down (we get −100 instead of
+200). Since both contracts figure to be close, double is clearly better

14
 THE MATHEMATICS OF MATCHPOINTS

than 3 r. What about pass vs. double? It is apparent that pass gains
nothing over double even when it is right, since −110 will be almost as
bad as −670, but pass loses heavily to double when it is wrong, because
+110 (or +140) falls between +100 and +200. Consequently double is
a standout, although this is far from obvious at first glance. The full
hand was:
♠ J973
r K763
Both vul.
q 10 6 2
♣ A3
♠Q 10 6 5 2 ♠ A8
N
r 95 r J 10 4
W E
q A93 q K85
S
♣ K85 ♣ Q7642

♠ K4
r AQ82
q QJ74
♣ J 10 9

3 ♣ doubled goes down 1 trick after a struggle, while hearts makes

9 tricks due to the favorable lie of the E-W hands. Few pairs competed
as vigorously as this E-W pair, so the swing between double and pass
was almost an entire board.

15
 MATCHPOINTS

♠ 7
r K742
Both vul.
q 95
♣ AQ9874

♠ AJ5
r A 10 5
q Q76
♣ KJ32

West North East South

— — — 1NT
Pass 2♠ 1 Pass 2NT 2
Pass 3r Pass 3♠
Pass 3NT Pass Pass
Pass
1. Transfer to clubs
2. Accepting if invitational

Opening lead: q 4.

East plays the q 10 on the opening lead and you grab your q Q. How
nice, a favorable lead for a change! You start to run your clubs, and
East discards the q K(!) on the third club. Not only a favorable lead, but
a clear blunder by East at trick 1. On the last three clubs West pitches
the r 3, r 6, and r J. This can’t really be happening (but it actually did
happen—I could never have made up a story like this). Now you can
guarantee 11 tricks by cashing the r K and, when West shows out as
expected, taking the marked heart finesse. Of course you can garner in
12 tricks by taking the first-round heart finesse. This play is virtually
certain to succeed, for no player in his right mind would pitch like this
from r QJxx. So what should you do? Here is a case where you should
make a play that is almost sure to be wrong, playing the r K, because
the cost of being wrong is probably zero. You have already been handed
2 tricks, and +660 just has to be a complete top, which it was in real
life. The first-round finesse will cost several matchpoints if it happens
to be wrong, for the defense can now cash some of the tricks that were
rightfully theirs. Incidentally, for those of you who are curious, the
first-round finesse would have won.

16
 THE MATHEMATICS OF MATCHPOINTS

♠ rq ♣ ♠ rq ♣

There is one situation where you may choose to take an action that
doesn’t minimize the expected matchpoint cost of being wrong. On the
last round or two of a session you may determine that you must get a
good score on a board in order to have a chance to achieve your goal for
the session, whether it be qualifying for the finals or winning the event.
In this case it may be correct to take an anti-percentage action, such as
trying to drop the stiff king offside on the first hand in this chapter, since
if the slam makes you will not get the necessary good score whatever
you do. However, this is seldom a good policy. Even the best estimators
are frequently one or two boards off in their estimates, simply due to the
innate randomness of bridge players and matchpoints. Consequently, it
is rare that you can determine exactly what score is necessary on a given
board.
It may seem difficult to estimate accurately what the rest of the
field will be doing on a hand, but it is often important to try to make
that estimate. Any time you have what appears to be a close choice
between two alternatives, consider the possible consequences of these
alternatives and compare them with likely results at other tables. You
may find that one alternative will cost far fewer matchpoints than the
other if it is wrong, and this will guide you in making your choice.

17
 TWO WAYS TO WIN

When you go to a horse race, you bet on a horse to win. If the horse
wins the race you get paid, otherwise not. Wouldn’t it be nice if, for the
same price, you would also collect on your bet if a second horse in the
race happened to win? Even if this second horse didn’t have much of a
chance, it would still improve your probability of cashing in your bet.
When an expert chooses a bid or play, he is always on the lookout
for that second horse. His action will usually have a primary way to
succeed, but if that fails he likes having another chance. Any action
that has more than one way to win is usually a better bet than one that
puts all the eggs in one basket.
A well known theme illustrating this is a play that combines two
chances, even though each individual chance may not be as good as an
alternative play that only gives one chance. For example, consider the
following play problem at IMPs:
 MATCHPOINTS

♠ 86
r J965
Both vul.
q KJ743
♣ 72

♠ A2
r A72
q A52
♣ A K J 10 3

West North East South

— — — 2NT
Pass 3♣ ∗ Pass 3q ∗
Pass 3NT Pass Pass
Pass

Opening lead: ♠ Q.

Declarer has 6 top tricks, and he needs to produce 3 more tricks from
the minor suits without losing the lead. Because there is only one entry
to dummy the diamond suit offers better prospects than the club suit,
since queen doubleton or tripleton onside is more likely with five cards
outstanding than with six outstanding. However, the astute declarer who
wants to bet on more than one horse will first try to drop the doubleton
q Q, and if that doesn’t work he will resort to the club finesse. The
chance of the q Q dropping doubleton combined with bringing in the
club suit is better than betting everything on the diamond suit. The
really careful declarer who likes to have as many horses running for
him as possible will time the play as follows:
a. Cash the q A. If the q Q drops singleton he is up to 7 tricks, and
now has two entries to dummy to take two club finesses for his
contract. If not:
b. Cash the ♣ A. If the ♣ Q drops singleton he is up to 8 tricks,
and simply takes the percentage finesse in diamonds for the ninth
trick. If not:
c. Lead a diamond to the q K. If the q Q comes down, claim. If not:
d. Take the club finesse, and hope for the best in that suit.

20
 TWO WAYS TO WIN

♠ rq ♣ ♠ rq ♣

Few players realize that it is possible to bet on more than one horse
in the bidding, even though they do it all the time. Even a lowly
1 ♠ opening bid on ♠ AQJ53 r K98 q 103 ♣ K106 has several things
going for it. The primary reason for opening is to start a constructive
auction to get to the best contract. However, if it is the opponents’ hand
or a competitive auction, the 1 ♠ opening bid has the following horses
in the running:

a. The auction starts at a level at which it may be inconvenient for

the opponents to bid constructively.

b. A good sacrifice might be reached.

c. Partner may get off to the best opening lead.

d. The 1 ♠ opening puts us in command in a competitive partscore

fight.

Naturally a player does not bother to think about all these things when
he picks up this hand, since the 1 ♠ opening is so automatic. Anyone
can make automatic bids; it is the right decision on borderline hands
that separates the winners from the losers. As we shall see later, the
previously mentioned factors do come into play when deciding on a
borderline action.

♠ rq ♣ ♠ rq ♣

Let’s consider some common bidding sequences that leave the bidder
with two ways to win:
You open with a 3 r preempt. Your main hope is that it is the
opponents’ hand (or that they think it is their hand), and that they get to
a bad contract because the preempt robs them of the necessary room to
investigate. It is not so good if your partner holds the strong hand, but
you still might win if he is able to place the contract accurately with the
help of the preempt. The bad case is when you are dropped in 3 r or, still
worse, doubled and left there. However, even if you go for a number
you still have one last horse to root for—maybe the opponents have a
21
 MATCHPOINTS

game or slam that outscores the number they collect on defense. This
explains why you should tend to preempt less with outside strength.
That outside strength may be just enough to stop the opponents from
making their game or slam, thus eliminating one of your horses from
the race.
The opponents stop in 2 r, and you risk a 2 ♠ balance. Your main
hope is to push them to 3 r, giving you a free shot at a plus score if only
8 tricks are available. However even if they double you, there is still the
chance that 2 ♠ will make or, as a last resort, if you are not vulnerable
you may get out for −100 when they can make their partscore. If you
are vulnerable that last horse is eliminated from the race once they apply
the ax.
You double the opponents’ Stayman bid with ♣ KQ109x. Your
main chance for gain is that they ignore you and bid on to their normal
contract, and that your double gets partner off to the best lead. However,
if they stop off to redouble 2 ♣, your strong club holding keeps open the
additional possibility that they just might not make it. Weaken your
club holding just a bit by making the ♣ 9 a small club, and the chances
of beating them when they redouble diminish considerably.
It must be noted that your extra chance to win often depends on
the opponents misjudging a competitive auction. Your opponents need
not be weak players; strong players will often go wrong if they are
presented with difficult bidding problems. The most difficult players to
play against are those who consistently give their opponents problems
that don’t have clear-cut solutions. What these players are doing, in
effect, is making bids that have two ways to win—whichever way
their opponent goes he might be wrong. If a competitive bid leaves
an opponent no choice then the bid can win only on its own merit: it
doesn’t have the additional chance of misjudgment by the opponents.

♠ rq ♣ ♠ rq ♣

Let’s now look at some actual examples of this concept. In each case,
look for the action that gives you extra chances to win by giving the
opponents problems.

22
 TWO WAYS TO WIN

N-S vul., South holds ♠ AQ8 r AQ10763 q 7 ♣ AJ5

West North East South

— — — 1r
3q 4r 5q ?

What now? Clearly you aren’t going to settle for a penalty against 5 q ,
since 11 tricks should be safe at hearts and you probably can’t get them
for 800. What about slam? Who knows? It might be cold if North
has ♠ KJxx r KJxx q xx ♣ Kxx or it might have no play if North
has ♠ KJxx r KJxx q x ♣ xxxx. If you don’t know, neither do your
opponents, and they certainly have a good save if your slam makes.
Therefore, you should just bid 6 r. You have two ways to win. Your
main bet is that slam will make, which it might very well. Even if the
slam is going down, you always have the possibility that the opponents
may misjudge and take a phantom save, and this second horse turns the
borderline decision to bid the slam into a winning percentage action.
Note that bidding this way doesn’t make it easy for the opponents to
decide, for you might choose the same action when you know slam is
cold. The full hand is:
♠ K962
r KJ85
N-S vul.
q2
♣ Q642
♠ 10 7 5 4
N
r 42
W E
♠ J3 q A 10 6 5
S
♣ K98
r9
q KQJ9843
♣ 10 7 3
♠ AQ8
r A Q 10 7 6 3
q7
♣ AJ5

It turns out that the slam goes down, but how would you like to be in
East’s shoes trying to determine that? In practice East took the phantom
save, and N-S collected +1100 and a well-deserved good board.

23
 MATCHPOINTS

N-S vul., South holds ♠ 52 r A1082 q QJ2 ♣ QJ107

West North East South

— — 1q Pass
1♠ Pass 2q Pass
3q Pass 3NT ?

Should South double? It looks close. The opponents are in a

marginal contract, they are not getting a favorable split in their long
suit (diamonds), South has a good opening lead, and any spade strength
North has lies over the spade bidder. On the other hand, South doesn’t
have a long suit to run; he just has to hope the opponents don’t have
9 tricks. If the double is wrong it will cost 1⁄2 matchpoint to other tables
in 3NT, if right it will gain 1⁄2 matchpoint from these same tables, and
there will be no effect on the score against tables where game isn’t bid.
From this analysis it seems that South should double if he thinks 3NT is
a favorite to go down, which makes his decision borderline. However,
the double has another way to win. The opponents can’t be sure that
South doesn’t have something like ♠ xx r Kxx q QJx ♣ KQJ10x, so
they might go wrong and run to 4 q when 3NT is cold. This second
horse makes the close double a percentage action. It might be argued
that South would not double with the solid club suit for fear of driving
the opponents out of 3NT, but if South is the type of player who will
double with both the solid club suit and the actual hand the opponents
won’t know where they are. The full hand was:

♠ KQ964
r J65
N-S vul.
q6
♣ 9543
♠ A 10 8 7 ♠ J3
N
r 743 r KQ9
W E
q K73 q A 10 9 8 5 4
S
♣ K86 ♣ A2

♠ 52
r A 10 8 2
q QJ2
♣ Q J 10 7

24
 TWO WAYS TO WIN

As you can see, 3NT is cold. East was nervous about his club holding,
however, and he ran to 4 q . This contract went down 1, a huge victory
for the double, which gained 1⁄2 matchpoint from the other tables in 3NT
and a full matchpoint from tables stopping in two or three diamonds.
The second horse paid off quite nicely.

N-S vul., South holds ♠ KJ3 r AQ103 q K105 ♣ J85

West North East South

— 1♣ Pass ?

If N-S have agreed that 1m-2NT is forcing, a 2NT response by South,

bypassing the heart suit, has a lot going for it. The hand is described
fairly accurately in one bid, notrump is played from what is probably
the right side, the heart suit is concealed from the opponents possibly
resulting in a favorable opening lead, and the non-vulnerable opponents
are frozen out of the auction. If N-S do not have a 4–4 heart fit,
2NT is almost certainly the winning response, so this is South’s main
bet if he makes the bid. However, the race is not necessarily lost if
North has four hearts. While a 4–4 heart fit will probably produce an
extra trick, South’s hand is balanced enough that 3NT could still be
the winning contract; for example, North might hold ♠ Qxx r KJxx
q AJ ♣ K10xx. The possibility of 3NT being right if North does have
four hearts combined with the clear gain if he doesn’t makes 2NT the
winning action in my opinion. If South had one more diamond and
one fewer club, the likelihood of taking the same number of tricks at
notrump as in a 4–4 heart fit would decrease considerably, so the 2NT
response loses much of its appeal since the second way to win is very
unlikely.

25
 MATCHPOINTS

N-S vul., South holds ♠ AQJ53 r K84 q 7 ♣ Q852

West North East South

— — 1q 1♠
1NT 2♠ Pass Pass
3q Pass Pass ?

It might not be immediately obvious, but pass is the bid that has two
ways to win. The problem with bidding 3 ♠ is that it puts all your
eggs into one basket. West, with his probable trump trick and fair hand
opposite his partner’s opening bid, is virtually certain to double you,
and if you go down you are −200 for a bottom. Are you willing to bet
everything on 3 ♠ making knowing that the ♠ K is offside? I’m not so
sure that I am. Passing, on the other hand, has two things going for it.
In addition to the possibility of 3 ♠ going for 200, you have the chance
that the opponents may be in the wrong contract—notrump might score
more. A look at the actual hand will show how right passing can be:

♠ 10 9 8 7
r Q J 10
N-S vul.
q A54
♣ 964
♠ K64 ♠ 2
N
r 652 r A973
W E
q KJ92 q Q 10 8 6 3
S
♣ K73 ♣ A J 10

♠ AQJ53
r K84
q7
♣ Q852

As you can see, notrump scores +120 for E-W, while diamonds scores
only +110. West certainly would have doubled 3 ♠, and this contract
goes down 1, −200, the worst possible result. Selling to 3 q was worth
a well-above-average score to N-S.

26
 LOADING THE DICE

All bridge players will occasionally make non-standard or unorthodox

bids. These range from mild deviations such as bypassing a 4-card
major when responding to partner’s opening minor-suit bid or opening
1NT with a 5-card major, to outlandish bids such as a preempt on a
jack-high suit, opening 1NT with a singleton, or an out-and-out psychic
bid. While it is possible to win without ever taking such actions, offbeat
bids should be part of a successful player’s repertoire. The trick is to
know when to use them. Many players simply make these bids when
the mood strikes them or when their “table feel” tells them that the time
is right. There is a much better approach, which can turn these bids into
winning percentage actions.
All non-standard actions come under the more general category of
“taking a position.” A position could be defined as an action that will
lead to an irrecoverably bad board if it turns out poorly. Examples of
this are bidding a thin minor-suit slam that the field won’t find, doubling
the opponents at a low level in the hopes that the penalty is greater
than an available sure game, or passing out an opening one-bid in the
balancing seat with a fair hand. All these actions risk very bad boards if
they are wrong; the justification for them can only be that we think they
are more likely to be right than wrong. In other words, we are willing
to shoot craps on a board if we think the dice are loaded in our favor.
Now, how can we load the dice in our favor on offbeat actions?
The previous chapter gave us the clue. We saw the importance of a
bid having two ways to win, so it is equally important that our bids
don’t give the opponents two ways to win. It is clear that an offbeat
bid is in grave danger of getting killed in some way—otherwise, it
would probably be the standard bid. The key to making these bids
winning actions is to use them only when everything else about the
hand is perfect for the bid. If anything else is the least bit wrong, wait
for another time. You just can’t afford to give the opponents a second
way to win on your questionable actions. I call this the Double-Flaw
 MATCHPOINTS

Principle. If a hand is doubly flawed for a bid, try to find some other
call.
Next time you are considering opening 1NT with a 5-card major,
take another look at your hand. Do you have any worthless doubletons?
If so, that is a second flaw, and the 1NT bid should not be made.
Considering a shaky 2-level overcall? Save it for when you are not
vulnerable, and make sure that you want that suit led. It would be silly
if you got away with the overcall, only to get partner off to the wrong
lead. Pondering a really light preempt? Make sure that the vulnerability
is favorable, and no queen-doubletons on the side, please. It would be
a shame if the opponents doubled you and you got out for less than the
value of their game, only to find that their game doesn’t make because
of that queen.
The following hand, which I held some time ago, is a perfect
illustration of the Double-Flaw Principle. It came up in an IMP match,
but the principle remains the same. As South, non-vul. vs. vul., I
held ♠ — r 3 q 109862 ♣ J1076542, and it was two passes to me.
Bids such as 3 ♣ and 4 ♣ would probably be normal in this situation, but
I decided to try something really outlandish—a psychic 1NT opening
bid! There was the danger inherent in any psychic bid that partner might
not read it in time, but the bid seemed to have everything else going
for it. The opponents have at least a game, the vulnerability is right,
partner is limited so he won’t go leaping to a slam, and if worst comes
to worst my runout might not be too expensive. However, there is a
hidden second flaw in the hand. Do you see it?
The psyche worked as well as could be expected. West had a
balanced 27(!)-count, and had no idea how to handle this situation. The
bidding proceeded:

West North East South

— Pass Pass 1NT
Double 2r Double 3♣
Double Pass Pass Pass

Down only 2, −300, when the opponents were cold for a slam. So
what happened? We lost 9 IMPs on the hand! Our teammates bid to
a very reasonable grand slam in spades, but it had no play because
North held ♠ J109x. The second flaw was the spade void, which was an

28
 LOADING THE DICE

indication that suits may not be splitting well for the opponents, so that
what happened at the other table was not that unlikely. One could say
that “the operation was successful but the patient died.” This is what
we are trying to avoid. If the operation is risky, it is important to be
sure that the patient will live if the operation succeeds; otherwise, the
risk of the operation is not worthwhile. A better hand for this particular
psyche would have been something like ♠ xx r xx q x ♣ J10xxxxxx.
If I had held this hand, I could be fairly confident that my teammates
could make whatever they bid.
Let’s look at a few examples and see how the dice are loaded:

None vul., South holds ♠ A754 r K103 q 93 ♣ KQ106

West North East South

— 1NT Pass ?

Is this the hand to bypass the possible 4–4 spade fit and go straight to
3NT? As we shall see later, the hand meets many of the requirements—
extra strength, lack of intermediates in the trump suit, and a strong
side 4-card suit. But what about that diamond holding? A worthless
doubleton is very dangerous, for a diamond ruff will produce an extra
trick if North has diamond length, and if he doesn’t the defenders may
be able to run the suit. The small doubleton in diamonds is the second
flaw that should convince South to take the normal action of using
Stayman. Change the hand to ♠ A754 r K103 q Q9 ♣ K1063 and that
second flaw is taken care of, so 3NT figures to be the winning action.

None vul., South holds ♠ 4 r 84 q J1085 ♣ KQ10952

West North East South

— — Pass ?

Feel like trying a 3-level preempt on a 6-bagger? Vulnerability is

probably OK, suit is excellent, side 4-card minor is not bad. Pity that
we are in second position. Partner is as likely to hold the strong hand
as West, and this flaw makes the opening a losing action. In third seat
it would be clear-cut, and in first seat it would be a winner more often
than not, but not in second seat. With ♠ 4 r 84 q J108 ♣ KQ109532 it

29
 MATCHPOINTS

would be correct. The hand has no flaws at all, so a second-seat preempt

is called for.

None vul., South holds ♠ AQ1062 r KJ105 q 963 ♣ 3

West North East South

— — — ?

This hand is about 1 to 1⁄2 HCP below my normal minimum opening

bid standards (if your standards differ substantially, add or subtract a
jack to make the hand the same relative to your framework). Is a super-
light first seat opening bid likely to be a winner? I believe it is. The
suits are strong, the majors are held, there are no rebid problems, and
the vulnerability is best for competitive action. You may get buried if
partner with a 12-count and no great fit drives to a no-play game, but
you will come out ahead in most variations. Any small change in the
hand, such as removing the ♠ 10 or reversing the major suits (causing
rebid problems) would make the light opening bid a likely losing action,
but this hand has no second flaw so the 1 ♠ opening bid is best.

Both vul., South holds ♠ K1053 r K q AQ108 ♣ KJ92

West North East South

— — — ?

Is this the hand for the feared 1NT opening with the singleton? There
are several good features. The hand is very notrump oriented, there
are several tenaces, the stiff king is ideal, and the strength is just right.
There are, however, two flaws. One is that there is no serious rebid
problem if you open 1 q and partner responds 1 r, so the clever 1NT
opening isn’t necessary. Also, if partner transfers to hearts how happy
will you be? Consequently, you should wait for another time. I would
certainly try it if the stiff king were in clubs, since both of these flaws
would be eliminated, and probably try it if the stiff king were in spades
(despite the risk of a transfer to spades) because the rebid problem after
a 1 ♠ response to 1 q is so severe.

30
 LOADING THE DICE

None vul., South holds ♠ K852 r 6 q QJ109764 ♣ 5

West North East South

— — — ?

The diamond suit, vulnerability, and position are all perfect for a 3 q
bid; the only question mark is the side 4-card major. If this were
the only problem I would take my chances with the preempt, but it
is worse than that. The side major is spades, which outranks the
opponents’ potential heart fit. We may need the spade suit to outbid
them. If the hand were either ♠ 6 r K852 q QJ109764 ♣ 5 or ♠ K852
r QJ109764 q 6 ♣ 5 I would go ahead with the preempt. The danger
of losing the best major-suit fit still exists, but at least our preempt suit
is as good as our side 4-card major for competing with the opponents’
potential fit. The actual hand was:

♠ AJ976
r 10 9
None vul.
q 832
♣ KJ7
♠ Q 10 3
N
r A85
W E
♠ 4 q5
S
♣ 10 9 6 4 3 2
r KQJ7432
q AK
♣ AQ8
♠ K852
r6
q Q J 10 9 7 6 4
♣ 5

4 ♠ is a good save over the E-W 4 r contract, and would have been
reached if South had passed and North had overcalled the 1 r opening
bid with 1 ♠.
Unfortunately, after the preempt the bidding proceeded:

31
 MATCHPOINTS

West North East South

— — — 3q
4r 4♠ Double Pass
Pass 5q Pass Pass
Double Pass Pass Pass

North tested the water with 4 ♠, but when he got doubled it seemed best
to run to the safety of the diamond suit. As you can see 5 q doubled
goes down 1 too many.

E-W vul., South holds ♠ QJ10983 r 7 q 6532 ♣ 96

West North East South

— Pass 1q ?

Any red-blooded bridge player would bid 2 ♠, but how many would
try 3 ♠? Yet, except for the lack of a seventh spade, the hand and the
conditions are perfect. Partner is a passed hand so the opponents surely
have at least a game, the vulnerability is right, and South has no outside
garbage. Sure, South might go for more than the value of a game, but
the opponents don’t know this. Even if 3 ♠ is set 800 there is still the
chance that E-W have a slam. If E-W don’t double, the extra level of
bidding consumed will hurt their constructive auction considerably, and
this will more than compensate for the few times that South is caught
and goes for too large a number. If the hand had any flaws, such as
♣ Qx or partner’s not being a passed hand, then the risk of the 3 ♠ bid
would not be worthwhile.

32
 GENERAL CONCEPTS
REVIEW PROBLEMS

Problem 1
You must walk a half mile to and from work. The
morning weather forecast predicts a 10% probability
of thunderstorms that afternoon. Should you take an
umbrella?
 MATCHPOINTS

Problem 1 Solution
You must walk a half mile to and from work. The
morning weather forecast predicts a 10% probability
of thunderstorms that afternoon. Should you take an
umbrella?

The expected cost of being wrong by not taking the umbrella is .10 X
cost of getting drenched. The expected cost of being wrong by taking
the umbrella is .90 X cost of unnecessarily lugging an umbrella around.
So it’s up to you. If getting drenched is more than 9 times as costly as
the nuisance of carrying an umbrella that isn’t needed, you should take
the umbrella; if not, leave it at home.

34
GENERAL CONCEPTS REVIEW PROBLEMS

Problem 2

Both vul., South holds ♠ A6 r AK86 q AQJ7 ♣ J87

West North East South

— — — 1q
Pass 1♠ Pass 2NT
Pass 3q Pass 3r
Pass 4r Pass ?

35
 MATCHPOINTS

Problem 2 Solution

Both vul., South holds ♠ A6 r AK86 q AQJ7 ♣ J87

West North East South

— — — 1q
Pass 1♠ Pass 2NT
Pass 3q Pass 3r
Pass 4r Pass ?

South has two main choices. He can pass the 4 r bid and play the 4–3
fit, or he can move on to a possible slam starting with a 4 ♠ control
bid. The 4–3 heart fit looks like it will play pretty well, with the strong
trump suit and the short hand able to take the ruffs. Some likely field
contracts are 3NT and 5 q , and 4 r figures to outscore both of them.
Consequently, it may not be necessary to bid the slam even if it is there.
If North holds the perfect hand such as ♠ KQxxx r Qxx q Kxxx ♣ x
then failing to bid the slam will cost 1⁄2 matchpoint against only those
pairs in 4 r or 6 q ; it will cost nothing against those in 3NT or 5 q since
the 4 r contract will beat them anyway. On the other hand, if North has
♠ KQxx r Qxx q Kxxx ♣ xx (his actual hand), 4 r is the best contract;
bidding 5 q costs 1⁄2 matchpoint to almost every other table, and bidding
6 q costs even more. Pass figures to be the winning action on balance.

36
GENERAL CONCEPTS REVIEW PROBLEMS

Problem 3

N-S vul., South holds ♠ QJ94 r A763 q QJ ♣ A86

West North East South

1q Pass 1♠ Pass
2♠ Pass 3♣ Pass
3NT Pass 4♠ ?

37
 MATCHPOINTS

Problem 3 Solution

N-S vul., South holds ♠ QJ94 r A763 q QJ ♣ A86

West North East South

1q Pass 1♠ Pass
2♠ Pass 3♣ Pass
3NT Pass 4♠ ?

Should South double? He figures to beat it, as his aces should cash
and his trump holding will produce 2 tricks more often than not. The
bidding indicates that 3NT is a possible contract, and South can tell by
looking at his hand that 3NT is likely to score as many tricks as 4 ♠. If
this is the case, South will lose a full matchpoint to each table in 3NT
by doubling and being wrong. The double also risks a full matchpoint
to tables not in game if it tells declarer how to play the hand to make the
contract. Since the double gains only against other tables in 4 ♠ when it
is right, and only gains 1⁄2 matchpoint from these tables, it is wrong to
double even though South is a favorite to set the contract. The full hand
was:
♠ 7
r 852
q 98762
♣ 10 5 4 3
♠ K 10 3 ♠ A8652
N
r K Q 10 4 r J9
W E
q A 10 5 4 q K3
S
♣ J7 ♣ KQ92

♠ QJ94
r A763
q QJ
♣ A86

38
 GENERAL CONCEPTS REVIEW PROBLEMS

With notrump easily taking 10 tricks and the double guaranteeing that
declarer played the spade suit correctly, the double was very costly.

39
 MATCHPOINTS

Problem 4

None vul., South holds ♠ 103 r J109852 q A5 ♣ AQ2

West North East South

1♠ Pass 2q ?

40
 GENERAL CONCEPTS REVIEW PROBLEMS

Problem 4 Solution

None vul., South holds ♠ 103 r J109852 q A5 ♣ AQ2

West North East South

1♠ Pass 2q ?

It certainly is risky against a two-over-one response, but do the gains

make it worthwhile? No, because South probably doesn’t want a heart
lead. West is likely to be declarer in spades or notrump, and South
would prefer a natural club lead from his partner. This is a very serious
flaw for the already risky overcall, so the overcall should be avoided.

41
 MATCHPOINTS

Problem 5

None vul., South holds ♠ AJ976 r KJ5 q 85 ♣ K53

West North East South

3q Pass Pass ?

42
 GENERAL CONCEPTS REVIEW PROBLEMS

Problem 5 Solution

None vul., South holds ♠ AJ976 r KJ5 q 85 ♣ K53

West North East South

3q Pass Pass ?

This is a common type of problem. South can pass, bid 3 ♠, or double,

and he is acting in the blind. Pass is the most committal action. It
bets everything on the hand’s belonging to the opponents. The 3 ♠ call
gets to spades or notrump when right, but eliminates the possibility of
playing hearts or clubs, or defending. Double is the bid with the most
horses in the running. Best would be if partner passed or bid spades,
but if partner bids 3 r or 4 ♣ South isn’t dead yet, for these could well
be winning contracts if North has a long suit and no spade support. The
double can work out badly when the hand belongs to E-W or when
N-S miss a superior 5–3 spade fit, but it has more ways to win than the
alternatives.

43
 MATCHPOINTS

Problem 6

None vul., South holds ♠ KQJ109 r 10 q 10753 ♣ 1043

West North East South

— — — ?

Should South open with a weak 2 ♠ bid?

44
 GENERAL CONCEPTS REVIEW PROBLEMS

Problem 6 Solution

None vul., South holds ♠ KQJ109 r 10 q 10753 ♣ 1043

West North East South

— — — ?

Should South open with a weak 2 ♠ bid?

A weak two-bid on a 5-bagger? This must be the hand for it. Excellent
suit, nothing on the side, neither vul., first seat. The only flaw is the
lack of a sixth spade, so South should try it.

45
 MATCHPOINTS

Problem 7

Both vul., South holds ♠ 105 r A42 q AQ107 ♣ KQ82

West North East South

— — — 1q
1♠ 3q 1 Pass ?
1. Limit raise

46
 GENERAL CONCEPTS REVIEW PROBLEMS

Problem 7 Solution

Both vul., South holds ♠ 105 r A42 q AQ107 ♣ KQ82

West North East South

— — — 1q
1♠ 3q 1 Pass ?
1. Limit raise

South certainly has the strength to consider moving towards game, and
5 q could be cold if North has, say, ♠ xxx r Kx q KJxx ♣ Axxx.
Consider what is likely to happen at other tables. Many if not most
players holding the South hand will open 1NT. West will probably not
risk entering at the 2-level, and North will undoubtedly jump to 3NT
with his presumed 10–12 points and no 4-card major. This is not likely
to be a success due to West’s spade suit. If 3NT is going down at other
tables South gains nothing by bidding on if 5 q makes, for both +150
and +600 beat −100. South does, however, risk 1⁄2 matchpoint against
the 3NT bidders if he is wrong in bidding a minor-suit game. Therefore,
South should pass and guarantee his plus score.

47
 MATCHPOINTS

Problem 8

Both vul., South holds ♠ K72 r A94 q J742 ♣ 952

West North East South

— — — Pass
1r Pass 1♠ Pass
3r Pass 4♣ Pass
4NT 1 Pass 5q 2 Pass
6NT Pass Pass ?
1. Blackwood
2. 1 ace

48
 GENERAL CONCEPTS REVIEW PROBLEMS

Problem 8 Solution

Both vul., South holds ♠ K72 r A94 q J742 ♣ 952

West North East South

— — — Pass
1r Pass 1♠ Pass
3r Pass 4♣ Pass
4NT 1 Pass 5q 2 Pass
6NT Pass Pass ?
1. Blackwood
2. 1 ace

Should South double for the spade lead? It’s far from clear that a spade
lead is a favorite to beat the contract, since East has only one ace. A
matchpoint cost analysis shows that the double is a big winner. The
reason is that the opponents are in their top spot, so South will be well
below average if 6NT makes, and the spade lead has a good chance of
being the difference between making and down 1. If South doubles and
is wrong he loses 1⁄2 matchpoint to other tables in 6NT, but loses nothing
to tables in any other contract, including 6 r. On the other hand, if
South fails to double and is wrong (i.e., the spade lead is the killer and
North now leads something else), South again loses 1⁄2 matchpoint to
other tables in 6NT, but costs a full matchpoint to tables in any other
contract, of which there figure to be several. Consequently the double
is by far the percentage bid, even if it succeeds somewhat less than half
the time.

49
 MATCHPOINTS

Problem 9

E-W vul., South holds ♠ AJ93 r A3 q A42 ♣ KQJ9

West North East South

— — — 1♣
Pass 1r Pass ?

50
 GENERAL CONCEPTS REVIEW PROBLEMS

Problem 9 Solution

E-W vul., South holds ♠ AJ93 r A3 q A42 ♣ KQJ9

West North East South

— — — 1♣
Pass 1r Pass ?

Time to bypass a 4-card spade suit with a 2NT rebid? Certainly

a potential gain if North doesn’t have four spades, as South would
welcome a spade lead. But what if North does have four spades? Then
4 ♠ is almost certainly worth an extra trick over notrump, for South has
strong spades and a weak diamond holding. Therefore, South’s best
rebid is 2 ♠. Put South’s jack in diamonds rather than spades, giving
him ♠ A9xx r Ax q AJx ♣ KQJx and the 2NT bid is much better,
since now notrump has a reasonable chance to be right even if N-S have
a 4–4 spade fit. The difference can be seen by looking at both hands
opposite something like ♠ 10xxx r QJ9x q 10x ♣ Axx.

51
 MATCHPOINTS

Problem 10

♠ AKJ98
r A4
Both vul.
q A Q 10
♣ 942

♠ 2
r Q752
q J9862
♣ AJ6

West North East South

— 1♠ Pass 1NT
Pass 2NT Pass Pass
Pass

Opening lead: r J.
Plan the play.

52
 GENERAL CONCEPTS REVIEW PROBLEMS

Problem 10 Solution

♠ AKJ98
r A4
Both vul.
q A Q 10
♣ 942

♠ 2
r Q752
q J9862
♣ AJ6

West North East South

— 1♠ Pass 1NT
Pass 2NT Pass Pass
Pass

Opening lead: r J.

Plan the play.

Both North and South bid very conservatively on this deal. There
are many roads to 3NT, and they will be taken at other tables.
Consequently, South can assume that 3NT will be the normal contract,
and he should plan his play accordingly. Overtricks are irrelevant, for
scores at other tables must fall outside the range of +120 to +180.
Undertricks might matter if 3NT goes down at some other tables.
Therefore, South should play absolutely safe for his contract. This
can be done by rising r A and playing on diamonds immediately,
guaranteeing 4 diamonds, 1 heart, 1 club, and 2 spades for 8 tricks. The
danger in ducking the first trick is that East may shift to the ♣ 10 and
shut out declarer’s diamond suit. The cost of being wrong by going up
is only an overtrick or two, which as we have seen is of little value, but
the cost of being wrong by ducking the first trick could be an undertrick
53
 MATCHPOINTS

or two, which might be meaningful if 3NT is being set at other tables.

The full hand:
♠ AKJ98
r A4
q A Q 10
♣ 942
♠ 764 ♠ Q 10 5 3
N
r J 10 9 8 3 r K6
W E
q 53 q K74
S
♣ Q53 ♣ K 10 8 7

♠ 2
r Q752
q J9862
♣ AJ6

The ♣ 10 shift is not that hard for East to find, after which East can
defeat the contract by refusing to win the first round of diamonds.

54
 Part II

Constructive Bidding
 THE BEST GAME

In this chapter, we shall examine situations where the only problem is

which game to bid. Slam is out of the question, and the combined hands
have enough strength that the entire field will certainly bid some game.
In most cases there will be only two choices, usually a major suit or
notrump, so matchpoint cost analysis won’t tell us anything—you will
simply lose 1⁄2 matchpoint to every table if you go wrong. Therefore,
the goal is simply to be right more often than not.
 MATCHPOINTS

THE 4–4 FIT

We all know that is it usually right to prefer a 4–4 major suit to notrump
at matchpoints. The reason is that the major suit will produce an extra
trick more often than not, and that extra trick translates into a better
score. The winning player wants to do better than “usually”: he wants
to be able to recognize hands on which the 4–4 fit will not produce that
extra trick and head for 3NT when these hands occur.
There are three reasons why a 4–4 fit might produce an extra trick
over notrump. They are:

1. A trump suit acts as a stopper so that the opponents can’t run their
long suit as they might in notrump.

2. If one player has a short suit, losers in that suit from the other
hand can be ruffed. The beauty of the 4–4 fit is that the ruff can
be taken in either hand, and the other hand still has the length to
draw trumps. It may only take one such ruff to produce the extra
trick.

3. A trump suit allows for more flexibility in the play, for such
things as end-plays, establishing a long suit by ruffing, etc.

These are pretty powerful arguments for the 4–4 fit. If none of
them produce the extra trick, however, then notrump, with that extra
10 points, becomes the winner. It is clear that if one of the partnership’s
hands contains a singleton the 4–4 fit is virtually always right, just on
the ruffing value alone, except in the rare case when the hand opposite
the singleton is loaded with secondary values such as KQJ9 in the suit.
The more interesting decisions occur when both hands are relatively
balanced. There are four factors a player should consider when deciding
whether or not to reject a 4–4 major-suit fit. They are:

1. Strength. If the partnership has only 25–26 HCP the 4–4 fit
is almost always right, whereas if the partnership has excess
strength, say 29–30 HCP, then notrump has a good chance of
producing the same number of tricks. There are two reasons
for this. First, the extra strength may be sufficient to stop the
opposition from running a suit while declarer is busy setting up
58
 GENERAL CONCEPTS REVIEW PROBLEMS

his winners. Second, the extra strength may produce the winner
that makes it unnecessary to ruff a loser. For example:

♠ K 10 6 5
r Q7
N-S vul.
q KJ74
♣ KJ2

♠ QJ83
r KJ2
q AQ8
♣ A 10 5

West North East South

— — — 1NT
Pass 3NT Pass Pass
Pass

There are 11 tricks available at notrump (3 spades, 2 hearts, 4 diamonds,

and 2 clubs.) Two aces must be lost at spades, so notrump is the winner.
However, a slight weakening of the hands changes the picture. For
example, make South’s ♣ A the ♣ Q; now 4 ♠ is cold for 10 tricks barring
bad splits, while notrump makes only 9 tricks after a heart lead, for
declarer doesn’t have time to knock out both black aces. Try making
North’s minor-suit jacks small cards in the same suits. Spades still
makes 11 tricks on reasonable splits, for North’s losing club can be
pitched on the long heart, but notrump now needs a 3–3 diamond split
to make the same 11 tricks.

2. Holding in short suits. We have seen that the extra trick in the
4–4 fit often comes from a ruff. If the card being ruffed is a
winner, no extra trick is gained. Aces and kings account for the
first two rounds of a suit, but queens and jacks (and often tens
and nines) determine the fate of the rest of the suit. Therefore,
if you have a doubleton it is often necessary for the doubleton to
include the queen or the jack for notrump to be the winner. For
example, opposite a holding such as AJ10x, king-doubleton will
produce no losers in a suit contract because the third round of
the suit can be ruffed, while in notrump a finesse must be taken.
59
 MATCHPOINTS

Queen-doubleton opposite AJ10x, on the other hand, will lose

either one trick or none at either notrump or a suit contract.
3. Holding in side 4-card suits. Exactly the same theme applies
as with doubletons, because the side suit may be ruffed out in
partner’s hand. If you have queens and jacks in the side suit you
may be ruffing winners, but if the suit is empty the ruff may well
be an extra trick. For example:
♠ K953
r A 10 2
Both vul.
q A 10 5 2
♣ Q7

♠ AQ84
r K986
q K4
♣ KJ3

West North East South

— — — 1NT
Pass 2♣ ∗ Pass 2♠
Pass 4♠ Pass Pass
Pass

Despite the extra strength and the balanced nature of the North hand,
his holdings in the red suits are a tip-off that a 4–4 spade fit is likely to
produce an extra trick. The reason is that any shortness in these suits
in partner’s hand can be used to ruff third- and fourth-round losers.
On this pair of hands notrump will make at most 11 tricks, probably
only 10, while spades is almost sure to make 11 tricks and may well
make 12 with good splits, for North’s third heart can be discarded on
the long club. Change North’s hand to ♠ K953 r AJ2 q QJ102 ♣ Q7.
Same strength, but the texture of the red suits now makes notrump a
likely winner. Opposite the same South hand there are 11 tricks at both
spades and notrump. With the second hand North should not bother
with Stayman; he should just bid 3NT.

4. Trump Holding. An 8-card fit will get a 4–1 split almost one-
third of the time. If the trump suit has good intermediates, such
60
 GENERAL CONCEPTS REVIEW PROBLEMS

as K10xx opposite QJxx, the bad split by itself won’t cost a

trump trick, but if the trump suit is empty such as Kxxx opposite
Axxx, that 4–1 trump slit will always cost the major-suit player a
trick. Consequently, the player rejecting the 4–4 fit should prefer
to have a trump holding weak in intermediates. It’s not that this
helps notrump; what is does is give the notrumper an extra way
to win when the suit player loses an extra trick due to a bad trump
split. For example:
♠ A542
r K97
N-S vul.
q K J 10 6
♣ Q3

♠ K873
r A5
q Q84
♣ AK84

West North East South

— — — 1NT
Pass 3NT Pass Pass
Pass

The North hand fits all the conditions for rejecting the 4–4 spade fit.
Unfortunately South produces the wrong heart holding, and there are
only 10 tricks available in notrump after a heart lead while spades easily
makes 11 tricks on a 3–2 split. However, the notrumpers have an extra
way to win even though they judged wrong—the spades might split 4–1.
Obviously you would prefer to be in 4 ♠ on the combined hands, but it
is nice to know that you have that additional chance of a bad spade split
if you opt for 3NT. If North’s spade holding were ♠ QJ10x instead of
♠ Axxx, that extra way to win would go out the window.

♠ rq ♣ ♠ rq ♣

At least three of the four conditions should be met in order to justify

rejecting a 4–4 major-suit fit, and meeting all four conditions is
preferable. In close cases notrump should be preferred if a 4–4 fit is
61
 MATCHPOINTS

only a possibility and the search would lead to a more revealing auction,
for you then gain by concealment when you don’t have the 4–4 fit. In
general, however, the 4–4 fit should be chosen.
Let’s look at a few examples of this type of decision.

Both vul., South holds ♠ KJ63 r K2 q Q1062 ♣ A32

West North East South

— — — 1q
Pass 1♠ Pass 2♠
Pass 3NT Pass ?

South should go to 4 ♠. His spade suit is strong (it contains the ♠ J),
and his heart holding doesn’t include the r Q or the r J, so the 4–4 fit
figures to be better. The heart suit is of particular concern, for if North
holds only four spades he doesn’t have four hearts, or he would have
responded 1 r rather than 1 ♠. Consequently, there is a danger that the
opponents may be able to set up their hearts at notrump.

N-S vul., South holds ♠ AJ r A964 q K76 ♣ QJ32

West North East South

— 1q Pass ?

If N-S are playing a 2NT response as forcing, I consider bypassing

the 4-card heart suit and bidding 2NT a worthwhile gamble. The hand
meets all the qualifications, so there is not likely to be an extra trick
coming from a ruff. There is some danger that the spade suit may be
inadequately stopped, but the extra strength may give South enough
fast tricks anyway. In addition, the 2NT bid will be best when North
doesn’t have four hearts, for South protects his spade tenace and the
N-S distribution isn’t revealed to the opposition.

62
 GENERAL CONCEPTS REVIEW PROBLEMS

Both vul., South holds ♠ 7632 r KQJ105 q KJ ♣ Q6

West North East South

— 1NT Pass ?

South should see that a 4–4 spade fit is likely to be of little value, for his
hearts don’t need ruffs to become established and South has strength in
his doubletons. A 5–4 or 5–3 heart fit, on the other hand, may produce
an extra trick, since South may get to ruff a losing spade. Consequently,
South should not bid Stayman but should instead transfer to hearts and
then bid 3NT, leaving the spade suit unmentioned.

63
 MATCHPOINTS

THE 5–3 FIT

Many of the same considerations are involved in deciding whether or

not to bypass a 5–3 major-suit fit to play notrump. The main difference
is that a ruff is now of value only in the hand with the short trumps.
Also, the hand with the 3-card trump support might have a side suit such
as AKxxx that can be established by ruffing without losing tricks in the
suit, while in notrump a trick or two may have to be lost establishing
the suit.
The hand with the 3-card support is almost always better placed
to make the final decision. If he is 4-3-3-3 notrump will probably
produce the same number of tricks, provided that the combined hands
contain extra strength and that all suits are adequately stopped. If the
hand with 3-card support has a doubleton, however, the 5–3 fit usually
scores an extra trick. For notrump to be right the doubleton suit must
be adequately stopped, yet have no valuable ruff in the short hand, and
there must be no side suit that can be established by ruffing.

Both vul., South holds ♠ K5 r A7632 q Q73 ♣ J84

West North East South

— 1NT Pass ?

Despite the lack of intermediates in hearts and the scattered strength,

South should transfer to hearts and then bid 3NT. There is no extra
strength, and the opening notrumper, who holds the potential 3-card
support, will be better placed to make the final decision.

N-S vul., South holds ♠ A93 r KQ5 q AJ7 ♣ J1075

West North East South

— 1♠ Pass 2♣
Pass 2NT Pass ?

With extra strength, no ruffing values, and adequate stoppers opposite

any doubletons, South should be quite willing to bypass a 5–3 spade fit.
He should raise to 3NT.

64
 GENERAL CONCEPTS REVIEW PROBLEMS

None vul., South holds ♠ Q63 r J4 q AKJ ♣ KJ1062

West North East South

— 1♠ Pass 2♣
Pass 2r Pass ?

Without the 2 r bid South would probably opt to play in spades because
of the heart weakness. After North bids 2 r this is no longer as much of
a problem, as with his super-strong diamond holding and extra strength
South has every reason to believe that notrump will take the same
number of tricks as the major, so he should just bid 3NT and not bother
supporting spades.

65
 MATCHPOINTS

THE LONG SUIT

A common problem occurs when you have a long solid or semi-solid

major suit but an otherwise relatively balanced hand and the partnership
has all suits stopped. Should you gamble that notrump will produce
the same number of tricks, or should you retreat to the safety of the
suit? It is unlikely that dummy has any ruffing value, and usually there
will not be a suit in dummy that can be advantageously established by
ruffing. Timing is the key to the value of the trump suit. The trump
suit serves as a stopper to fend off the enemy attack while you are busy
setting up tricks in the other side suits. If there are no such tricks to
be set up, then notrump will take the same number of tricks. What
this means is that honor cards have their usual roles reversed. Aces,
usually more desirable for suit play, are now preferred at notrump since
they don’t have to be set up. Kings and queens, usually indicative of
a notrump hand, now point toward the suit contract since they require
time to establish.
The following hand is a good illustration:

♠ 3
r A64
E-W vul.
q AJ852
♣ J953

♠ AKQJ842
r 97
q7
♣ A76

West North East South

— Pass Pass 1♠
Pass 2q Pass 3♠
Pass 3NT Pass ?

Since South’s only side card is an ace, passing 3NT is a reasonable

gamble, in the hope that there won’t be any side tricks to be set up while
the opponents are attacking their best suit. On the actual hand there are
10 tricks available in either contract, and spades has little chance to
score an extra trick. However, change South’s hand to ♠ AKQJ842

66
 GENERAL CONCEPTS REVIEW PROBLEMS

r 97 q 7 ♣ KQ7, and now a 4 ♠ call is more prudent. The clubs may

have to be established, as in the actual hand, and there may not be time
to do this at notrump.

E-W vul., South holds ♠ 43 r QJ5 q A10653 ♣ Q105

West North East South

— 1♠ Pass 1NT
Pass 3♠ Pass ?

I would take a shot at 3NT, since my slow tricks are in my short suits, the
suits the opponents figure to lead. However, with ♠ 43 r A65 q QJ653
♣ Q105, 4 ♠ is preferable, since I may need time to establish diamonds.
If you look at both hands opposite a typical 3 ♠ bid such as ♠ KQJ10xx
r Kx q Kx ♣ AJx it should be quite clear why notrump is more likely
to be a winner on the first hand than on the second one.

♠ rq ♣ ♠ rq ♣

This is a good place to clear up a misconception that exists in the

minds of most bridge players about the difference between IMPs and
matchpoints. I have often heard a player who is faced with a close
choice of games say, “At matchpoints I’d try 3NT and hope to take
the same number of tricks, but at IMPs I’d play the safer major-suit
contract.” This is backwards from the way one should think. On most
hands, one should lean toward notrump more heavily at IMPs than at
matchpoints.
The reason one prefers a suit contract to notrump is because of the
possibility that the suit contract might produce more tricks. Suppose
both contracts score the same number of tricks. Then notrump is
superior at both forms of scoring. It is worth 10 more points for
matchpoint purposes and might earn a game swing at IMPs if that
number of tricks is 9. If the suit contract scores 2 more tricks than
notrump it is better at both forms of scoring, for it is then worth 50
more points at matchpoints and a possible game swing at IMPs when
the difference is 10 tricks verses 8. If the suit contract takes exactly
1 more trick it is again better at matchpoints, since it scores 20 more
points, but the difference is negligible at IMPs. Both contracts either

67
 MATCHPOINTS

make or go down, as the notrumpers are contracting for one fewer trick,
and the 1-IMP swing for the overtrick is inconsequential. Thus, the
major-suit game needs to score 2 extra tricks to be superior to notrump
at IMPs, while it needs to score only 1 extra trick to be superior at
matchpoints, and notrump is superior at both forms of scoring if the
contracts take the same number of tricks. Therefore, one should lean
toward notrump more at IMPs than at matchpoints.
The following hand is an illustration of this principle:

♠ 7
r A986
Both vul.
q K 10 6 5 3
♣ J97

♠ A K Q J 10 4
r 43
q A7
♣ Q 10 6

West North East South

— — — 1♠
Pass 1NT Pass 3♠
Pass 3NT Pass ?

One would certainly want to be in 4 ♠ at matchpoints, since there are

10 tricks available unless the opponents can negotiate a club ruff, while
notrump will probably only make 9 tricks if the defense leads or shifts
to hearts, which they probably will. However, at IMPs 3NT is better, for
4 ♠ can go down on a club ruff while 3NT is cold. South has a very close
decision on his final bid, and it might well be right to let the scoring be
the deciding factor—bidding 3NT at IMPs and 4 ♠ at matchpoints. If
you think that the decision is not so close and that either 4 ♠ or 3NT is
likely to be better at both forms of scoring I won’t argue, but it would
definitely be wrong to bid 4 ♠ at IMPs and pass 3NT as matchpoints, as
many players do.

♠ rq ♣ ♠ rq ♣

68
 GENERAL CONCEPTS REVIEW PROBLEMS

There is one type of hand that should play in notrump at matchpoints

and a suit contract at IMPs. This is the hand that will certainly make
game in the major suit simply because it is so strong, but where notrump
might get buried by a wide open suit. For example:

Both vul., South holds ♠ K32 r A10532 q KJ5 ♣ QJ

West North East South

— 1NT Pass ?

In a matchpoint game the player holding the South hand decided not
to try for slam (I agree), and to bypass the possible 5–3 fit by bidding
Stayman and then jumping to 3NT when North didn’t have four hearts.
This didn’t work too well, for North held ♠ AQxx r KQx q AQxx
♣ xx and the defense ran the first 5 club tricks. I must confess I wouldn’t
have done too well on the South hand either; in fact, I would have just
bid 3NT and not even looked for the 5–4 heart fit, since I think it far
more likely than not that notrump will score the same number of tricks.
At IMPs, however, I would transfer the hand and then bid 3NT, for if
North holds three hearts 4 r must be virtually cold, so there is no reason
to risk the opponents being able to run a suit.

69
 MATCHPOINTS

SEVEN-CARD TRUMP FITS

Many of the same principles apply to choosing between a 4–3 major-

suit fit and notrump, although the arguments in favor of the suit contract
must be stronger because of the shorter trump holding. Trump strength
is of prime importance, particularly in the 4-card holding. This is the
big distinction between 7- and 8-card trump fits. J9xx opposite Qxxx
can be a quite satisfactory trump fit, but J9xx opposite Qxx can be a
nightmare to play and should be avoided.
The main advantage of the 4–3 fit is the ability to score an extra trick
by ruffing a loser in the short hand. Consequently, side-suit texture is
very important in the hand with the 4-card trump holding. A suit such
as Axxx has several losers to be ruffed, while a suit such as QJxx is
not so good because it is likely that only winners will be ruffed. Also,
it is important that the 4-card trump holding be strong. AKJx opposite
Qxx is quite comfortable, since you can take a ruff in the short hand
and then draw trumps, but Axxx opposite KQJ is not as good, for
every ruff in the short hand may cost a trump trick.
A 4–3 fit is often chosen because one suit is inadequately stopped
for notrump. It is important that the hand with the short trumps be able
to take the ruffs in the weak suit. If the long hand is forced to take the
taps, you are reduced to playing a 3–3 trump fit, and if the opponents’
trumps don’t split 3–3 the roof will cave in. For example:

♠ KJ9
r 94
q KQ532
♣ KJ9

♠ AQ85
r 10 6
q AJ98
♣ Q52

On these combined hands 4 ♠ is the only game with any play, and the
chances are pretty good since the North hand with three trumps can
take the force in hearts if necessary, so the game will go down only
if the spades split 5–1 or the defenders can make declarer pull trumps

70
 GENERAL CONCEPTS REVIEW PROBLEMS

before establishing a club trick by threatening a diamond ruff. However,

change the hands to:

♠ KJ9
r 964
q KQ53
♣ KJ9

♠ AQ85
r 10
q AJ982
♣ Q52

Now 4 ♠, while not the worst contract ever reached, pretty much needs
a 3–3 spade split, since the South hand is forced immediately in hearts.
On this hand 5 q is virtually cold, so it is a superior contract.
It is usually the hand with the 4-card trump suit that is better placed
to make the decision about a possible 4–3 fit. Consequently, it is
generally correct to raise a major with 3-card support only if there is
any ruffing value at all, and let partner carry the ball. For example:

None vul., South holds ♠ A4 r J73 q K10875 ♣ KQ10

West North East South

— — — 1q
Pass 1r Pass ?

South should rebid 2 r, not 1NT. If North holds ♠ 97 r AQ104 q QJ6

♣ AJ64 he would simply raise 1NT to three, but after the 2 r raise a
likely bidding sequence might be:

West North East South

— — — 1q
Pass 1r Pass 2r
Pass 3♣ Pass 3q
Pass 4r Pass Pass
Pass

71
 MATCHPOINTS

When South bids 3 q rather than 3NT North can diagnose the weakness
in the spade suit, and he will bid 4 r even though he knows it might be
a 4–3 fit. On this pair of hands 3NT requires the hearts to behave to
take even 9 tricks, while 4 r has chances even with the r K offside or
the hearts 4–2 and will certainly make whenever 3NT does. However, if
North instead holds ♠ KQ97 r AQ104 q QJ6 ♣ 64 the bidding would
be:
West North East South
— — — 1q
Pass 1r Pass 2r
Pass 2♠ Pass 2NT
Pass 3NT Pass Pass
Pass

On this auction South will bid 2NT with only three hearts and the unbid
suit, clubs, well under control. Now North has no reason to play the
4–3 fit, so he bids 3NT. Notrump will make at least as many tricks as
hearts more often than not, since all suits are well stopped and there is
no useful ruffing value in the short hand.

None vul., South holds ♠ 7 r AK107 q 8752 ♣ K876

West North East South

— — — Pass
Pass 1♠ Pass ?

It seems at first glance that South’s only choice is 1NT; 2NT would be
far too aggressive opposite a third-seat opener (especially without a fit)
and 2 ♣ is reserved for Drury (and would not be particularly appealing
if it were natural). Careful reflection shows that 2 r is likely to be the
winning call, even though this promises five hearts. The hand is perfect
for a potential 4–3 fit—strong trump holding, two side suits that can use
some ruffs, and little danger of a force in the long hand. I held this hand
in the Life Master Pairs, tried a 2 r response, and was raised to 4 r. The
full hand is quite illuminating:

72
 GENERAL CONCEPTS REVIEW PROBLEMS

♠ K 10 6 4 3
r Q63
None vul.
q AK4
♣ A4
♠ 82 ♠ AQJ95
N
r J9854 r2
W E
q 10 6 q QJ93
S
♣ J 10 9 3 ♣ Q52

♠ 7
r A K 10 7
q 8752
♣ K876

West led the ♣ J. I took two rounds of clubs ending in my hand, and led
a spade to the ♠ K and East’s ♠ A. East returned a trump. I went up with
the r A, ruffed a club, cashed two top diamonds, ruffed a spade, ruffed
a club with the r Q, and exited with a diamond. West was forced to ruff
his partner’s trick and give me the tenth trick with my r 10, for 23 out
of 25 matchpoints. Admittedly this was all a bit lucky, but considering
that the 4–3 trump fit took 10 tricks despite a 5–1 split while at notrump
even 8 tricks are far from certain, the hand illustrates how powerful a
4–3 fit can be if the right conditions are met.

♠ rq ♣ ♠ rq ♣

A 5–2 trump fit usually will not produce a ruff in the short hand, the
extra trick that is so valuable in a 4–3 fit. The 5-card trump holding
provides protection against an enemy suit being run. Therefore, the
main reason for playing a 5–2 fit is to give yourself time to set up tricks
while the trump suit serves as a stopper. The conditions that make a 5–2
major-suit fit superior to notrump are:
1. A reasonably strong trump holding.

2. An unstopped or singly stopped side suit that the defense can run
at notrump.
A 5–2 fit is often backed into after all other possibilities have been
explored and rejected. For example:
73
 MATCHPOINTS

♠ K7
r A 10 8 7
N-S vul.
q Q 10 9 8 3
♣ A6

♠ AQ854
r K5
q KJ72
♣ 10 4

West North East South

— — — 1♠
Pass 2q Pass 3q
Pass 3r Pass 4q
Pass 4♠ Pass Pass
Pass

After South’s 3 q bid North is concerned about the club situation, so

he probes with a 3 r call. If South had ♣ Kx and two small hearts he
would bid 3NT, which would then be the right spot. However South
has nothing in clubs, so notrump is out of the question. South’s spades
are too weak in intermediates to rebid, so he retreats to 4 q . North isn’t
willing to settle for a minor-suit game yet, so he makes one more try
for a better score with 4 ♠. South’s spades are barely strong enough to
play the 5–2 fit, so he passes. It should be noted that 5 q is a much
safer contract than 4 ♠; in fact, 6 q will make if the spades split 3–3.
However, 4 ♠ will usually make, and will make 5 on a 3–3 spade fit, so it
is the winning contract at matchpoints. If South’s hand were ♠ A10xxx
r KQ q KJxx ♣ xx he should reject the 5–2 spade fit and move on to
the safety of 5 q . The trump suit would be too weak to handle a 5–2 fit.
It should be noted that North’s sequence on the previous hand
(bidding diamonds, then hearts, then raising spades) should not be
construed as showing a hand with a singleton club and slam interest.
It is far more important to leave open all possibilities to arrive at the
best game when strain is in doubt. However, had South rebid 3NT
instead of 4 q , then a 4 ♠ bid by North would show a stiff club and have
slam implications. The logic is that if North was sure he wanted to play
spades but had no slam interest he would have simply bid 3 ♠ over 3 q ,

74
 GENERAL CONCEPTS REVIEW PROBLEMS

and if the 3 r call was a probe for the best game then North would have
passed 3NT. The general principle is that any bid that can be interpreted
as looking for the best game is of overriding importance. It is only when
the trump suit is definitely established that slam tries can be made and
understood.

75
 MATCHPOINTS

MINOR SUIT GAMES

The lowly minor suit, at 20 points per trick, should be avoided if at all
possible when choosing the best game. The usual reason for playing a
minor-suit game is that there is a suit inadequately stopped for notrump
and no major-suit fit is strong enough to play, but the combined hands
have enough strength to contract for 11 tricks. For example:

♠ K964
r AQ52
Both vul.
q J52
♣ 42

♠ A75
r K4
q AKQ763
♣ 10 7

West North East South

— — — 1q
Pass 1r Pass 3q
Pass 3♠ Pass 4q
Pass 5q Pass Pass
Pass

After South’s 3 q rebid, North probes for the best game with a 3 ♠ call.
South doesn’t have a club stopper for notrump, doesn’t have a heart
preference, and his spades are too weak to suggest a 4–3 spade game,
so he must content himself with 4 q . North is far too strong to pass, and
the only game left open to him is 5 q , so he bids it. It should be noted
that 3NT is down off the top with a club lead, 4 ♠ depends on a 3–3
spade split, and the winning contract of 4 r(!), which depends only on a
4–3 heart split, is basically impossible to find, so 5 q is a good contract
and is likely to result in an above-average score.

♠ rq ♣ ♠ rq ♣

There is one type of hand that may belong in five of a minor even though
all suits are adequately stopped. If both partners have singletons, and
76
 GENERAL CONCEPTS REVIEW PROBLEMS

the suits opposite these singletons can’t be easily established by force,

then a 5–4 or even a 4–4 minor-suit game may be the winner, even
at matchpoints. The reason is that there just isn’t any source of tricks
at notrump, while in the minor-suit contract declarer may combine a
crossruff with setting up a long suit and take several extra tricks. For
example:

♠ KQ963
r7
E-W vul.
q A74
♣ Q 10 8 2

♠ 2
r AJ865
q K96
♣ AK43

West North East South

— — — 1r
Pass 1♠ Pass 2♣
Pass 3♣ Pass 3q
Pass 3♠ Pass 3NT
Pass 5♣ Pass Pass
Pass

After North raises clubs, South can realize that this might be the type
of hand that belongs in the minor, so instead of barging into 3NT he
temporizes with 3 q . North rebids his fair spade suit, and now South
suggests 3NT. North has help in diamonds, but he, too, can see that
a trump suit may be very important because of his singleton heart. If
South were sure of notrump he would have bid it immediately, so North
can infer that South must have a suit-oriented hand. Consequently,
North makes the good decision to bid 5 ♣. The minor-suit game will
probably come to 11 tricks by attacking spades, but where are the tricks
in notrump? If West doesn’t have ♠ Axx notrump will probably take
only 8 tricks, and if he does the club game could easily score 12 tricks,
to beat 3NT just making, so 5 ♣ is a much better contract.

77
 GAME OR PARTSCORE

Deciding whether or not to bid a close game can involve many factors.
Usually the simple approach of bidding game only if you think it
more likely to make than not will lead to the correct decision, since
at matchpoints the object is to beat the other players holding your cards
by any amount. At IMPs you like being in a 40% vulnerable game,
since the cost of being wrong by bidding the game when it goes down is
6 IMPs, while the cost of not bidding game when it makes is 10 IMPs.
Therefore, the total cost of being wrong bidding the game is 60% X
6 IMPs = 3.6 IMPs; for not bidding it the cost is 40% X 10 IMPs =
4.0 IMPs. At matchpoints, however, the cost is 1⁄2 matchpoint against
every other pair whether you bid game or not, assuming no other result
is possible, so you would prefer to stay out of a 40% game.
There are three different types of situations in which you may have
to decide whether or not to bid a game. First, there may be no other
possible strain—i.e., everybody in the room will be playing in spades,
the only question is whether or not to bid the game. Secondly, there
may be several different possible game or partscore contracts available
for your side. Thirdly, there may be some possible opposing contracts
to consider. There is some overlap among these three cases, but we will
examine each one separately.
 MATCHPOINTS

NO ALTERNATIVE CONTRACT

When there is no alternative strain possible, the decision of whether or

not to bid a game should, in theory, be based on your judgment as to
whether game is a favorite to make. However, it is not necessarily that
simple. If you find yourself in a contract that is different from the other
tables, you may not get a chance to make the most of your superior
declarer skills. We are all egotistical enough to believe that we play the
dummy as well or better than the next guy. If we stop short of game
and the rest of the field bids the game, all our skill in making overtricks
won’t help if the game makes. For example:

E-W vul., South holds ♠ AKJ r Q872 q KJ73 ♣ QJ

West North East South

— — — 1NT
Pass 2NT Pass ?

17 HCP, but what a terrible hand. Eleven of the points are stuffed into
the short suits, there are no intermediates, and the hand contains only
one ace. Despite this being a “maximum” 1NT opening in high cards,
I would bet that 3NT will fail more often than not. However, I would
still choose to bid the game. The reason is that the rest of the field will
also bid it, since most of the other players just count points in notrump
auctions.
Normally you should try to win a board in the bidding when the
opportunity arises rather than make the “field bid” if you think your
choice is better, since you must take advantage of superior bidding
judgment. Also, it is usually not possible to determine that accurately
what the field will do anyway. On this hand, however, it is clear that
the rest of the room will be in 3NT, and the contract will certainly
have some play even if it is not a favorite. If you pass 2NT you stake
everything on your bidding judgment and the defense at the other tables.
It won’t make any difference how well you declare your hand. The hand
is:

80
 GENERAL CONCEPTS REVIEW PROBLEMS

♠ 10 9 8
r J53
E-W vul.
q A 10 9
♣ K 10 8 3

♠ AKJ
r Q872
q KJ73
♣ QJ

West leads the ♠ 2 (fourth best), and you have to overtake dummy’s ♠ 8
with your ♠ J when East follows small. You lead the ♣ Q, West plays
the ♣ 2, and East wins the ♣ A and returns a spade. It appears as though
West has four spades and three clubs. If he had four hearts he might
have chosen a heart lead, while with four diamonds he would have
preferred the spade lead since North didn’t bid Stayman. Therefore
West is more likely to have long diamonds than East, so you back this
analysis by leading a diamond to the q 10, which holds. You cross back
with a club, lead a diamond to the q 9, cash dummy’s winners, and
come back with a high spade to collect your q K for the tenth trick. Well
played! West’s hand was ♠ Q942 r A4 q Q842 ♣ 752. With a normal
opening lead, some help from the defense (East would have done better
to duck the ♣ A to foul up your entries), and sound reasoning you made
one more trick than the rest of the field is likely to make. If you passed
2NT, however, your +180 will be a goose-egg, since everybody else will
be in game (North certainly had the right tens and nines), and anybody
can take 9 tricks, so your good play was wasted. In this case, failing to
bid 3NT costs a full matchpoint to those pairs who are +400, for you
get +180 instead of +430. Had partner had the wrong hand and 3NT
gone down you would only cost 1⁄2 matchpoint to the other tables by
bidding the game, or possibly nothing if your skillful play gets you out
for down 1 while the field is going down 2.
A similar problem may exist when you stretch for a game that the
field doesn’t bid. If you are good enough to squeeze that extra trick out
of the hand, it may not be necessary to bid the game to get a good score.
+170 is as good as +620 if the rest of the room is +140. The danger
in stretching too high is that the hand may break badly, and your extra
trick may only mean one less undertrick. For example:

81
 MATCHPOINTS

Both vul., South holds ♠ AKJ73 r 63 q K1052 ♣ K4

West North East South

— — — 1♠
Pass 2♠ Pass ?

Is the South hand worth a game try? Perhaps it is (although personally

I don’t think so), but you can be pretty sure the point-counting field
won’t move toward game with only 14 HCP and no great distribution.
The combined hands are:
♠ 10 8 6 4
r KQ7
Both vul.
q 643
♣ A93

♠ AKJ73
r 63
q K 10 5 2
♣ K4

North will certainly accept if you invite, and while game is not hopeless
you would rather not be there. The r J lead is covered by the r Q
and r A, and the heart return goes to dummy’s r K. You recognize the
possibility of an end-play later in the hand, so you carefully ruff a heart
before playing top spades. Unfortunately, East pitches a heart on the
second spade. You continue with three rounds of clubs, ruffing the
third, before exiting with a trump. East pitches his last heart on the
spade, and West plays a fourth club which you ruff in dummy as East
follows. Everybody is now down to nothing but diamonds, so you duck
a diamond to West. This is a winning line of play, for West started with
♠ Q92 r J102 q AQ9 ♣ Q1065. Most other declarers will probably
take only 8 tricks, but your −100 still loses to their +110 even though
you took an extra trick. Stretching to the game cost a full matchpoint
against those declarers who stopped in 2 ♠ and made 8 tricks, for you
went from a potential +140 to −100. The overbid completely nullified
your good play.
To summarize: If there is no other possible strain and the question
is simply whether or not to bid game, you should usually go with your

82
 GENERAL CONCEPTS REVIEW PROBLEMS

judgment. In theory, you want to be in the game if it is better than

even money, and stay out of the game if it is worse than even money.
However, there may be some close decisions where your judgment is
clearly contrary to what the field will be doing. In these cases it may be
better to make the “field bid” if you fancy yourself a superior declarer.

83
 MATCHPOINTS

OTHER CONTRACTS AVAILABLE

When there are several different contracts that might be reached at other
tables, deciding whether or not to bid a game can get very tricky. No
longer is it simply a question of deciding if game is a favorite to make.
If the game you must bid is not likely to be the best game or if your
partscore will probably outscore other partscores, it is better to settle
for the partscore. Conversely, if your partscore is probably inferior to
other partscores or if your game will definitely be the best result if it
makes, it is better to stretch to the game. We shall examine several
examples of these situations.

Both vul., South holds ♠ KJ76 r KQ975 q 6 ♣ K105

West North East South

3♣ 3q Pass 3r
Pass 4q Pass ?

South had a difficult choice at his first turn. Perhaps 3NT is the better
call, but 3 r (forcing) leaves open the possibility of game in either major
suit. Unfortunately North bid 4 q , the bid South least wanted to hear.
South would like to go back and bid 3NT but that would be insufficient,
so what should he do? 5 q figures to be a favorite, but slam looks pretty
unlikely. The problem is that if 5 q makes 3NT will also make, probably
with an overtrick or two. A large portion of the field figures to get to
3NT; either West won’t make a vulnerable preempt or South will try
3NT rather than 3 r at other tables. Consequently South will gain little
by bidding 5 q , so his best bet is to pass and hope the cards lie badly so
that no game makes. By passing, South gains 1⁄2 matchpoint from the
3NT players when he is right and nothing makes, while if South bids 5 q
he probably loses to the notrumpers even when he is right in bidding.
The full hand:

84
 GENERAL CONCEPTS REVIEW PROBLEMS

♠ A92
r 64
Both vul.
q A K Q 10 9 5 2
♣ 3
♠ Q 10 8
N
r A J 10 8
W E
♠ 543 q J874
S
♣ 72
r 32
q3
♣ AQJ9864
♠ KJ76
r KQ975
q6
♣ K 10 5

Against 3NT West leads the ♣ Q, and the contract is doomed by the
bad diamond split. The same bad split defeats 5 q , but if the diamonds
behave 3NT will take at least 10 tricks, so pass is definitely the winning
call.

E-W vul., South holds ♠ 62 r AQJ5 q AQ83 ♣ J32

West North East South

— Pass Pass 1r
Pass 2♣ 1 Pass ?
1. 3-card invitational raise

South has somewhat more than a minimum opening bid, so it seems as

though he might move on toward a heart game in the 4–3 fit. However,
it looks like he has already done well by opening the 4-card major in
third seat. Many pairs cannot find the 4–3 fit because system or personal
preference will prevent them from opening 1 r, and the 4–3 fit seems
right on this hand. Consequently, bidding 4 r would be overkill. South
should bid 2 r and take his good result for the best partscore, since
he will probably beat the notrump and club partscores. The combined
hands are:

85
 MATCHPOINTS

♠ J9
r K62
E-W vul.
q J42
♣ A Q 10 7 6

♠ 62
r AQJ5
q AQ83
♣ J32

4 r is a very reasonable contract, which will probably make if either

minor-suit finesse succeeds and suits split normally. However, 2 r
making whatever it makes will outscore all other partscores, and no
other game is likely to make. Bidding 4 r when it is wrong will cost
a full matchpoint to tables in notrump or club partscores, which are
likely contracts, and 4 r gains nothing against these results when it is
right since +170 would still be a winner.

N-S vul., South holds ♠ 7 r A104 q Q962 ♣ Q10642

West North East South

3r 3♠ Pass ?

North can have a wide variety of hands for his overcall. On balance,
I would guess that 3NT or 4 ♠ will fail more often than not. However,
South should still bid 3NT. The reason is that there is no assurance that
3 ♠ is a decent contract, so South might as well go for all the marbles,
even if it is slightly against the odds. Other tables may not face the
preempt, and they will be able to stop low if they don’t have a game, so
it is best to shoot the works rather than languish in 3 ♠. If North has a
distributional hand he may pull to 4 ♠ or best of all bid a minor.

♠ rq ♣ ♠ rq ♣

The previous hand is an illustration of a very important principle in

constructive bidding: When you have a close decision between an
overbid and an underbid, choose the bid that gets you to the proper
strain if the bid is successful. Usually this will mean probing on towards

86
 GENERAL CONCEPTS REVIEW PROBLEMS

game in close situations to make sure you play in the right suit. Here is
another example of this principle:

Both vul., South holds ♠ AK1062 r AQ85 q Q2 ♣ K5

West North East South

— — — 1♠
Pass 1NT Pass ?

On strength the hand is clearly worth a game invitation—a 2 r call

doesn’t come close to doing justice to the hand. The hand is not,
however, worth a game force. The value bid is, therefore, 2NT. The
problem with a 2NT rebid is that the heart suit will probably be lost
and the hand committed to notrump. It is better to make the slight
overbid of 3 r. This may get you overboard on occasion when partner
has a minimum, but at least you figure to land in the right strain. If you
overbid but play the right suit you may get lucky and make it, giving
you a second chance to win, but if you bid 2NT and find partner with
something like ♠ x r Kxxxx q xxx ♣ QJxx nothing can save you—4 r
is cold and 2NT is probably down 1.
Weaken the hand slightly to ♠ AK1062 r AQ85 q 72 ♣ K5 and
now the 3 r rebid is simply too optimistic. A 2NT rebid still has the
same problem, so it is best to make the mild underbid of 2 r. You may
occasionally miss a good game, but you will never land in a ridiculous
contract.

N-S vul., South holds ♠ Q10874 r Q985 q 64 ♣ 107

West North East South

— 2NT 1 Pass ?
1. 20–21 HCP

Bid Stayman, followed by 3 ♠ over 3 q (or 3 r if you use Smolen here).

The worst that can happen is that you get one trick too high, and maybe
you’ll get lucky and make it anyway. This is much less serious than
missing the right major-suit fit if you have one.

87
 MATCHPOINTS

Both vul., South holds ♠ Q7 r AKJ964 q A83 ♣ KQ

West North East South

— — — 1r
Pass 1♠ Pass ?

The hand is just about worth a game force, but which game? 4 r, 4 ♠,
and 3NT are all candidates, so a jump to 3NT or 4 r is unilateral and
may shut out the winning contract. A phony 3 q bid won’t really solve
the problem, and may lead to other complications. It is best to bid 3 r,
a slight underbid. This leaves partner room to rebid his spades, bid
3NT, or raise to 4 r, thus maximizing your chances of getting to the
best game. You may miss an odds-on game when partner passes on
a minimum, but you still have a second way to win—that game you
missed just might not make. Any time partner bids over 3 r you will be
pleased with your choice, for you are now more likely to get to the right
game than if you had guessed yourself.

N-S vul., South holds ♠ KQ953 r 93 q 96 ♣ K1032

West North East South

— 1NT Pass 2r ∗
Pass 2♠ Pass ?

Even though 3 ♣ is a game force, it is a much better bid than 2NT.

Now North can evaluate his hand intelligently, and get to 4 ♠ with ♠ AJ
r xxx q AKxx ♣ Axxx but play 3NT with ♠ AJx r KQ10x q AQ10
♣ xxx. Over a 2NT rebid North can do little more than count his points
and his spades to choose his bid. Getting to the right game on these
hands can swing a full matchpoint against those in a partscore and
1⁄ matchpoint against other pairs in game; stopping at 2NT or 3 ♠ can
2

only gain 1⁄2 matchpoint from other tables if you are exactly right and
neither game makes.

88
 GENERAL CONCEPTS REVIEW PROBLEMS

OPPONENTS IN THE AUCTION

When your opponents are in your constructive auction, there is another

possibility to consider when deciding whether or not to bid a game.
At other tables, players holding your cards may choose to defend,
collecting a plus score that way. You must determine how large their
score on defense is likely to be. If it figures to be only +50 or +100 you
will be glad to settle for a partscore, while if it is likely to be higher you
must bid the game to try and beat their score.
A common situation occurs when the opponents have a possible
sacrifice against your game contract. For example:

N-S vul., South holds ♠ AQ10742 r A74 q 9 ♣ KJ5

West North East South

— — — 1♠
2q 2♠ 3q ?

Without the enemy competition this hand would only be worth a game
try, since 4 ♠ figures to be an underdog if partner has a minimum.
However, on the actual auction you should just go ahead and bid 4 ♠.
This bid has two things going for it. First of all the opponents might
take the 5 q sacrifice, which you wouldn’t mind at all since you are not
that sure of making 4 ♠. Even if your opponents don’t save, the save
may be taken at other tables, and 5 q doubled will probably go down 2
or 3 for +300 or +500 your way. If this is happening at other tables, you
must get +620 to beat them. Failing to bid the game when it makes will
cost a full matchpoint against tables at which the save is taken, while
bidding the game and going down costs nothing against these tables.
Of course at the time you bid 4 ♠ you don’t know whether West has a
save-oriented hand or not, but simply this possibility justifies the slight
overbid.

♠ rq ♣ ♠ rq ♣

Another time to stretch to a marginal game is when you have missed

an opportunity to collect a number against an enemy overcall, and the
opportunity may be taken at other tables. In this case you must bid the

89
 MATCHPOINTS

game and hope to outscore the result that could have been achieved on
defense. For example:

Both vul., South holds ♠ 92 r J7 q Q1097 ♣ AQ1062

West North East South

— 1♠ 2q Pass
Pass 2r Pass ?

You made a Trap Pass of 2 q , hoping partner would reopen with a

double. Unfortunately, he chose 2 r. Perhaps his bid is reasonable,
but many pairs play a reopening double is almost automatic and will
get a chance at 2 q doubled. This will certainly go down at least 1 trick,
for a minimum of +200. On value your hand is only worth a 2NT bid,
for if partner has a distributional minimum game will not be a favorite.
However, I would shoot the works with 3NT, since a partscore won’t
beat the number that may be collected at other tables. If you are wrong
and game doesn’t make you won’t cost yourself anything against those
tables in 2 q doubled, but if you are wrong by failing to bid the game
you cost yourself a full matchpoint against every +200 or +500 at other
tables. It may seem as though the 3NT bid is in rage at partner’s failure
to reopen with a double (which it would be at IMPs or rubber bridge).
At matchpoints it is so important to try to beat results at other tables that
an “anti-percentage” call may be the winning bid if the “percentage”
call won’t get many matchpoints even if it is correct.

♠ rq ♣ ♠ rq ♣

The time you should settle for a partscore is when you feel that at
other tables players holding your cards may be selling out to enemy
competition and only collecting +50 or +100. If this is the case, the
partscore will be sufficient to garner the matchpoints, while stretching
to game could cost points that were already locked up. For example:

90
 GENERAL CONCEPTS REVIEW PROBLEMS

None vul., South holds ♠ 4 r K109832 q 843 ♣ KQ5

West North East South

— 1♣ 1♠ 2r
2♠ 3r Pass ?

The 2 r bid could have worked out very badly if partner didn’t have
heart support, but this time you are in luck. Partner clearly has a
minimum or he would have bid a game himself, but even opposite a
minimum opening bid with heart support game might be on—I would
estimate it at about 50–50 from South’s point of view. On this auction,
however, it is clearly correct to pass 3 r. Many players will not risk the
2 r call on your hand, and will sell to an undoubled spade partscore.
+140 or +170 will beat these results just the same as +420 will, but
−50 will lose to them if the spade partscore goes down. Consequently
you gain nothing against these tables by bidding game even if you are
right, but you lose a full matchpoint to those defending and beating a
spade partscore if you are wrong in bidding a game. This is not one
of those hands on which you have to worry about beating 4 ♠ saves at
other tables, for East’s pass of 3 r indicates that the opponents simply
do not have enough distribution to compete at that level, so 4 ♠ doesn’t
figure to be bid at other tables.
Let’s carry this example further. Suppose West now competes to
3 ♠, and this is passed back around to you. It is now quite correct to
bid 4 r, even after declining to do so earlier. You no longer have the
good +140 or +170 available; now all you have is a small set of 3 ♠ (if
it goes down at all!). Therefore it is now best to shoot out the 4 r bid.
This gains a full matchpoint from pairs who buy it for 3 r when game
makes, and loses nothing to these pairs if game doesn’t make, for you
could never have beaten them once West bid 3 ♠. The full hand:

91
 MATCHPOINTS

♠ 853
r AJ4
None vul.
q QJ7
♣ AJ32
♠ K 10 6 2 ♠ AQJ97
N
r Q76 r5
W E
q K62 q A 10 9 5
S
♣ 974 ♣ 10 8 6

♠ 4
r K 10 9 8 3 2
q 843
♣ KQ5

As you can see, 4 r depends on guessing the r Q, while 3 ♠ is down 1.

It is instructive to look at the results at all the tables on this hand. They
were:

N-S Score Matchpoints

+ 420 111⁄2
+ 420 111⁄2
+ 300 10
+ 170 9
+ 140 7
+ 140 7
+ 140 7
+ 100 5
+ 50 3
+ 50 3
+ 50 3
− 50 1⁄
2

− 50 1⁄
2

Looking at the matchpoints makes it clear that it is right to buy the hand
for 3 r if possible, but push on to 4 r if the opponents compete to 3 ♠.
If you pass 3 r and guess the r Q you cost yourself 21⁄2 matchpoints by
failing to bid game (from 111⁄2 to 9), but if you bid on and misguess the
r Q the 4 r call costs 61⁄2 matchpoints (from 7 to 1⁄2). If the opponents
bid to 3 ♠, however, it is another story. If you bid on and misguess the

92
 GENERAL CONCEPTS REVIEW PROBLEMS

r Q the 4 r call costs only 21⁄2 matchpoints (from 3 to 1⁄2), while if you
bid 4 r and guess the r Q you gain 81⁄2 matchpoints from the bid (from
3 to 111⁄2). The drastic change in the odds on the 4 r call when the
opponents bid 3 ♠ and when they don’t is an illustrative example of the
importance of matchpoint cost analysis in competitive auctions.

93
 THE BEST PARTSCORE

When the most your side can make is a partscore, the opponents hold
close to half the high-card strength. This means that at other tables
the players holding your opponents’ cards may be in the auction and
might buy the contract. If this happens, any plus you earn for your
partscore, particularly if it is greater than +100, will beat many pairs at
other tables who defend and collect only small penalties. Furthermore,
if the hand lies favorably for the opponents they may be able to make
their partscore, so if you avoid going down more than −100 you may
still get a reasonable board. It is important to understand that there may
be competition at other tables even if your present opponents are silent.
Others may take a more aggressive view with their cards.
If any plus score may be sufficient for a good board, it follows that
you should aim for the safest plus score rather than the highest one in
close situations. Minor suits, virtually ignored for game bidding, come
into their own in constructive partscore auctions. For example:

N-S vul., South holds ♠ 1094 r Q5 q AJ76 ♣ J654

West North East South

— 1r Pass 1NT
Pass 2q Pass ?

Many players would bid 2 r on the South hand, going after the higher-
scoring partial, but I think that they are wrong. Let’s look at the
combined hands and try to estimate some likely results:
 MATCHPOINTS

♠ Q8
r AK842
N-S vul.
q Q843
♣ K8

♠ 10 9 4
r Q5
q AJ76
♣ J654

If everything works you might make +140 in hearts, but if the hearts
aren’t 3–3 there is a reasonable chance that 2 r will go down (losing
2 spades, 2 clubs, 1 diamond, and 1 heart), while 2 q is safe against
most distributions. If hearts make 8 tricks and diamonds make 9 tricks,
not an unlikely result, it doesn’t matter which contract you choose. This
is a major difference between game and partscore bidding: in a game
contract if the minor makes one extra trick over the major, the major
still scores more; in low partscores this is not true. The big swings
on partscore hands occur when the opponents buy the contract at some
other tables, which is not at all unlikely. Let’s have a look at the full
hand:
♠ Q8
r AK842
N-S vul.
q Q843
♣ K8
♠ KJ76 ♠ A532
N
r J 10 7 6 r 93
W E
q 92 q K 10 5
S
♣ Q92 ♣ A 10 7 3

♠ 10 9 4
r Q5
q AJ76
♣ J654

This particular East chose not to bid, but it wouldn’t be hard to imagine
other Easts either making an initial takeout double or doubling 2 q . If
this happens, E-W will probably buy the hand for 2 ♠ undoubled. At

96
 GENERAL CONCEPTS REVIEW PROBLEMS

this contract they will probably lose 2 tricks in each red suit and 1 trick
in each black suit for down 1. 2 q should make comfortably for N-S,
but in 2 r declarer has 6 losers. Consequently, bidding 2 r rather than
passing 2 q costs a full matchpoint against tables in 2 ♠, while if 2 r had
made there would be no gain against these tables, since making 2 or 3
diamonds also beats those defending against a spade partial.
If the combined hands have near-game strength (23 or 24 HCP) then
it becomes more important to strive for the higher-scoring partscore.
The riskier contract is more likely to make on sheer power, and the
chance of enemy competition at other tables is decreased. In the
previous example, make South’s hand just a bit stronger, say ♠ 1094
r Q5 q AJ76 ♣ Q654. Now you would want to be in hearts, for 2 r
is pretty secure and there is a reasonably good chance for 9 tricks and
+140, which beats any possible result in diamonds. With this hand
South can recognize that this extra strength exists, so he should now
bid 2 r. If South’s hand is in the game zone, say ♠ 1094 r Q5 q AJ76
♣ A654, hearts is clearly superior to diamonds. Not only does 4 r have
a better chance to make than 5 q , but it scores more when it does make.
If it appears that your side holds only a bare majority of the points,
however, it is better to settle for the partscore that is most likely to
make.

N-S vul., South holds ♠ 53 r Q952 q KJ42 ♣ 876

West North East South

— 1q Pass ?

South should raise to 2 q (if not playing Inverted Minors) rather than
respond 1 r. In addition to the preemptive value of the raise, this hand
is so weak that South should be content to play in the safer partscore
if North is unable to take another bid. A 4–3 heart fit could be pretty
scary, and the risk of missing a 4–4 heart fit is not very serious. If
North has four hearts and a minimum opening bid there is a good chance
that the opponents can make some spades, so any plus score should be
an excellent result. If South’s hand were stronger, say ♠ K3 r Q952
q KJ42 ♣ 876, he should respond 1 r. The extra strength makes it

97
 MATCHPOINTS

worth going after a higher-scoring partscore in hearts or notrump. The

combined hands are:
♠ Q64
r KJ4
N-S vul.
q AQ653
♣ 93

♠ K3
r Q952
q KJ42
♣ 876

North is likely to raise the 1 r response to 2 r. This will probably make

in spite of the flimsy trump holding, and if the hearts split 3–3 South
may make +140 for a great score. However if South’s ♠ K were a small
spade, as in the first hand, 2 r would probably be a disaster, for the
opponents could tap South in spades. Also, other E-W pairs will be
getting into the auction and making some spades. So N-S should be
quite happy to buy the contract for 2 q rather than risk some bad splits
and −200 in 2 r.

Both vul., South holds ♠ J1062 r Q95 q J1087 ♣ 97

West North East South

— 1NT Pass ?

2 ♣, Stayman, is South’s percentage bid. If North has four spades it is

great, if North has four hearts the 4–3 heart fit may well play better
than 1NT, and if North responds 2 q he is likely to have at least four
diamonds, in which case 2 q figures to be better than 1NT since South’s
hand is so weak. But with ♠ J1062 r Q95 q J1087 ♣ K7 South should
pass 1NT. With the extra strength, it is better to go for the higher-scoring
notrump contract. Now, only spades is likely to be better than notrump.
The 4–3 heart fit will probably not produce an extra trick, and notrump
will almost certainly score better than a diamond partscore.

98
 GENERAL CONCEPTS REVIEW PROBLEMS

None vul., South holds ♠ J52 r A106 q 543 ♣ J1062

West North East South

— 1♠ Pass ?

South should raise to 2 ♠ rather than respond a non-forcing 1NT. In

addition to the preemptive value of the 2 ♠ bid, spades will probably
take one more trick than notrump if North has a minimum opening
bid, and if North has a strong balanced hand there is still time to get
to notrump if that is the right game. But if South had held ♠ J52
r A106 q QJ3 ♣ J1062 a 1NT response might be superior. The extra
strength and stoppers in every suit make it likely that notrump will take
as many tricks as spades if North has a balanced minimum, and if North
is unbalanced he will take another call and South can get back to spades.

♠ rq ♣ ♠ rq ♣

A 4–4 major-suit fit is virtually always superior to notrump at the

partscore level. The most important requirement for notrump to be
superior, excess strength, can’t be met when there is not enough strength
to make a game. For this reason, it is correct to probe for a 4–4 major-
suit fit even with a suitable notrump hand. For example:

Both vul., South holds ♠ Q104 r J973 q K6 ♣ Q654

West North East South

— 1q Pass ?

South should respond 1 r, not 1NT, despite the balanced nature of his
hand. If North has four hearts and a minimum opening bid the 4–4
fit will probably take at least one more trick than notrump, and South
can’t afford to risk missing this fit if it exists. This is one of the major
disadvantages of weak notrumps at matchpoints: a 4–4 major-suit fit is
often missed on a normal partscore hand.

♠ rq ♣ ♠ rq ♣

The requirements for a 4–3 major-suit fit to be superior to notrump at

the partscore level are not nearly as strict as at the game level. All that

99
 MATCHPOINTS

is usually needed is a reasonable trump suit and any kind of potential

ruffing value in the short hand. The reason is that it is not necessary
to keep control of the hand at the partscore level. Declarer can often
scramble home on a semi-crossruff for an extra trick. For example:

E-W vul., South holds ♠ A5 r K95 q AQ74 ♣ J632

West North East South

— — — 1q
Pass 1r Pass ?

South should raise to 2 r rather than rebid 1NT. If North has some
nondescript minimum such as ♠ Jxxx r Axxx q xx ♣ Kxx the 4–3 fit
figures to bring in an extra trick over notrump one way or another, even
though the trump suit is relatively weak, because declarer can score
tricks with small trumps that wouldn’t be available at notrump. The
hand may be difficult to play, but it will be just as difficult to defend. If
North happens to have five hearts the raise will almost certainly be right.
An added bonus of raising on hands such as these is that responder
won’t feel compelled to rebid a 5-card major after a 1NT rebid for fear
of missing a 5–3 fit, since he knows opener will raise on 3-card support
more often than not. If responder has a game-going hand with only four
hearts, there is plenty of time to get to notrump if that is where the hand
belongs.
There is one type of hand on which the 4–3 major-suit fit should
be avoided at the partscore level. When you have another suit that will
produce tricks, it is better to play notrump so that suit can comfortably
be run. For example:

Both vul., South holds ♠ K3 r Q98 q KQJ107 ♣ K63

West North East South

— — — 1q
Pass 1r Pass ?

South should rebid 1NT, not 2 r. The diamonds will score in notrump,
but they may not be so valuable in a 4–3 heart fit because North won’t
be able to draw trumps conveniently and run the diamonds. If South

100
 GENERAL CONCEPTS REVIEW PROBLEMS

had instead held ♠ K3 r Q98 q A10765 ♣ K63 a 2 r bid would be

preferable. The diamonds don’t figure to run at notrump, so the ruffing
value will probably produce the extra trick even if North has only four
hearts.

101
 PAVING THE WAY

This section will not be of much interest to some readers. Players

who use methods that restrict choice of early actions such as 5-card
majors, forced response in any 4-card major, required 4-card support
to raise a 1-level response, etc. do not have the opportunity to make
the types of decisions discussed in this chapter. These methods have
their advantages, but they eliminate the option of using good judgment
to make favorable delicate decisions early in the auction.
Players who play a free-wheeling style can base their initial actions
on more than the number of cards in a suit. They can look at the quality
of suits, the overall hand, and how the auction is likely to proceed to
help them make their choice. Intelligent early decisions can shift the
odds in favor of the player making the bid if he chooses wisely. We
shall examine a few of these decisions in detail. Similar reasoning can
be applied to all early decisions.
 MATCHPOINTS

THE FOUR-CARD MAJOR OPENING

Many methods do not permit discretion in deciding whether or not to

open a 4-card major. Eastern Scientific and Precision forbid opening
any 4-card major, while many Canape structures compel a major-suit
opening on a 5432 suit. Let us suppose you play a natural style that
allows leeway on this decision. What should the criteria be?
1. Trump strength. This is the most important consideration. Even if
you play “4-card majors,” partner will bid his hand as though you
have 5, since this is what you will have the great majority of the
time you open a major. Therefore, you must be prepared to play a
4–3 fit—possibly at the game level—if partner has a good hand,
3-card support, and some distribution, so your trump suit had
better be good enough to stand it. A hand such as ♠ 73 r AKQ4
q A972 ♣ 1092 is fine for a 1 r opening bid, but something like
♠ A3 r K864 q A972 ♣ Q92 is better opened 1 q . The latter
hand is not the type that you are anxious to play in a 4–3 heart
fit. There is little danger that a 4–4 heart fit will be lost. If the
opponents don’t bid partner can certainly bid a 4-card heart suit
now or on the next round, and if the opponents enter the auction
a Negative Double will help to locate the 4–4 heart fit.

2. Side-suit texture. One of the advantages of playing a 4–3 major-

suit fit is the ability to ruff losers in the short hand. If the hand
with the 4-card major has 4-4-3-2 distribution (the most common
distribution when considering whether or not to open a 4-card
major), the side 4-card suit is the suit most likely to be ruffed
in dummy. If this suit has late-round losers we will be happy to
ruff them, but if it is strong in intermediates the 4–3 fit is likely
not to be of value. So with ♠ J84 r AK106 q A763 ♣ 106 a
1 r opening bid is right because the third and fourth rounds of
diamonds may be ruffed in dummy, but ♠ Q84 r AK106 q QJ63
♣ 106 is better opened 1 q . The heart suit is strong, but the texture
of the diamond suit suggests that a 4–3 heart fit will not take an
extra trick over notrump.

3. Rebid considerations. In close decisions, the question of whether

or not to open a 4-card major is often decided by rebid consider-
104
 GENERAL CONCEPTS REVIEW PROBLEMS

ations. Assume that partner will respond in your shortest suit. If

you don’t have a convenient rebid, it is time to reconsider your
choice of opening bids.

None vul., South holds ♠ AQJ5 r 974 q 94 ♣ AJ54

West North East South

— — — ?

The black suits are quite satisfactory for opening the 4-card major, but
South should prefer a 1 ♣ opening bid. If South opens 1 ♠ and North
responds 2 q , South will have a serious rebid problem. He can’t very
well rebid his 4-card major, so he must rebid 2NT. Even if this shows a
minimum opening bid, as it does with most players today, it is not really
satisfactory due to the lack of a heart stopper. Also, there is virtually
no chance to get to a club partscore after opening 1 ♠. A 1 ♣ opening
bid does not suffer from any of these defects. South can easily rebid 1 ♠
over either red suit and get both his suits in while staying at the 1-level.
The only things wrong with the 1 ♣ opening are the loss of the 4–3
spade fit when it is right (and it doesn’t have to be right—another way
for the 1 ♠ opening to lose) and the loss of the preemptive value of the
1 ♠ opening if the opponents are considering entering the auction. In
my opinion, these defects don’t compensate for the potential cost of the
1 ♠ opening when an awkward rebid causes problems.
A small change in the suits makes the 4-card major opening bid
much more attractive. With ♠ AQJ5 r 974 q AJ54 ♣ 94 a 1 ♠ opening
bid is best. There are no rebid problems. 2 q can be bid over 2 ♣,
and 2 r can be raised to 3 (since 2 r shows a 5-card suit over a 1 ♠
opening bid). It is the 1 q opening that creates rebid problems after a
2 ♣ response. All the possible rebids (2 q , 2 ♠, or 2NT) have serious
drawbacks. Consequently, a 1 ♠ opening is clear on this hand.

4. Looking ahead. If the decision on whether or not to open the

4-card major is close, a look at how the auction might develop
often provides the answer. Each situation must be analyzed
individually. Here are a few examples.

105
 MATCHPOINTS

None vul., South holds ♠ 98 r AKJ6 q AKQ ♣ K832

West North East South

— — — ?

Some players would open 2NT despite the worthless doubleton and the
fact that the hand doesn’t really evaluate to 20 points, but let’s suppose
that you reject the 2NT opening. Which is better, 1 ♣ or 1 r? If partner is
so weak that he will pass either opening bid then you would rather play
1 r than 1 ♣, while if he responds the 1 r bid may work out well, since
a 4–3 heart game could well be right. It could be argued that partner
is more likely to keep the bidding open over 1 ♣ than over 1 r. This
is true, but it is actually an argument in favor of the 1 r opening. The
reason you didn’t open 2NT in the first place was so that you wouldn’t
get too high if partner had a bad hand. Suppose he holds something like
♠ KJxx r xxx q Jxx ♣ xxx. He will certainly pass 1 r, but he might
try 1 ♠ over a 1 ♣ opening bid, and now your 2NT rebid will put the
contract at a dangerous level.

None vul., South holds ♠ KQ107 r A983 q 42 ♣ KJ9

West North East South

— — — ?

A 1 ♣ opening bid is best. The quality of suit clearly argues against a

1 r opening bid. A 1 ♠ opening bid is OK as far as playing a 4–3 fit
goes, but there is too great a danger of missing a 4–4 heart fit if partner
responds 1NT. You don’t dare rebid 2 r, for this could easily land you
in a silly 4–2 spade fit. So the 4-card major opening should be rejected
on this hand. However, holding ♠ A983 r KQ107 q 42 ♣ KJ9 a 1 r
opening bid is quite satisfactory. Now a 4–4 spade fit won’t be missed
because partner can respond 1 ♠, so all the lights are green for the 4-card
major opening.

106
 GENERAL CONCEPTS REVIEW PROBLEMS

None vul., South holds ♠ 4 r AQ106 q A10654 ♣ KJ8

West North East South

— — — ?

Minimum opening bids with a singleton spade often present problems.

Many 4-card majorites would open 1 r, but I think they are wrong.
Their argument is that if you open 1 q and partner responds 1 ♠ you have
no convenient rebid, while if open 1 r you can rebid 2 q easily enough.
This sounds convincing, but you might have second thoughts when
partner takes a preference to 2 r with ♠ Kxxxx r xx q Qx ♣ Qxxx
or ♠ Kxxxx r Kx q xxx ♣ Qxx and the 4–2 fit is not the greatest
contract in the world. It is true that opening 1 q and rebidding 2 q (or
perhaps 2 ♣) over 1 ♠ may not work out—partner might have to pass
with ♠ Kxxxx r Kxxx q x ♣ Qxx for example—and rebidding 1NT
(my personal preference) has the disadvantage that partner will play
you for a doubleton spade, so there are problems whatever you open.
Suppose it is the opponents who bid spades. If you open 1 q the 4–4
heart fit can be discovered via the Negative Double, but if you open 1 r
the diamond suit may get lost forever, particularly if partner has three
hearts. For example, suppose North has ♠ Jxxx r Jxx q Kxxx ♣ Qx
and the bidding goes:

West North East South

— — — 1r
1♠ 2r 2♠ ?

You would like to be in 3 q , but you can’t get there from here. 3 q by
you would be forcing, so you are stuck and must either sell to 2 ♠ or
land in a very skinny 3 r contract when 3 q is clearly best. Had you
opened 1 q partner would have raised to 2 q on the actual hand, while
had his red suits been reversed he would have made a Negative Double
and you could then get to hearts. A 1 q opening bid leaves you better
placed in any competitive auction. The old philosophy of bidding your
longest suit first has much to recommend it.
Interestingly enough, change the hand to ♠ KJ8 r AQ106 q A10654
♣ 4 and the 1 r opening now has much more going for it. A 1 ♠ response
suits you fine as you can raise, and the enemy is less likely to have a
spade fit. If partner responds 1NT you can pass and have a reasonable

107
 MATCHPOINTS

chance of being right, while if partner responds 2 ♣ you are happier with
your 2 q rebid if you open 1 r than if you open 1 q .

108
 GENERAL CONCEPTS REVIEW PROBLEMS

1NT WITH A FIVE-CARD MAJOR

Whether or not to open a strong notrump when holding a 5-card major

is often quite a problem. The main considerations are suit quality, rebid
problems, and other flaws.

1. Suit quality. The main danger in opening 1NT with a 5-card

major is that you are very likely to miss a 5–3 major-suit fit if you
have one, for partner simply will not play you for a 5-card major
and most conventional methods over notrump openings do not
allow the 5-card major to be shown. Consequently, you should
choose hands on which notrump has a good chance of being
superior to the 5–3 fit. Such a hand is one that has a very weak
trump suit. With ♠ AQ10 r J6532 q KJ ♣ AJ2 a 1NT opening
bid is surely best. If partner has some scattered collection such
as ♠ Kx r Kxx q A10xxx ♣ 10xx, 3NT is certainly better than
4 r, and if you take away one of his high cards 1NT is likely to
take as many tricks as 2 r.
Another type of hand with which to consider opening 1NT
is a hand with a very strong major and intermediates on the side.
For example ♠ Q65 r AKQ74 q K93 ♣ K2. A 1NT opening bid
will probably work out well. Even if partner has three hearts there
may not be any ruffing value opposite this hand. Your hearts are
ready to run, and you can go right after your side tricks. Unless
the opponents can establish a suit on the opening lead, notrump
will probably take as many tricks as hearts.
The type of suit that calls for the suit opening bid rather
than 1NT is one of intermediate strength. For example ♠ AQ4
r KJ1053 q K2 ♣ QJ9. Despite the notrump orientation of the
hand, a 1 r opening bid is probably best, You can’t afford to miss
a 5–3 heart fit with this hand. The hearts are strong enough that
you will be using the suit for tricks at notrump if you have a 5–3
heart fit, but weak enough that they will take time to establish.
Consequently, the 5–3 fit is likely to produce an extra trick even
if there is no ruffing value in dummy, simply because time will
be required to establish tricks and the trump suit is needed as a
stopper.

109
 MATCHPOINTS

2. Rebid considerations. If you fail to open 1NT with a 5-card

major, you must consider your rebid problems after opening the
major. There are seldom any great problems with a spade suit, but
the heart suit can present difficulties if partner responds 1 ♠. For
example ♠ A9 r KJ1072 q KJ7 ♣ A72. The hand isn’t ideal for
a 1NT opening bid, due to the intermediate strength of the heart
suit. Nevertheless, 1NT is the best opening bid. The reason is
that if you open 1 r and partner responds 1 ♠ you are really stuck.
1NT is a clear underbid, 2NT is an equally large overbid, and
two of a minor is likely to lead to a disastrous result, with partner
possibly passing you out in a bad 4–3 or 3–3 fit. These problems
are solved with a 1NT opening bid.
It must be noted that this hand is the worst sort of hand for a
1 r bid. Any slight change makes the 1 r opening more attractive.
Weaken the hand a bit, and now the 1NT rebid is adequate,
strengthen the hand a bit and a 2NT rebid isn’t too much of a
distortion, give South three spades and now it may be feasible to
raise the 1 ♠ response to an appropriate level, particularly if you
play Flannery and the 1 ♠ response implies a 5-card.

3. Other flaws. When you open 1NT with a 5-card major you risk
a loss if partner has three of the major, so you would like to be
pretty sure of showing a profit if he doesn’t. Consequently, the
hand should be perfect for notrump outside of the 5-card major.
This is another illustration of loading the dice in your favor:
you don’t want to give yourself two ways to lose. For example
♠ AKQ42 r 87 q KJ7 ♣ QJ2. South should open 1 ♠ , not 1NT.
The worthless doubleton in hearts is the second flaw. It would be
sad if North had something like ♠ Jx r xx q Axxx ♣ AKxxx.
After a 1NT opening bid he would bid 3NT and the heart suit
would be wide open. After a 1 ♠ opening, an intelligent auction
might be:

110
 GENERAL CONCEPTS REVIEW PROBLEMS

West North East South

— — — 1♠
Pass 2♣ Pass 2♠
Pass 3q Pass 4♣
Pass 4♠ Pass Pass
Pass

If you had opened 1NT you would have been right in that you didn’t
miss a 5–3 spade fit, but you would still come out a loser because of the
second flaw, the worthless doubleton.

None vul., South holds ♠ K8 r J9642 q AQ108 ♣ AQ

West North East South

— — — ?

South should open 1 r. The 5-4-2-2 distribution is too much of a

flaw. Can’t you just picture North holding something like ♠ Q10xx r x
q Kxxx ♣ Jxxx. You would certainly like to be in a diamond partscore,
but it’s tough to get there after a 1NT opening bid. Once again you got
away with not opening the major (no 5–3 fit), but the second flaw (the
4-card diamond suit) came back to haunt you. Avoid opening 1NT with
a 5-card major unless the hand is otherwise perfect for the action.

111
 MATCHPOINTS

THE INITIAL RESPONSE

The proper choice of initial response is often the key to get to the
right contract. The main consideration is the rebid problem. You must
think of partner’s two or three most likely rebids and anticipate how
comfortable the auction will be after he makes them.
There are basically three types of hands to consider: minimum
strength responses, game-invitational hands, and forcing hands. We
shall examine problems with all three types.

♠ rq ♣ ♠ rq ♣

With a minimum strength response, directness is the key. Keep in mind

that you will not bid again if partner makes a limited rebid, so you
should hope that these bids will be the best contract. Also, you want
to avoid complex auctions in which you have difficulty limiting your
hand.

None vul., South holds ♠ KQ107 r 73 q J10872 ♣ 32

West North East South

— 1♣ Pass ?

A 1 ♠ response is better than 1 q . If partner raises to 2 ♠ you will be

quite happy to play a possible 4–3 spade partial, and any non-forcing
bid by partner can be easily passed. A 1 q response loses the possibility
of finding a 4–3 spade fit, and can lead to other complications. For
example:

West North East South

— 1♣ Pass 1q
Pass 1♠ Pass ?

Should you raise to 2 ♠? If you do, partner will play you for a better
hand, and bid a bad game with something like ♠ Jxxx r Ax q Kx
♣ AKJxx. If you pass, however, you may find that partner has ♠ AJxx
r xx q x ♣ AKQxxx. He would have jumped to game over your 1 ♠
response, but was not strong enough to do more than bid 1 ♠ over a 1 q
response since a fit had not yet been found. Another problem with the

112
 GENERAL CONCEPTS REVIEW PROBLEMS

1 q bid is that partner might hold ♠ xxxx r AQx q Kxx ♣ KQx. A 1 ♠

rebid on this balanced hand is just too much of a distortion, so he will
bid 1NT and the superior 4–4 spade fit will be lost. Also, if North rebids
1 r you must now rebid 1 ♠. This might work out OK, but it is better
not to have to make two forcing bids on a 6-count if it can easily be
avoided.
Contrast with the following hand, same auction:

None vul., South holds ♠ 10874 r KJ9 q QJ62 ♣ 53

West North East South

— 1♣ Pass ?

Over the 1 ♣ opening bid, South should respond 1 q . All the arguments
are in the other direction. South does not want to play a 4–3 spade fit, he
can pass a 1 ♠ rebid secure that there is no game opposite a distributional
black two-suiter, a 1 r rebid can happily be passed, and missing a 4–4
spade fit if North has a balanced hand with bad spades might not be so
bad. Players who use good judgment on hands like this will get better
results than those who always bid up the line or always respond with
any 4-card major.

♠ rq ♣ ♠ rq ♣

If you have a hand of intermediate strength, you are probably going to

take another bid after partner rebids. The likely course of the auction
must be examined carefully. For example:

N-S vul., South holds ♠ AQ8 r 9752 q KQ85 ♣ 63

West North East South

— 1♣ Pass ?

There is nothing particularly wrong with a 1 q response, but 1 r figures

to make life easier for you and more difficult for the opponents. The
point is that you plan to rebid 2NT over 2 r or any other minimum rebid.
This describes your hand accurately—game invitational, balanced, only
four hearts—so there is little danger of getting stuck in a bad heart
contract if partner has only 3-card support. In addition, if partner
113
 MATCHPOINTS

doesn’t have heart support your response may stop a heart lead against
notrump, and this is likely to be the best lead for the defense. Another
subtle problem with the 1 q response is that if partner bids 1 r it is not
clear whether the hand is worth driving to game, but if the bidding goes:

West North East South

— 1♣ Pass 1r
Pass 2r Pass 2NT

You have described your hand accurately, and partner has a good chance
to make the right decision.

Both vul., South holds ♠ KJ4 r J74 q A532 ♣ J53

West North East South

— 1r Pass ?

If partner has a 5-3-3-2 minimum either 1NT or 2 r could be right—

there are arguments both ways. The problems occur when partner
doesn’t have the balanced hand. Suppose you respond 1NT and he
rebids 2 ♣. What now? If you emerge from your shell with a 3 r bid,
it will be your luck to find partner with ♠ xx r Axxxx q Qx ♣ AKxx
and 3 r will be in quite a bit of jeopardy. On the other hand, if you
merely bid 2 r partner will undoubtedly hold ♠ Ax r KQxxx q xx
♣ AKxx and a good 4 r game will be missed, for partner can’t afford
to bid again since your sequence doesn’t promise 3-card support. The
solution to this dilemma is to bid 2 r immediately. This doesn’t show
more strength than the 1NT response, but will be more encouraging to
partner if he holds an intermediate hand with some distribution. He will
pass with the first hand but make a game try with the second, and you
will get to the right contract in both cases.
Conversely, suppose you had the same scattered strength but
weaker, say ♠ KJ4 r J74 q J532 ♣ J53. Now a 1NT response may
turn out OK if partner passes, and will certainly be better if he has
an unbalanced hand. You don’t want him making a game try with a
moderate hand, and the sequence:

114
 GENERAL CONCEPTS REVIEW PROBLEMS

West North East South

— 1r Pass 1NT
Pass 2♣ Pass 2r

will slow him up, which is what you want.

♠ rq ♣ ♠ rq ♣

With a game-going hand, you can assume that the only objective is to
find the best game. Slam may be possible if partner has a strong opening
bid, but that can be worried about later. The overwhelming majority of
these hands will be game contracts. Since matchpoints is a game of
frequency, finding the best game is more important. Consequently, you
should look ahead so that the auction is likely to develop in such a way
that opener will be able to make the right decision. For example:

E-W vul., South holds ♠ AQ85 r A4 q 83 ♣ KJ962

West North East South

— 1r Pass ?

South should respond 2 ♣ rather than 1 ♠. There are many possible game
contracts, and you want to keep all avenues open by creating an early
game-force. Over partner’s likely 2 q or 2 r rebid you can now bid 2 ♠
and plan to follow this up with a 3 r bid, painting an accurate picture
of your hand and pinpointing the diamond weakness; whether the right
game is 3NT, 4 r, or 4 ♠ (or possibly 5 ♣), partner will be well placed
to judge correctly. The main difficulty with a 1 ♠ response is that the
auction might go:

West North East South

— 1r Pass 1♠
Pass 2♠ Pass 3♣
Pass 3♠ Pass ?

You knew you were probing for the best game when you rebid 3 ♣.
Unfortunately, from partner’s point of view 3 ♣ was a game try, so he
signs off with a minimum (and possibly 3-card spade support), and you
have to guess what to do. Some hands he might hold for this auction

115
 MATCHPOINTS

are ♠ KJx r KJxxx q Jx ♣ Axx (4 ♠ is best), ♠ Jxx r KQ10xx q xx

♣ AKx (4 r is best), or ♠ Kxx r Qxxxx q KQ10 ♣ Kx (now 3NT is
the winner). After a 2 ♣ response you would have a good chance to get
to the right contract opposite all of these hands.

116
 SLAM BIDDING

Slams play a very important role in IMP matches. Any time there is a
slam in the air, there are always at least 10 IMPs up for grabs if the slam
is bid at one table and not bid at the other table. Good slam bidding
mechanisms are very important to any partnership. That gadget that
gets the partnership to a slightly safer six of a minor contract or keeps
them out of an inferior slam because a critical weakness is diagnosed
can swing many IMPs, and this large IMP swing justifies the caution
and accurate methods even though opportunities to employ them may
be infrequent.
At matchpoints, however, slam bidding takes a definite back seat
to all other constructive or competitive bidding. Getting to the best
game is at least as important as bidding a brilliant 60% slam in a 4–4
minor fit, since beating the field by 10 or 20 points is as big a win
as a 750-point victory. Since slam decisions occur very infrequently
relative to game and partscore decisions (perhaps 10% of all hands
are in the slam zone), accurate slam bidding just isn’t very important.
Consequently, a player’s first thoughts about a hand should be about
how he will do in a constructive game auction or a competitive partscore
battle, not how a slam sequence might work out. For example:

None vul., South holds ♠ AKQJ86 r 74 q AQJ9 ♣ 4

West North East South

— Pass Pass ?

South should simply open 4 ♠ at matchpoints. This is very likely to

be the right contract opposite almost any North hand, and the bid puts
tremendous pressure on the opponents. It is true that there may be a
slam if North has the right hand. For example, as little as ♠ xx r Axxx
q K10xx ♣ xxx gives an excellent play for 6 q . For this reason, it is
probably correct to open 1 ♠ or 2 ♣ at IMPs. The swing of a missed
slam is just too great to risk. At matchpoints, frequency is the name of
 MATCHPOINTS

the game. If you can shut the opponents out of a good save, or goad
them into a bad one, or cause them to double you when you are cold, or
escape a double when the hand belongs to the opponents, you will gain
just as much as if you had reached a hard-to-bid slam by starting slowly,
and one of these small gains is much more likely than getting to a good
slam. At IMPs the gains from opening 4 ♠ when the opponents take the
wrong action are usually only 2–5 IMPs as opposed to a double-figure
swing for bidding a slam, but at matchpoints each swing is the same
size.

118
 GENERAL CONCEPTS REVIEW PROBLEMS

WHICH SLAM TO BID

Slam bidding is one area where it is very important to try to determine

what the field will be doing. If it is clear that the rest of the field will also
get to slam, then it is very important to get to the highest-scoring slam.
If a substantial part of the field is likely to miss the slam, it is crucial to
get to the safest slam. If you gamble 6NT and go down 1 when you are
cold for six of a minor, you cost yourself a full matchpoint to any table
that stops in game, while if 6NT had made you gain nothing against
these tables for playing notrump instead of the minor.
When the room is in slam, notrump is king. Six of a minor can
never score more than 6NT, and six of a major only scores more if it
makes 7. For example:

None vul., South holds ♠ 76 r J95 q AQJ5 ♣ KQ83

West North East South

— 2NT Pass ?

At IMPs you would haul out all your fancy slam-bidding machinery
over 2NT opening bids and attempt to find a 4–4 or 5–4 minor-suit fit,
for if such a fit exists it is probably safer than the notrump slam. At
matchpoints you can throw your machinery out the window and bid a
direct 6NT. Your hand is strong enough that the field will bid slam, and
6NT certainly figures to be a favorite even if it is not cold, so you must
bid 6NT even if it is not aesthetically pleasing. The combined hands:

♠ A Q 10
r AK2
None vul.
q K 10 9 7
♣ AJ4

♠ 76
r J95
q AQJ5
♣ KQ83

119
 MATCHPOINTS

6 q is virtually cold, while 6NT basically depends on one of two spade

finesses or the r Q dropping doubleton. At IMPs you would be pleased
to arrive at 6 q , but at matchpoints is essential to get to 6NT. It will
make more often than not, and the occasional top you get for bidding
6 q when 6NT goes down won’t compensate for the more likely bottom
you get when 6NT makes.
Weaken the South hand somewhat to make it ♠ 76 r 985 q AQJ5
♣ K983. Now it is far from clear that the field will bid a slam, so
you should be happy to play a minor-suit slam if you can locate a
fit. Opposite the actual North hand 6 q a pretty good contract. If the
diamonds split 3–2 it will make if clubs are 3–3 or if either spade honor
is onside. 6NT is just fair—it needs both spade finesses or one spade
finesse and a club split, with some outside squeeze possibilities. If the
whole room bid a slam 6NT would not be bad, but since many pairs will
stay out of slam 6 q should be a reasonable result even if 6NT makes,
and a top if 6NT goes down when 6 q makes. Bidding 6NT and being
wrong on this hand can be very costly, turning a near top into a near
bottom, while getting to 6 q and being wrong (i.e., 6NT also makes)
will not cost nearly as much. The reason is that bidding 6NT as opposed
to 6 q risks a full matchpoint but gains nothing against those pairs who
stop short of slam.

♠ rq ♣ ♠ rq ♣

When it is clear that the field will bid a slam and you have a choice
between a safe major-suit slam and a good but not secure notrump
slam, it is sometimes correct to choose the major slam even if 6NT is a
favorite to make. The reason is that the major suit has an extra way to
win—even if 6NT makes, the major may make 7 and outscore notrump.
This should only be done when you have all the aces, of course. If an
ace is missing the major can’t make 7, so you should bid 6NT if you
think it more likely to make than not. For example:

120
 GENERAL CONCEPTS REVIEW PROBLEMS

Both vul., South holds ♠ A86 r AKQ3 q J73 ♣ K62

West North East South

— 1NT Pass 2♣
Pass 2r Pass ?

It would not have been unreasonable for South to have bid 6NT directly,
as he certainly would have done had his 4-card suit been a minor. South
reasoned that if North had four hearts, a doubleton in the North hand
might be valuable since South has no queens in his 3-card suits, so a
third-round ruff could be worth an extra trick. Well, North came up
with a 4-card heart suit. What should South do? His percentage action
is to ask for aces (via whatever ace-asking mechanism his partnership
uses in this situation). South isn’t worried about being off two aces,
which is virtually impossible opposite a strong notrump, and he isn’t
trying for a grand. There are North hands that will produce a good
grand, but it would take better bidding methods than most of us have to
determine if North has such a perfecto. What South is planning to do is
to bid 6NT if he is off an ace, but 6 r if he has all the aces. 6NT figures
to be a favorite in any event, but 6 r might make 7 if North has the right
hand and no aces are missing. The combined hands:

♠ K4
r J985
Both vul.
q AK8
♣ A J 10 5

♠ A86
r AKQ3
q J73
♣ K62

6NT depends on running four club tricks unless the q Q is nice enough
to drop doubleton, but if four club tricks are available, 7 r will make.
However if North is missing an ace, he might hold ♠ Kx r Jxxx
q AKQ ♣ QJ10x. Now both 6 r and 6NT are cold, so you would want
to be in notrump.

121
 MATCHPOINTS

TO BID A SLAM OR NOT

The problem of whether or not to bid a close slam is quite similar to that
of bidding a close game. When there is only one possible strain, you
should usually bid it if you think it will make and not otherwise. Once
again, really close decisions should be resolved in favor of making the
“field bid” if possible. You can assume that most of the field will bid a
“point-count” slam, but not one depending on distribution, so use this
information to judge accordingly.
When alternative game contracts are possible, you must estimate
how well your game contract will fare relative to other game contracts
if you don’t bid the slam. When the auction has forced you into a
minor suit and you think that notrump will do well, it is better to shoot
out a thin slam than to languish in five of a minor only to lose to the
notrumpers. For example:

E-W vul., South holds ♠ J7 r 532 q AQ962 ♣ AQ10

West North East South

— 1♠ Pass 2q
Pass 2r Pass 2NT
Pass 3q Pass 3NT
Pass 4q Pass 4♠
Pass 5♣ Pass ?

Just what you didn’t want to hear! You bid 4 ♠ in the hope that it might
be the best game. Perhaps this wasn’t too good a decision; signing
off in 4NT might have been better. You were never really trying for
slam, but things seem to have gotten out of hand. Now you know that
partner is probably 5=4=4=0, but you seem to have the wrong cards for
slam purposes. At IMPs you would just bid 5 q and take your likely
plus score, but you can’t afford to do that at matchpoints. One way
or another, most of the field will get to 3NT. Unfortunately for you, it
appears as though the field will be right. With the fit in diamonds and
powerful club holding, 3NT is likely to do very well. Your only hope
for a decent board is to bid 6 q and hope to get lucky, even though the
slam figures to be an underdog. The combined hands:

122
 GENERAL CONCEPTS REVIEW PROBLEMS

♠ A9542
r A Q 10 7
E-W vul.
q K873
♣ —

♠ J7
r 532
q AQ962
♣ A Q 10

6 q is not a favorite to make, but it has reasonable play. Stopping in

5 q when 3NT makes 10 tricks, as it will most of the time, will cost a
full matchpoint against the notrumpers when slam makes (from +920
to +420 against their +430), while it costs nothing to bid 6 q when the
slam goes down, as you were dead anyway.

♠ rq ♣ ♠ rq ♣

The converse applies when you believe you have reached a high-scoring
game contract that the field is not likely to find. A good 4–3 major suit
fit is a common example. When this happens, stopping in game will net
you many matchpoints even if the slam makes. For example:

Both vul., South holds ♠ AJ10 r AQJ6 q AJ43 ♣ 62

West North East South

— — — 1r
Pass 2q Pass 3q 1
Pass 3r Pass 3♠
Pass 4♣ Pass 4q
Pass 4r Pass ?
1. Extra strength

So far, the auction has worked out very well. Partner probably has the
♣ A but not great club strength (since he didn’t bid 3NT), only three
hearts, and a fair but not particularly strong hand since he couldn’t
do more than bid 4 r at his last turn. 6 q certainly figures to have a
reasonable play whatever he holds, and it might even be cold. However,

123
 MATCHPOINTS

you have only about 29–30 HCP between the combined hands and no
great distribution, so most of the field will not get to slam. Most players
cannot or will not open 1 r on your hand, so the 4–3 heart fit will be lost
at other tables. The most popular contract will be 3NT, with 5 q being
tried at a few tables. The 4–3 heart fit will probably outscore these other
game contracts, so it is best to pass 4 r. Even if 6 q makes, 4 r should
prove to be a winning matchpoint contract. The combined hands:

♠ Q76
r K32
Both vul.
q K 10 8 7 5
♣ A8

♠ A J 10
r AQJ6
q AJ43
♣ 62

6 q is a pretty good contract. If the diamonds come in it is cold,

and if the diamonds don’t behave it can still make if three rounds of
hearts survive and the spade finesse works, so at IMPs you would like
to be there. At matchpoints, however, 4 r figures to be the winner.
Notrump only does well when both spades and diamonds behave, while
hearts can survive an unfavorable lie in either or both of these suits,
particularly if the hearts are 3–3. Consequently, 4 r will score well on
most lies of the cards, and it may be the best contract of all.

♠ rq ♣ ♠ rq ♣

124
 GENERAL CONCEPTS REVIEW PROBLEMS

If your opponents might take a save against your slam, it is best to bid
it even if you are not all that sure of making it.

N-S vul., South holds ♠ AQ9764 r A6 q 6 ♣ AJ62

West North East South

— — — 1♠
2q 3♠ 1 4q ?
1. Limit raise

Who knows if slam will make. It probably depends a lot on North’s club
holding. I would guess that slam is a slight underdog and that careful
control-bidding might give us a better idea, but I would just go ahead
and bid 6 ♠. Now the opponents have to decide what to do. If they
save I’m happy since I wasn’t all that sure of making and the slam may
not be bid at other tables anyway, while if they defend I may still make
it. If I had held ♠ AQ9754 r A6 q 6 ♣ AKJ2, I would expect 6 ♠ to
make, but 7 to be off a trick somewhere. I would also expect most of
the field to bid the slam. Once again, I would leap to 6 ♠. This time the
opponents should take the save, but how can they tell when I bid both
hands the same way?
When you are bidding a slam, how you get there is often more
important than whether you get there. The opening lead, vital at any
contract, can be the whole difference in a slam because the defense
may not get a second chance. Overtricks are also important, so a cashed
or uncashed ace can mean a lot if the rest of the field is also in slam.
Consequently, it is often better to just blast away rather than conduct a
careful scientific auction. If you blast to a slam off two cashing tricks
you may still win if the defense doesn’t cash them, but if your accurate
auction tells the opponents what to lead you not only lose to the slam
bidders who get away with it, but also to the pairs who stop in game
and don’t get the best opening lead. Also, when the slam is a good
contract, a blasting auction may stop the lead to hold in the overtrick.
For example:

125
 MATCHPOINTS

None vul., South holds ♠ KQ75 r AKQ108 q K9 ♣ J5

West North East South

— 1♠ Pass ?

No doubt you and your partner have careful methods to stay out of
slam if partner holds ♠ AJxxx r xx q AQx ♣ Qxx, and at IMPs these
methods should definitely be employed. However, this is matchpoints!
It is a fact of life that the field will be in slam on this hand unless they
are off two aces. The practical player will simply bid Blackwood. If
partner produces three aces, South should probably gamble out 7NT.
This will make whenever the hearts produce 5 tricks, which is basically
on a finesse at worst, and there may be other chances if North has
extra strength. If partner has only two aces, South has several possible
approaches. One is to take his chances with a 6NT bid, which is likely to
work out but will look silly if North has a singleton small club. Another
approach would be to try 5NT. I know this “promises” all the aces,
but so what? There is no chance that North will leap to a grand on
his solid suit when we hold the king and queen of that suit. If North
shows the ♣ K we can now bid 6NT expecting to be a favorite, while
if North has no king South can try 6 ♠ and hope for the best. Another
approach is to bid 5NT, planning on bidding 6NT whether or not North
has a king! This bluff may stop an opponent from making an aggressive
lead from the ♣ K, since South “promised” all the aces, particularly
after South bids 6NT when his partner doesn’t have any kings. Any of
these approaches is reasonable, and they are all better than bidding a
mundane 6 ♠ after the two-ace response and better than trying to bid the
hand scientifically and winding up telling the opponents what to lead
whether or not slam is reached. The point is that at matchpoints the
extra 10 points for notrump and the potential overtrick from a friendly
opening lead are so important that accuracy and scientific bidding in
slam auctions is much less likely to be rewarded than at IMPs, where
getting to the game or slam most likely to make has number-one
priority.

126
 LOW-LEVEL INTERFERENCE

FREE BIDS

When the opponents enter your auction at a low level, it is extremely

important to describe immediately any distributional features or trump
support that you have, even if this involves overbidding your values.
The partner of an overcaller or takeout doubler may be waiting ready to
throw a preemptive bid or raise at you, and if you don’t get your suits
bid now you will be forced to guess later at an uncomfortably high level.
The concept of the “free bid” or “free raise” showing extra strength has
been discarded by most good players. Today, a free bid, and especially
a free raise, is often made on less strength than normal. For example:

Both vul., South holds ♠ 862 r K654 q 83 ♣ J1064

West North East South

— 1r 2q ?

If East had passed, South should pass. If he raises, North may bid a
game or make a game try that will get N-S overboard. If West reopens
South can compete to 2 r, and North will carry the ball from there.
However, East’s overcall changes the picture. Passing is too dangerous,
for West may raise to 3 q . Now what should South do if it comes
around to him? If he tries 3 r he may catch his partner with a balanced
minimum, get doubled, and go for a number. On the other hand, if
he sells to 3 q it will be his luck to find North with ♠ Ax r AQxxxx
q xx ♣ K9x. If North has this hand E-W can almost certainly make
3 q , and 3 r will go down at most 1 and might make, so N-S should
compete to 3 r. Unfortunately, it is difficult for North to bid on his own
over 3 q , for without heart support his hand could be a disaster at the
3-level. The solution is for South to raise to 2 r immediately. Now
North can compete to 3 r with the above hand but sell to 3 q with a
5-3-3-2 minimum, and he will be right in both cases.
 MATCHPOINTS

This illustrates an important principle of competitive bidding that

will be discussed more in the next section. In a competitive auction,
accuracy in game bidding is not nearly as important as in a totally
constructive auction. A stretch to a marginal game has more than one
way to win—it just might be the opponents’ hand if the cards lie well
for them, and you may steal the contract, often undoubled. In addition,
you may induce a phantom sacrifice. The important point is to get your
suits in fast when the competition starts.
Let’s switch to opener’s side and see how he should continue after
an overcall and a raise. If the partner of the overcaller passes, his
bidding is fairly normal. So if with none vul. the auction goes:

West North East South

— — — 1r
2q 2r Pass ?

a. ♠ 432 r AQ865 q 96 ♣ AK8. South should pass and subse-

quently sell out if the enemy competes, since he has no extra
strength or distribution.

b. ♠ 742 r AKJ76 q 5 ♣ AQJ8. South should bid 3 ♣, his normal

game try.

c. ♠ K9 r AKQ963 q 9 ♣ AQ108. South should just bid 4 r. This

has the effect of putting pressure on West if he has a distributional
hand—he is forced to guess at a high level whether or not to save.

d. ♠ 104 r AKQ92 q KQ2 ♣ A75. South should bid 2NT, an

accurate value bid.

e. ♠ A5 r AQ9854 q 94 ♣ K63. South should pass. He will

compete to 3 r if West bids again, but South shouldn’t bid 3 r
now when there is a good chance of buying the hand at 2 r. If the
overcall had been in spades there would be more reason to bid
3 r, for it would be more likely that West would bid again.

128
 GENERAL CONCEPTS REVIEW PROBLEMS

Now, let’s look at the same hands if East raises the overcall. The bidding
has gone:

West North East South

— — — 1r
2q 2r 3q ?

a. ♠ 432 r AQ865 q 96 ♣ AK8. South should pass. A balanced

minimum is not the right hand with which to compete. If North
has extra defense he may double. If he has extra offense (in
particular a fourth trump) he will compete to 3 r. But if North has
a normal minimum raise N-S are better off selling to 3 q —which
they might beat—than competing to 3 r, which almost certainly
won’t make and might get doubled.

b. ♠ 742 r AKJ76 q 5 ♣ AQJ8. South should bid 4 r. Who

knows? It might make, it might be the opponents’ hand, or they
might be goaded into a phantom sacrifice. Let them guess what
to do.

c. ♠ K9 r AKQ963 q 9 ♣ AQ108. South should bid 4 ♣. This

is not a slam try, as it would be if East had passed. South is
concerned about competition above 4 r, so he is describing a
game-going hand with club length and strength to his partner,
in the hope that North will know what to do if the opponents take
a sacrifice.

d. ♠ 104 r AKQ92 q KQ2 ♣ A75. South should double. This is

basically a penalty double, although South is unlikely to have a
real diamond stack after the suit has been bid and raised. The
double shows a strong balanced hand. North will usually pass
with an average flattish raise, but bid three or four hearts with a
more offensively oriented hand.

e. ♠ A5 r AQ9854 q 94 ♣ K63. South should bid 3 r. With the

sixth trump, competing is a must. It is not clear who can make
what, but 3 r figures to be right regardless.

129
 MATCHPOINTS

♠ rq ♣ ♠ rq ♣

It should be noted that some bids change in meaning in the face of

competition. For example:

Both vul., South holds ♠ 8 r AQ874 q Q4 ♣ AJ872

West North East South

— — — 1r
1♠ 2r 2♠ ?

South should bid 3 ♣. This is not a game try, as it would be if RHO had
remained silent. South is simply showing his values so that North may
be able to make an intelligent decision if the enemy competes to 3 ♠.
North should jump to game only with a perfect fitting maximum. The
3 ♣ bid does not create a force—the partnership can sell to 3 ♠ if there
is nothing more to be said. Let’s look at some North hands after the
auction:
West North East South
— — — 1r
1♠ 2r 2♠ 3♣
3♠ ?

a. ♠ K42 r J105 q 876 ♣ K1093. North will sell to 3 ♠. He doesn’t

have the offensive strength to compete to 4 r, and the club bid
warns him that the enemy hands are likely to fit well, so the best
bet is to sell and hope to beat it. Even if E-W score up +140 N-S
may not get a bad board, for some N-S pairs may do worse by
taking action over 3 ♠.

b. ♠ K42 r J105 q K1093 ♣ 654. North should double 3 ♠. There

are no guarantees with this double, but South’s 3 ♣ bid indicates
that North’s diamond cards are likely to produce tricks on de-
fense, since South figures to be short in diamonds. A “matchpoint
double,” protecting the potential +140 N-S might make in hearts,
is clearly called for.

130
 GENERAL CONCEPTS REVIEW PROBLEMS

c. ♠ 742 r K532 q K876 ♣ K3. North should try 4 r. This is the

type of perfect hand on which N-S figure to have a good play for
game even if South just has a distributional minimum.

An obvious question is: What should South do if he has a game try in

clubs? The answer is: Just bid a game and hope for the best. Maybe it
will make, or maybe the opponents will do the wrong thing (sacrificing
or selling) when they had a better auction available. There are several
ways for this kind of bid to win, so the game try is not so important.

♠ rq ♣ ♠ rq ♣

When you have a suit to show, it is better to show it right away than
to risk getting frozen out of the auction, even if you must overbid
somewhat. For example:

N-S vul., South holds ♠ 93 r 75 q KJ10652 ♣ KJ6

West North East South

— 1r 1♠ ?

Without the overcall 1NT is the correct response. The overcall changes
matters considerably. A 1NT response is no longer feasible, since South
has no spade stopper. If South passes, West may raise to 2 ♠ or even 3 ♠.
When this comes back to South, he will be playing guessing games. It
makes much more sense to bid 2 q immediately, even if the point count
is not there.
The importance of bidding your suits or raising partner when the
opponents step into the auction is so great that some bids that would
show extra strength without competition should be made on minimum
hands after an overcall. For example:

131
 MATCHPOINTS

N-S vul., South holds ♠ 83 r AQ754 q AJ85 ♣ J3

West North East South

— — — 1r
Pass 2q 2♠ ?

If East had passed, South would be required to rebid 2 r in many

partnerships, as a bid at the 3-level shows extra strength. After the
overcall, however, it is important to get the 3 q bid in now. Passing
would be all right if West also passes, for North will take some action
and now South can bid 3 q . Unfortunately, there is the danger that West
will raise to 3 ♠, and if this gets passed around to South (or even if North
doubles) South won’t know what to do. Life is much easier if South can
bid 3 q (without promising any extra values) on this hand.

Both vul., South holds ♠ 10 r KQ954 q AQJ96 ♣ 84

West North East South

— — — 1r
Pass 1NT 2♠ ?

South must bid 3 q . If South passes North will almost certainly sell to
2 ♠, and this figures to be a bad result. Without the overcall the 3 q bid
would have been a game-forcing jump shift, but the overcall makes it
strictly competitive, so North shouldn’t move on without an exceptional
hand.

N-S vul., South holds ♠ 5 r J4 q AQJ4 ♣ KQ10942

West North East South

— — — 1♣
Pass 1r 2♠ ?

South must risk a 3 ♣ bid. There is too great a danger that North will
sell to 2 ♠ with a flat 7- or 8-count. Obviously South doesn’t have the
values for a 3 ♣ rebid over 1 r, but after the 2 ♠ bid South must stretch a
bit on hands such this. South could still have a stronger hand, of course,
but North should take the conservative view if he has a close decision
as to whether or not to bid a game. As in most competitive decisions,

132
 GENERAL CONCEPTS REVIEW PROBLEMS

this conservatism by the partner of the free-bidder has an extra way

to win—it might be the opponents’ hand for 3 ♠ making or down 1.
Consequently bidding 3NT and going down when 3 ♣ was making may
be a more costly error than normal. There could be a full matchpoint
cost against tables where other N-S pairs sold to a spade partial and
collected a small plus score on defense.

♠ rq ♣ ♠ rq ♣

It is important not to overdo it with this kind of free bid if the bid takes
you to a higher level. For example:

None vul., South holds ♠ A5 r 984 q AQ10754 ♣ J2

West North East South

— — — 1q
Pass 1♠ 2♣ ?

South might as well bid 2 q , since this is what he would have normally
bid had East passed. The bid shows nothing more than a sixth diamond.
However, after:

West North East South

— — — 1q
Pass 1♠ 2r ?

South should pass. This hand just doesn’t have the playing strength to
make it worth going to the 3-level. If North takes another call South
can bid his diamonds later, while if North sells to 2 r or if West bids 3 r
and North passes the hand probably belongs to E-W and selling will be
all right.

♠ rq ♣ ♠ rq ♣

It must be noted that certain flat minimum hands do not need to be

bid in competition, even though they would have been bid without the
interference. For example:

133
 MATCHPOINTS

Both vul., South holds ♠ Q974 r Q82 q 96 ♣ Q974

West North East South

— 1q 1r ?

A Negative Double isn’t exactly a mistake, but pass is probably best.

This hand is so square that if North can’t find another bid then N-S
probably don’t have any reason to continue contesting the auction. Of
course if East had passed, a 1 ♠ response would be mandatory. On this
hand, however, action by South over the overcall is likely to induce
North to compete to too high a level.
Contrast with the following hand on the same auction:

Both vul., South holds ♠ K962 r 42 q J1096 ♣ 532

West North East South

— 1q 1r ?

South should certainly make a Negative Double after East overcalls his
partner’s 1 q opening with 1 r. It is important for South to get into the
auction because of the good diamond fit. Had East passed, South would
probably get best results by passing. A bid might get the partnership too
high.

134
 GENERAL CONCEPTS REVIEW PROBLEMS

PENALTY DOUBLES

Any time the opponents enter your constructive auction, you have
the option of doubling them. Obviously you must expect them to be
going down when you double, but this may not be sufficient. If all
you can make is a partscore, then down 1 vulnerable or down 2 non-
vulnerable is quite sufficient. If your side can make a game, however,
then your vulnerability is also a factor. In order to beat your game
score, you need a 2-trick set at favorable vulnerability, a 3-trick set at
equal vulnerability, and a 4-trick set at unfavorable vulnerability. If the
decision is close, you should go for the penalty if you are quite sure of
beating their contract for at least +200. This gives you several ways to
win. The penalty may be large enough to compensate for your potential
game, the game might not make, or even if game does make it might
not be bid at other tables. It only costs 1⁄2 matchpoint to other tables if
you double and are wrong (provided you collect at least +200), but if
you bid a marginal game and go down when you could have collected
more than the value of a partscore on defense you lose a full matchpoint
against other tables playing a partscore with your cards.
The key factor in close decisions is your side’s overall high-card
strength. With marginal game values you should double, with minimal
but sufficient game values you should tend to bid the game, and with
excess values for game you should usually double and go for the big
set. For example:

N-S vul., South holds ♠ 52 r KJ62 q A762 ♣ 542

West North East South

— 1NT 2r ?

South should double, assuming that double is defined as penalty. The

values for game are marginal, and there certainly is a likely +300
available. Strengthen South’s hand a bit to ♠ 52 r KJ62 q A762
♣ Q42. Now South should bid 3NT (via whatever mechanisms the
partnership has to show a heart stopper). This contract figures to
be secure, but collecting +800 is a bit optimistic. Here, South has
minimum but sufficient game values. Improve South’s hand a bit more
to ♠ 52 r KJ62 q A762 ♣ A42. Now South should go back on the

135
 MATCHPOINTS

doubling trail. Game is certain, of course, but the excess strength and
the heart stack figure to be enough to set East 4 tricks for +800 and a
top. This is a gamble, of course, but it seems worthwhile. A look at the
full hand illustrates the correctness of this strategy:

♠ AKQ3
r 53
N-S vul.
q KQ9
♣ K 10 8 6
♠ 10 9 8 6 ♠ J74
N
r4 r A Q 10 9 8 7
W E
q J 10 8 4 3 q5
S
♣ 973 ♣ AQJ

♠ 52
r KJ62
q A762
♣ 542

2 r doubled is down 2, +300, and N-S have no game. Give South East’s
♣ Q and 3NT is on, but 2 r doubled only nets +500. But give South
East’s ♣ A and now careful defense collects +800 against 2 r doubled,
beating the score for the 3NT game.

♠ rq ♣ ♠ rq ♣

When the opponents choose the wrong time to enter the auction, it
is important to double them and collect a number. This is usually
difficult to do at the 1-level, but penalty doubles at the 2-level can
be very profitable when your side holds the balance of strength. The
most important factor is your holding in the enemy trump suit. A good
general guideline is to go for a penalty if your side has six trumps,
but to simply bid to your normal contract if your side has only five
trumps. When the trumps are stacked behind the overcaller, it is easy.
For example:

136
 GENERAL CONCEPTS REVIEW PROBLEMS

♠98
r 52
None vul.
q AJ63
♣ AKJ85

♠AQJ64
r KJ87
q 10 7 2
♣ 2

West North East South

— — — 1♠
Pass 2♣ 2r Double
Pass Pass Pass

This should produce a comfortable penalty, and South’s hand is so

minimum that N-S might not be able to make a game, despite the game-
forcing two-over-one response. If the cards do lie favorably enough for
N-S to make a game, they may score +500 against 2 r doubled.
The last hand was easy, and most pairs would properly collect the
penalty. Let’s shift the hands slightly:

♠ 98
r J52
None vul.
q A 10 3
♣ AK875

♠AQJ64
r K87
q J62
♣ J2

N-S have the same cards, and once again it is right to defend. But how
do they know? Clearly South can’t double, for North would sit with a
weaker heart holding as in the first example and be wrong. The answer
is that North must do the doubling. The bidding should go:

137
 MATCHPOINTS

West North East South

— — — 1♠
Pass 2♣ 2r Pass
Pass Double Pass Pass
Pass

The double in front of the overcaller does not show a trump trick, as
does the double behind the overcaller. North’s double in this sequence
simply shows a fair balanced hand, no fit for partner, and three trumps.
Most players looking at the North hand would not consider a double,
yet it is clearly the winning action provided South understands that it
does not show a trump stack, and so he won’t sit without a reasonable
trump holding.
Lastly, suppose the player in front of the overcaller holds the long
trumps:

♠ 98
r KJ52
None vul.
q J3
♣ AK875

♠ AQJ64
r 87
q A 10 6
♣ Q62

You can’t get them all the time. With these combined hands and the
same auction the opponents will escape unscathed, for South should
not sit for the double with two small hearts. Note that this may not be
so bad for N-S. With the heart honors lying well for E-W the penalty
may not be so secure, and the club fit gives 3NT a good chance to make.
Since South holds club support, he shouldn’t even pass the 2 r bid
around to North. A normal auction might be:

West North East South

— — — 1♠
Pass 2♣ 2r 3♣
Pass 3NT Pass Pass
Pass

138
 GENERAL CONCEPTS REVIEW PROBLEMS

Without the overcall South would rebid 2 ♠, but once the opponents are
in the auction it is more important to show the club fit immediately in
case the enemy competes to a higher level. The 3NT contract has no
guarantees, but it is certainly reasonable and better than defending 2 r
doubled.

♠ rq ♣ ♠ rq ♣

The second main consideration is support for partner’s suit. If you are
short in partner’s suit you should be quick to defend, but if you have
support it is usually correct to show it first.

None vul., South holds ♠ K84 r 84 q J52 ♣ AQ963

West North East South

— 1♠ Pass 1NT
2r Double 1 Pass ?
1. Penalty

South should bid 3 ♠, just as he was planning to do before the overcall

complicated the issue. It is true that an occasional number may be
missed if North has a large heart stack, but there is a great danger that
the opponents have a double fit (hearts and diamonds) and the penalty
against 2 r may not be sufficient to compensate for a potential partscore
or game in spades for N-S. An additional argument for the 3 ♠ bid
is that even if North has a good double of 2 r, the spade fit may be
enough to allow 4 ♠ to make, which may score more than the penalty.
Consequently the 3 ♠ bid (as opposed to the pass) has an extra way to
win—there may be a spade game. If South’s third spade were a small
diamond or club, passing the double would be the indicated action.
The importance of a fit in partner’s suit when deciding whether or
not to defend is so great that it is necessary to strain to raise partner in
any competitive auction in order to avoid later problems.

139
 MATCHPOINTS

♠ 87
r J 10 5
N-S vul.
q KJ52
♣ AK43

♠ KQJ52
r 74
q A4
♣ Q862

West North East South

— — — 1♠
Pass 2♣ 2r 3♣
3r Double Pass Pass
Pass

If East hadn’t overcalled South would have rebid 2 ♠, planning on

continuing with 3 ♣ if North rebid 2NT. Once East bids, however, it
is essential for South to show his support now. If South had bid 2 ♠
he would be very uncomfortable passing the double of 3 r, for North
would also double with something like ♠ 87 r J105 q K2 ♣ AKJ543.
If North holds this hand, both 3 r and 4 ♣ are likely makes. Once South
has shown his club support, however, he can confidently pass the double
for he knows that North won’t be doubling with a long club suit.

♠ rq ♣ ♠ rq ♣

If the enemy interference is in the form of an artificial call such as a

takeout double, Michaels cuebid, or Unusual Notrump, the next hand to
bid has available a double of the artificial bid or redouble of the takeout
double to show strength and suggest willingness to defend. This should
be used on hands that have all the opponents’ suits but one (or better,
all the opponents’ suits) under control. The partner of the doubler or
redoubler should double himself if the opponents land in his second
suit; otherwise, he should pass it around to his partner unless he has a
singleton in the enemy suit. If he has a singleton a low-level double
will probably not be profitable, since the opponents are likely to have at
least eight trumps, so he should bid out his shape.

140
 GENERAL CONCEPTS REVIEW PROBLEMS

Most pairs can handle the following type of hand correctly:

♠ 54
r A Q 10 8 6
Both vul.
q 82
♣ KJ85

♠ AK973
r 43
q Q 10 5 3
♣ A4

West North East South

— — — 1♠
Double Rdbl Pass Pass
2q Pass Pass Double
Pass Pass Pass

North has shown his strength and willingness to defend by the redouble,
so he can pass it around to South. South has four trumps so he must
double despite his minimum opening bid, and the penalty is likely to be
worthwhile. If East had tried 2 ♣ South would pass it around and North
would double.
Problems arise on more balanced hands. For example:

Both vul., South holds ♠ A6 r 9852 q A873 ♣ Q32

West North East South

— 1♠ Double Rdbl
2♣ Pass Pass ?

Should South double? Most players would, but if North has only two
clubs he will pass the double and it could easily backfire. A better
approach is for South to pass first, then double. This should show a fair
balanced hand, but no great stack in the probable enemy suits. So after
the auction:

141
 MATCHPOINTS

West North East South

— 1♠ Double Pass
2♣ Pass Pass Double

North will pass with ♠ KQ852 r AJ4 q Q4 ♣ J74 but will pull to 2 r
with ♠ KQJ82 r AJ104 q Q4 ♣ 74 and he will probably be right both
times. On the first hand 2 ♣ should go down, on the second hand it could
well make and N-S have a comfortable heart partscore.
Let’s see how this might work against the Unusual Notrump. None
vul., the bidding has gone:

West North East South

— 1♠ 2NT ∗ ?

a. ♠ 63 r Q10854 q A6 ♣ KJ42. South should double. He plans to

double 3 ♣, and hopes his partner can double 3 q . If the opponents
have a safe landing spot in 3 q , N-S must bid on and hope to find
a decent contract.

b. ♠ 63 r Q54 q A1086 ♣ KJ42. South should double 2NT, plan-

ning to double either 3 ♣ or 3 q . Unless North is very short in the
minors, the penalty should be lucrative.

c. ♠ 643 r AJ54 q J86 ♣ A42. South should pass first, planning to

double either 3 ♣ or 3 q . This sequence shows a hand like this—a
balanced fairly strong hand but no stack in either minor. North
will know to sit for the double with ♠ AQJxx r Kx q Qxx ♣ Jxx
but to pull with ♠ AQJxx r KQxx q xx ♣ Jx.

142
HIGH-LEVEL INTERFERENCE

SACRIFICES

You have bid to a game that you felt you could make. The opponents
now proceed to take what is clearly a sacrifice. What should you do?
Should you double them and take the penalty, or should you bid on
in the hope of making a higher score on offense? This is a common
problem at matchpoints. Analysis of the potential matchpoint costs of
the various auctions will help put the problem in the proper perspective.
When your opponents take a sacrifice, they are betting on a parlay.
First, your game must make, and second, the save must not go for too
much. Actually there is a third part to the parlay—the field must bid
the game. This third part is not a problem for the game bidders: if they
have bid a marginal game that may not be bid at other tables, they won’t
have much trouble doubling. So for the rest of this discussion we will
assume that the rest of the field will surely bid the game, even though it
may not necessarily be a make. If the sacrifice is wrong on any part of
its parlay, it will be a very bad board for the savers, provided that their
opponents don’t take the push and bid one more.
If your opponents take a 5 q save over your 4 ♠ bid and you choose
to push on to 5 ♠, now you are the one who is betting on a parlay. You
are betting that 5 ♠ will make, and also that the 5 q sacrifice will not go
for more than the value of the game. If you are wrong on either count,
you should have doubled rather than taken the push.
It might seem right to bid on if you think that you are a favorite
to make your contract and if you believe, but are not sure, that the
opponents have taken a good save. This is not necessarily correct. The
problem is that the two are often closely inter-related. The finesse that
allows you to make your contract when it works may be the same finesse
that beats the save more than the value of the game, for it will lose for
the opponents. The following example illustrates this point:
 MATCHPOINTS

♠ K 10 6 4
r Q 10 2
None vul.
q 95
♣ AJ32

♠ A7532
r AJ85
q2
♣ KQ8

West North East South

— — — 1♠
2q 3♠ 1 5q ?
1. Limit raise

Let’s suppose that South could see North’s hand. What should he do?
A superficial analysis would be: If the spades split 2–2 or if the r K is
onside, then 5 ♠ will make. This makes 5 ♠ a favorite, so it is the right
bid since I’m not sure whether or not we can beat 5 q enough. The full
hand:
♠ K 10 6 4
r Q 10 2
None vul.
q 95
♣ AJ32
♠ QJ9 ♠ 8
N
r K9 r 7643
W E
q AQJ874 q K 10 6 3
S
♣ 10 4 ♣ 9765

♠ A7532
r AJ85
q2
♣ KQ8

Unlucky! Both suits fail to behave, and 5 ♠ goes down. Furthermore

E-W have taken a good save; they only have 4 losers for −300. But it
isn’t only bad luck. Suppose that either hearts or spades behave for N-S.
In that case, E-W would have one more loser in the suit that behaves,

144
 GENERAL CONCEPTS REVIEW PROBLEMS

and would be set 500 in 5 q , more than the value of a game. On this hand
if E-W do not have a singleton in hearts or clubs, it is never right to bid
on. Either 5 ♠ will go down or 5 q will be set at least 500, depending on
the lie of the major suits. The fate of both 5 q and 5 ♠ depends on the
same things—the E-W distribution and the location of the r K. Even if
N-S were vul. vs. not it would be very questionable to bid on. Twenty-
five percent of the time 5 ♠ will go down, and some of the time it makes
the same friendly lie of the cards will set the opponents 800.
Now, let’s look at the matchpoint costs. Suppose we choose to
double 5 q instead of bidding 5 ♠ and are wrong—5 ♠ makes and 5 q
goes down less than the value of the game. In this case we get +300
instead of +450, so we lose 1⁄2 matchpoint to every other table since
these are the only possible scores. It might be noted that we lose nothing
to any table that somehow fails to bid the game.
Suppose we bid 5 ♠ and are wrong. There are several ways to be
wrong. If 5 ♠ is down 1 but 5 q is a good save we go from +300 to −50.
This costs 1⁄2 matchpoint to other tables at which the decision was at the
5-level, but does not cost anything against tables where the save was not
taken, since we had no chance against them. Suppose 5 ♠ makes, but
5 q would have gone for 500. Then we go from +500 to +450, losing
1⁄ matchpoint to every other table since these are the only two possible
2

results. This indicates that we should be more inclined to take the push
when the question is whether or not 5 ♠ will make, rather than whether
we can get them enough. This is quite logical. In the first case the
assumption is that the opponents have taken a good save which means
that if we double them we may get a poor score even if we are right,
while in the second case it is not clear whether or not their save is a
good one.
There are, unfortunately, more serious errors to be considered.
Suppose both that 5 q was going for too much and 5 ♠ doesn’t make.
Now the decision to bid 5 ♠ swings our score from +500 to −50, costing
a full matchpoint to those in 4 ♠ making, which figures to be a popular
result. Suppose that the sacrifice is a phantom—4 ♠ was, in fact, going
down. Now our call turns our +300 into −100, which loses a full
matchpoint to the common result of 4 ♠ down 1 for −50.
What does all this tell us? If you are sure that your game was
making, and sure that either you can make at the 5-level or that the

145
 MATCHPOINTS

opponents’ save will not go for more than the value of the game then it
may well be right to bid on, with more of a tendency to bid on when you
know the opponents have taken a good save. However, if it is possible
that 4 ♠ wasn’t making or possible that 5 ♠ is going down and they will
go for too much in their save, you should definitely double. Bidding
on may cost a full matchpoint to most tables if wrong, while doubling
can cost at most 1⁄2 matchpoint to any other table. Only judgment and
experience can really answer the question on any given hand, but it is
better to know what we are risking by various actions.

N-S vul., South holds ♠ AJ96 r 96 q A98 ♣ KJ52

West North East South

— — — 1♣
Pass 1♠ 2r 2♠
3r 4♠ 5r ?

South should double. He has a minimum opening bid with no great

distribution. There is no reason to expect 5 ♠ or even 4 ♠ to make, and
no reason to be sure the opponents won’t go for 800. Admittedly E-W
have probably taken a good save, but South’s proper course of action
is to suggest defending. With a singleton heart South should pass the
decision to his partner.

N-S vul., South holds ♠ Q1042 r 94 q 86 ♣ KQ742

West North East South

— 1♠ 2q 2♠
3q 4♠ 5q ?

South should pass. He has an offensively oriented hand with a fourth

trump and good club suit, so he should not suggest defending. On the
other hand he can’t bid 5 ♠ himself, since for all he knows North was
stretching for the 4 ♠ bid. As a general rule, a limited hand should
never bid on over a sacrifice in the direct seat, unless his partner has
previously made a bid that suggested such an action on a suitable hand.
South can only pass or double, and North should be well-placed to make
the final decision.

146
 GENERAL CONCEPTS REVIEW PROBLEMS

N-S vul., South holds ♠ AQJ762 r KQ53 q 5 ♣ K2

West North East South

— — — 1♠
3q 3♠ 5q ?

South should bid 5 ♠. There are no guarantees, but it seems certain that
E-W will not go down 800 in 5 q , and equally certain that 4 ♠ would
have made. Consequently, doubling 5 q will probably lead to a poor
score even if it is right, since the save may not be taken at all the other
tables. Note that South cannot pass. This is not a forcing situation,
because N-S have not voluntarily bid game. South must decide one
way or the other, and the percentage bid seems to be 5 ♠.

♠ rq ♣ ♠ rq ♣

There is one type of situation where it is correct to push on in a close

decision. Occasionally it may sound like the opponents are sacrificing.
In fact they may think so themselves, but their save just might make. In
this case bidding on now has an extra way to win—maybe it is a good
sacrifice!

None vul., South holds ♠ 7 r AQJ632 q KJ107 ♣ A9

West North East South

— — — 1r
1♠ 2r 4♠ ?

South was planning to bid a game in hearts, expecting to be a favorite.

5 r will probably not make unless partner has the right hand, so it seems
as if South should double 4 ♠. However, there is a reasonable possibility
that 4 ♠ will make. I would bid 5 r as a two-way shot—maybe it will
make, or maybe it will be a good save. The full hand:

147
 MATCHPOINTS

♠ 864
r 10 7 4
None vul.
q AQ32
♣ 854
♠ AK952 ♠ Q J 10 3
N
r K9 r 85
W E
q 9865 q4
S
♣ QJ ♣ K 10 7 6 3 2

♠ 7
r AQJ632
q K J 10 7
♣ A9

East meant his bid as preemptive, and West’s hand is nothing to get
excited about. The hands fit well, however, and the offside r K that
defeats your 5 r contract is the onside r K that allows 4 ♠ to make. Once
again the bid that has more than one way to win comes through with a
surprising victory.

148
 GENERAL CONCEPTS REVIEW PROBLEMS

ENEMY PREEMPTS

High-level preempts by the opponents early in the auction can make

life very difficult. Not only do you not have enough room to find
your proper suit or level, but it might not even be your hand. There
is no absolute way to guarantee being right all the time, but there
are guidelines that will increase your chances of landing on your feet.
Remember that other pairs holding your cards are likely to face the
same problem.
The most important thing to realize is that extreme accuracy is
not nearly as important against preempts as it is in totally constructive
auctions. If you have the auction to yourself, it is crucial to get to the
very best spot most of the time. Second best will not score well. In
a jammed auction, however, second best will often score quite nicely.
This is because other pairs will do worse and land in third, fourth, or
fifth best contracts. Consequently, the cost of being wrong by finding
the second best as opposed to the best contract may not be great, for you
will still beat many pairs who land in yet worse contracts. For example:

♠ Q543
r AJ962
N-S vul.
q 65
♣ 83

♠ AK76
r 73
q K2
♣ K Q J 10 5

West North East South

— — 3q 4♣
Pass Pass Pass

I’m not endorsing the N-S auction; in fact, I have no idea what is best.
In an uncontested auction it would be a close decision between 3 ♠ and
4 ♠, and any pair who arrived at a contract other than a partscore or
game in spades could expect a very poor matchpoint score. After the
preempt, however, I would expect the 4 ♣ contract to score pretty well

149
 MATCHPOINTS

even though it isn’t optimal. Look at some of the bad things other N-S
pairs might do:

• They might speculate 3NT with the South hand and play it there.

• They might start with a takeout double and have North drive to
4 r, which is certainly not a favorite.

• They might go conservative and defend 3 q , doubled or undou-

bled.

• They might find their spade fit via a cuebid by North in response
to a takeout double, but overestimate their combined strength and
push past 4 ♠ looking for slam.

In addition to all this, if spades split badly, as they are more likely than
normal to do after the preempt, then 4 ♠ will not make and the club
partscore looks better still. The principle is to settle for any decent
landing spot after a preempt rather than look for perfection. Often this
means simply bidding a long suit instead of exploring other avenues.
Another important principle is to be willing to defend if the decision
is close. The preemptor doesn’t like this. Put yourself his shoes. Let’s
say that RHO opens 1 ♣ and you shoot out 3 r on a marginal hand.
If the opponents continue bidding you are always happy. Maybe they
will stumble into the right contract, but you know that your preempt has
done its dirty work of hampering their constructive auction and that you
have escaped unscathed. On the other hand, suppose you are dropped
in 3 r, doubled or undoubled. Now you are not so happy. You may be
a winner, but if partner puts down the wrong dummy you won’t like it.
You no longer have the feeling that you definitely have the best of it.
If the preemptor isn’t happy being dropped there, it follows that it
is often right to defend when an opponent preempts. Obviously I am
not advocating always defending; quite often it is your hand and you
simply have to try to get to the best contract. What I am saying is that
it is often right to sell when facing a close decision. For example:

150
 GENERAL CONCEPTS REVIEW PROBLEMS

Both vul., South holds ♠ K932 r J4 q AJ ♣ KJ643

West North East South

— — — 1♣
3r Pass Pass ?

Sure, N-S might have a partscore available in spades, clubs, or

diamonds, or there might be a large penalty available if North has a
penalty pass of a takeout double. The problem is that South has no
real reason to believe that there is a playable contract above 3 r, or that
North will find the right bid even if there is one. The one piece of
information that South has is that North was unable to act over 3 r, so
it is unlikely that N-S are missing a game. It is a close decision, but I
would pass it out at 3 r. This can’t be too far wrong, while a takeout
double or a 3 ♠ bid could be very wrong if North has the wrong hand,
and even if he has the right hand he may not find the winning response,
whatever it is.

None vul., South holds ♠ AJ6 r QJ9 q 74 ♣ AQ1082

West North East South

— — — 1♣
3r Double 1 Pass ?
1. Negative Double

South should pass. This should be at worst a reasonable result, and

could easily be the best spot available for N-S. Any of the other possible
actions (3 ♠, 3NT, or 4 ♣) could work out better, but they could all be
calamities. Even if it is right to bid, can South be so sure of finding the
right bid?

♠ rq ♣ ♠ rq ♣

When making a choice-of-games decision against an enemy preempt,

strain to play notrump if it is at all reasonable. There are two reasons
for this. First of all, the bad splits that are suggested by the preempt
may doom an otherwise sound suit contract by unlucky ruffs or terrible
trump divisions. Secondly, notrump will often put the preemptor’s hand
out of commission if you can hold up on his suit until his partner is out.

151
 MATCHPOINTS

The preemptor often does not have any side entries, so the usual timing
advantages of a suit contract are no longer needed. A simple holdup
play can suffice to stop the run of the enemy suit. For example:

None vul., South holds ♠ A3 r KJ84 q KQ2 ♣ KQ107

West North East South

— — 3♠ ?

South certainly must act, and a takeout double seems automatic. But
is it? Somebody has to bid notrump if it is right, and North is unlikely
to be able to do so. Even if N-S have a heart fit, the hand could well
belong in notrump. I recommend a 3NT overcall with the South hand.
It could really backfire if West has three spades or East has a side entry,
but the odds favor success. The full hand:
♠ 52
r Q653
None vul.
q J 10 9 7
♣ A86
♠ 86 ♠ K Q J 10 9 7 4
N
r A 10 9 2 r7
W E
q A84 q 653
S
♣ J942 ♣ 53

♠ A3
r KJ84
q KQ2
♣ K Q 10 7

Looking at just the N-S hands without any opposing bidding, 3NT
would be a terrible contract compared to 4 r. The 3 ♠ opening bid
changes matters entirely. Note that 4 r may go down if East has a
singleton in hearts, diamonds, or clubs. At 3NT, however, declarer
calmly ducks one round of spades and then goes about knocking out
West’s aces as East sits helplessly by, unable to contribute to the
defense. To add insult to injury West eventually gets squeezed in
clubs and hearts when declarer cashes dummy’s fourth diamond, and
notrump makes 10 tricks while 4 r goes down. This sort of result is

152
 GENERAL CONCEPTS REVIEW PROBLEMS

quite common when there is an enemy preempt, so it is wise to get to

notrump if possible. Make very aggressive notrump overcalls against
preempts, particularly if you hold the guarded ace of the preemptor’s
suit. You can anticipate holding up until the partner of the preemptor
is out, and then going about your business setting up tricks with one
opponent completely out of the hand.

♠ rq ♣ ♠ rq ♣

Another important point when dealing with preempts is for the strong
hand to be optimistic about game chances. If the decision is close, it is
usually right to overbid to get to game. For example:

N-S vul., South holds ♠ AJ64 r 8 q KQ42 ♣ A642

West North East South

— — 3r Double
Pass 3♠ Pass ?

If South had doubled an opening 1 r bid on this hand and heard his
partner respond 1 ♠, he would barely be worth a raise to 2 ♠. North could
have a yarborough, and any higher contract could be overboard. Against
the 3 r opening bid, however, it is correct to drive the hand to game!
South simply must assume that North has a fair hand, and South cannot
expect North to jump to game on some scattered collection such as
♠ Q10xxx r xxx q Jxx ♣ Kx. As a corollary, the partner of the strong
hand should discount his first 6 or 7 points when choosing his response,
because his partner will be playing him for them. On balance, this
approach will work well. The preemptor is assumed to have a weak
hand, so there are plenty of points to go around and the partner of the
strong hand will usually have at least a few of them. This approach will
permit fairly light action against preempts without fear that partner will
hang you when he has what you were hoping for when you overcalled
or doubled. For example:

153
 MATCHPOINTS

N-S vul., South holds ♠ A87 r AQJ1072 q 6 ♣ AQ9

West North East South

— — 3q ?

South should bid 4 r. He hopes that his partner “has his bids” (i.e., holds
a few high cards even though he hasn’t actually bid anything). If North
has a yarborough—unlucky! Against an opening one-bid South could
start slowly with a 1 r overcall, knowing that if North does nothing
there is probably no game, but against the preempt South must gamble.
It should be noted that at other tables E-W may be taking sacrifices
(phantom or not) at 5 q , so it is all the more important to shoot out a
game.

Both vul., South holds ♠ K63 r J53 q 1075 ♣ K952

West North East South

3q 3r Pass ?

South should pass. He “has his bid”—just what North is playing

him for. North may have risked an overcall on as little as ♠ Axx
r KQ109xx q xx ♣ Qx. This is dangerous, but defending 3 q when
both 3 q and 3 r are cold won’t be worth many matchpoints, and South
certainly wouldn’t have balanced. If South raises to 4 r, however, all
the good work of the 3 r overcall is undone. By bidding 4 r on this
hand South is in effect giving his partner a choice of defending 3 q or
playing 4 r. This is inconsistent, since 3 r could easily be the winning
contract, as it is here.

154
CONSTRUCTIVE BIDDING
REVIEW PROBLEMS

Problem 1

Both vul., South holds ♠ KQ8 r QJ63 q KQ76 ♣ 53

West North East South

— — — 1q
Pass 1r Pass 2r
Pass 3NT Pass ?
 MATCHPOINTS

Problem 1 Solution

Both vul., South holds ♠ KQ8 r QJ63 q KQ76 ♣ 53

West North East South

— — — 1q
Pass 1r Pass 2r
Pass 3NT Pass ?

4 r. South should go for the 4–4 heart fit. His intermediates in hearts
are strong, and his worthless doubleton in clubs also points to the suit
contract. Lack of aces does not necessarily mean that the hand is
notrump oriented, for N-S may need time to knock out those missing
aces, and only a trump suit can provide that time.

156
CONSTRUCTIVE BIDDING REVIEW PROBLEMS

Problem 2

E-W vul., South holds ♠ A9543 r K6 q 105 ♣ KQ106

West North East South

— 1q Pass 1♠
Pass 2♠ Pass 3♣
Pass 3♠ Pass ?

157
 MATCHPOINTS

Problem 2 Solution

E-W vul., South holds ♠ A9543 r K6 q 105 ♣ KQ106

West North East South

— 1q Pass 1♠
Pass 2♠ Pass 3♣
Pass 3♠ Pass ?

3NT. South is certainly strong enough to drive to game, and notrump is

still a possibility. Nothing can yet be concluded from North’s 3 ♠ bid
except that he has a minimum, for at that point South’s 3 ♣ bid was just
a game try. Now when South bids 3NT he clarifies his intentions to
North—choice of games with club concentration—and North can make
the final decision.

158
CONSTRUCTIVE BIDDING REVIEW PROBLEMS

Problem 3

Both vul., South holds ♠ Q8742 r 72 q 96 ♣ AQJ5

West North East South

— 1r Pass 1♠
Pass 2♣ Pass ?

159
 MATCHPOINTS

Problem 3 Solution

Both vul., South holds ♠ Q8742 r 72 q 96 ♣ AQJ5

West North East South

— 1r Pass 1♠
Pass 2♣ Pass ?

Pass. There is a chance that a marginal game will be missed, but a

3 ♣ bid is more likely to push the partnership overboard. North would
certainly bid 3NT with something like ♠ x r AK10xx q KQx ♣ Kxxx
and it won’t have much play. A heart preference could work if North has
strong hearts, but South has no way of knowing. The general principle
of taking the sure plus score on a partscore hand prevails here.

160
CONSTRUCTIVE BIDDING REVIEW PROBLEMS

Problem 4

None vul., South holds ♠ 943 r QJ87 q 76 ♣ A654

West North East South

— 1NT Pass 2♣
Pass 2r Pass ?

161
 MATCHPOINTS

Problem 4 Solution

None vul., South holds ♠ 943 r QJ87 q 76 ♣ A654

West North East South

— 1NT Pass 2♣
Pass 2r Pass ?

Pass. South should be happy where he is. His Stayman bid was
certainly questionable, but it seems to have worked well. Many players
will pass 1NT and probably take fewer tricks than N-S will take in
hearts, so South should not jeopardize his already favorable position
by pushing to a close game. Even if 4 r makes, +170 will beat those
playing 1NT.

162
CONSTRUCTIVE BIDDING REVIEW PROBLEMS

Problem 5

None vul., South holds ♠ 62 r Q85 q KJ7 ♣ AQ974

West North East South

— 1♠ Pass 2♣
2r Pass Pass ?

163
 MATCHPOINTS

Problem 5 Solution

None vul., South holds ♠ 62 r Q85 q KJ7 ♣ AQ974

West North East South

— 1♠ Pass 2♣
2r Pass Pass ?

Double. In front of the bidder this does not show a trump stack, but
a fair balanced hand with some defense. If North has three hearts
defending 2 r doubled will probably be lucrative, and North should
usually pull the double with a doubleton heart.

164
CONSTRUCTIVE BIDDING REVIEW PROBLEMS

Problem 6

E-W vul., South holds ♠ A32 r J2 q AQ2 ♣ AQ984

West North East South

— — — 1♣
Pass 2♣ 1 Pass 2q
Pass 3r 2 Pass 3♠
Pass 4♣ Pass 4q
Pass 4♠ Pass ?
1. 10+ points
2. Singleton heart

165
 MATCHPOINTS

Problem 6 Solution

E-W vul., South holds ♠ A32 r J2 q AQ2 ♣ AQ984

West North East South

— — — 1♣
Pass 2♣ 1 Pass 2q
Pass 3r 2 Pass 3♠
Pass 4♣ Pass 4q
Pass 4♠ Pass ?
1. 10+ points
2. Singleton heart

5 ♣. It looks like South must have at least a reasonable play for slam,
for the worst North could have is something like ♠ KJx r x q Kxxx
♣ Kxxxx, and any improvement makes slam laydown. However, South
should settle for 5 ♣. He made a fortunate decision not to open 1NT;
this will certainly be the popular choice at most tables. The likely result
at other tables is 3NT down 1 or 2 on a heart lead, for North probably
has a normal 3NT response despite his singleton heart. 5 ♣ making 5
or 6 should be worth plenty of matchpoints if 3NT goes down at other
tables, so South should not jeopardize his favorable position by bidding
the slam.

166
CONSTRUCTIVE BIDDING REVIEW PROBLEMS

Problem 7

N-S vul., South holds ♠ Q1086 r K10832 q 984 ♣ K

West North East South

— 1♣ 3q ?
(You are playing Negative Doubles.)

167
 MATCHPOINTS

Problem 7 Solution

N-S vul., South holds ♠ Q1086 r K10832 q 984 ♣ K

West North East South

— 1♣ 3q ?
(You are playing Negative Doubles.)

Pass. Either 3 r or a Negative Double could work well, but there just
may not be a safe landing spot. If North also sells out it might be the
best action available to N-S, and if North reopens South can reconsider
depending on North’s choice of actions.

168
CONSTRUCTIVE BIDDING REVIEW PROBLEMS

Problem 8

Both vul., South holds ♠ J4 r J74 q KJ3 ♣ AQ1092

West North East South

— 1♠ Pass ?

169
 MATCHPOINTS

Problem 8 Solution

Both vul., South holds ♠ J4 r J74 q KJ3 ♣ AQ1092

West North East South

— 1♠ Pass ?

2 ♣. With all those jacks, South is not quite worth a force to game.
His alternative is to start with 1NT and then show invitational values—
probably with a 2NT bid—at his next call. The problem with this
approach is that South conveys nothing about his hand besides his
point count, leaving North to guess not only whether to bid game,
but also which game to choose. How is North to know to bid 4 ♠
with ♠ KQ109x r 10x q AQx ♣ Kxx but to bid 3NT with ♠ KQ109x
r AQx q 10x ♣ Kxx? South should make the slight overbid of 2 ♣,
planning to continue with 3 q over North’s likely 2 q , 2 ♠, or 2NT rebid,
driving the hand to a possibly thin game but increasing the chances of
arriving in the best game.

170
CONSTRUCTIVE BIDDING REVIEW PROBLEMS

Problem 9

Both vul., South holds ♠ KQJ5 r 10762 q A2 ♣ A104

West North East South

— 1q Pass ?

171
 MATCHPOINTS

Problem 9 Solution

Both vul., South holds ♠ KQJ5 r 10762 q A2 ♣ A104

West North East South

— 1q Pass ?

1 ♠. A 4–3 spade fit looks good here, while a 4–4 heart fit might
not be missed too much. For example, suppose North holds ♠ Axx
r Axxx q KQxx ♣ xx. 3NT will probably only make 9 tricks, 4 r will
make 10 tricks unless the hearts split 4–1, but 4 ♠ will always make
10 tricks unless there are very bad splits, and possibly 11 tricks if a
heart-diamond squeeze materializes. There are other hands North might
hold, of course, but the 1 ♠ response just has to be the percentage action.

172
CONSTRUCTIVE BIDDING REVIEW PROBLEMS

Problem 10

N-S vul., South holds ♠ 98 r AK1062 q A6 ♣ KQ93

West North East South

— — — 1r
Pass 2r 2♠ 4r
4♠ Pass Pass ?

173
 MATCHPOINTS

Problem 10 Solution

N-S vul., South holds ♠ 98 r AK1062 q A6 ♣ KQ93

West North East South

— — — 1r
Pass 2r 2♠ 4r
4♠ Pass Pass ?

Double. South has committed himself to this action. If he had wanted

to probe intelligently he should have bid 3 ♣ rather than 4 r, to get some
help from his partner. Instead South blasted to 4 r, which certainly isn’t
necessarily cold, in order to make the opponents guess what to do. They
made their guess, and now South must assume that they have guessed
wrong and double. A 5 r bid would be very costly if it turned out that
4 ♠ were a phantom.

174
 Part III

Competitive Bidding
 THE FIGHTING GROUND

Most matchpoint enthusiasts will agree that more matchpoints are

swung by competitive partscore decisions than any other phase of
bidding. There is a logical reason for this. Most bidding decisions such
as which game to bid, whether or not to bid a game, whether to take
a sacrifice, etc., involve only two or three possible contracts at other
tables. Consequently, a losing auction is unlikely to cost more than
1⁄ matchpoint against another table. In a partscore competitive auction,
2

however, there are many different possible contracts and results. It is

not unusual to see the scores on such a board look as follows:

+ 470
+ 200
+ 140
+ 120
+ 110
+ 100
+ 90
– 50
– 90
– 100
– 110
– 300
– 670

With such a variety of results at other tables, any action is likely

to swing several matchpoints. Thus, accurate competitive partscore
bidding has high priority for a winning matchpoint player.
Where does the battle take place? If you stop at a cozy 2 ♣ or 2 q
contract, competent opponents simply won’t let you play there if they
have their share of the high-card strength and any kind of fit at all.
The opponent in the balancing seat will usually find some excuse to
 MATCHPOINTS

get into the auction. Conversely, suppose you bid up to 4 ♣ or 4 q in

a partscore battle. You may get doubled, but it is unlikely that the
opponents will compete higher unless the hand is very distributional.
It is at the 3-level where the critical partscore decisions are made. The
decision to compete to the 3-level over the opponents’ 2- or 3-level
contract is one of the most important decisions in matchpoints, and one
that recurs frequently. The 3-level is the battleground where partscore
fights are won and lost, and your bidding should be geared towards
making the right decision when this battleground is reached.
The most important principle is: In a competitive auction (or one
that is likely to become competitive), if you know that you will want
to compete to the 3-level and if it is probable that the opponents will
make you go there, then it is best to do it immediately rather than wait
for the enemy to describe their hands to each other. This often occurs
when you have a good trump fit, the enemy has a fit or a potential fit in
a higher-ranking suit than yours, and the high-card strength seems to be
evenly divided. Vulnerability may be a factor, but it is usually correct to
get to the 3-level yourself on all vulnerabilities in this type of situation.
That quick bid to the 3-level prevents your left-hand opponent from
bidding his suit or raising his partner at a comfortable level. When
the opponents hold the higher-ranking suit, they will have to make
the critical decision of whether or not to compete to the 3-level over
your 3-level contract. Your quick action at the 3-level often prevents
the opponents from exchanging the information necessary to make an
intelligent decision.

None vul., South holds ♠ 108 r KJ102 q K653 ♣ J97

West North East South

— 1r Double ?

South should bid 3 r. Call it preemptive, call it jamming, call it anything

you want, but it is the correct bid. We all know that if South bids 2 r
E-W will find their way to 2 ♠, either by a direct 2 ♠ bid by West or a
reopening double by East. After this happens, South will undoubtedly
want to compete to 3 r anyway, so he should do so now. Imagine poor
West holding ♠ KQxx r xx q Q10xx ♣ xxx. Over a 2 r call he has an
easy 2 ♠ bid, but over a 3 r call he is not so happy. 3 ♠ could easily

178
 THE FIGHTING GROUND

overshoot the mark, and even if 3 ♠ is right his partner may carry on
to 4 ♠ going down; East can’t be sure just how competitive the 3 ♠ call
is. On the other hand, if West passes his partner may have a moderate
distributional hand such as ♠ AJxx r x q AJxx ♣ K10xx. Should he
reopen over 3 r? It could be catastrophic if his partner doesn’t have a
fit, but selling out could be equally disastrous. Let’s look at the entire
hand:
♠ 754
r AQ9643
None vul.
q9
♣ AQ5
♠ KQ32 ♠ AJ96
N
r 85 r7
W E
q Q 10 7 2 q AJ84
S
♣ 832 ♣ K 10 6 4

♠ 10 8
r K J 10 2
q K653
♣ J97

It is clear that E-W belong in 3 ♠. If South bids only 2 r, they will

have no problem. West will bid 2 ♠, and East will compete to 3 ♠ over
North’s 3 r call. If South bids 3 r, however, E-W could easily either sell
out or overbid to 4 ♠ without either one of them making a particularly
bad bid. Notice that it is correct for either East or West to compete
over 3 r on this hand, but if the opposite hand weren’t quite so suitable
competition by East or West could be disastrous. The vulnerability was
set at none vul., but the 3 r call would be equally correct and effective
at any vulnerability.

179
 MATCHPOINTS

E-W vul., South holds ♠ 1063 r A8 q K32 ♣ KQ1075

West North East South

— — — 1♣
Pass 2♣ 1 Pass ?
1. Non-forcing.

What a nice cozy contract. Do you really think that South will be
permitted to buy it there? If West has played enough matchpoints
to know that defending 2 ♣ contracts is not the road to winning pair
games, he will be in there. South will definitely want to compete to
3 ♣ over whatever E-W get to on the 2-level, so he should bid it now.
South doesn’t know if E-W should be competing over 3 ♣ or not, and
he doesn’t care what they do. The important point is that an immediate
3 ♣ bid makes it more difficult for E-W to determine their best action,
whatever it might be.

Both vul., South holds ♠ Q4 r 876 q 108 ♣ KJ8765

West North East South

1r Double Rdbl ?

Without the redouble, South should bid 2 ♣. 3 ♣ would be an invitational

bid, and South’s hand simply isn’t worth it. The redouble changes
matters. Now, South should immediately bid as high as he is willing to
compete, which is to 3 ♣. Let the opponents try to find their best fit from
here, and they are more likely to do the wrong thing. The strength for
this kind of bid can vary from almost completely preemptive to nearly
invitational. It doesn’t really matter. All that is important is that South
is willing to compete to the 3-level. Of course if he tries the bid on
something like ♠ xx r QJ10x q xx ♣ AQxxx (I think he should), he
will then double the opponents if they bid 3 r, so the jump to the 3-level
can also be used as a trap. Only the bidder knows; the opponents must
guess.

180
 THE FIGHTING GROUND

None vul., South holds ♠ 102 r Q1065 q 973 ♣ A985

West North East South

1♣ 1r Double ∗ ?

South should go right to 3 r. It is virtually inconceivable that the

opponents will sell to 2 r, and South will compete to 3 r on his next
turn, so he should put the pressure on before they can determine whether
or not they have a fit. Again, the strength of this bid can vary from very
preemptive to nearly game invitational. Don’t worry about occasionally
missing a game. As we have seen this is not so important in competitive
auctions, and the matchpoints gained when the opponents misjudge
the situation because they have to guess at the 3-level will more than
compensate for any bidding inaccuracy on your part.

♠ rq ♣ ♠ rq ♣

Sometimes it is not necessary to go directly to the 3-level, even though

you are willing to compete there if pushed. This usually occurs when
your side holds the higher-ranking suit and there is some chance that
you may buy the contract at the 2-level. It is then not so important to
shut the opponents out of their fit, for they will have to go to the 4-level
to buy the contract, and if your assessment of the hand as a partscore
battle is correct this will suit you fine. Careful attention to the tempo
of the bidding and awareness of opponents’ problems will help you
determine whether or not the immediate 3-level bid is necessary. Some
examples:

E-W vul., South holds ♠ Q87 r A876 q K9 ♣ K632

West North East South

— 2♠ Pass ?

3 ♠ would not be a bad bid, but I think it is better to pass. There is

a reasonably good chance that you will buy the hand for 2 ♠, and this
might be all you can make. It is silly to push voluntarily to 3 ♠ down 1
when the opponents were about to sell to 2 ♠. The only argument for
bidding 3 ♠ immediately is to trap them into trying something at the
4-level, which you will be willing to test out with a double, but this is

181
 MATCHPOINTS

not too likely to happen. Of course if E-W find their way to 3 q you
will be forced to compete to 3 ♠, but it is not at all inconsistent to pass
now, planning on bidding 3 ♠ over 3 q . If East makes a takeout double,
however, you should definitely bid 3 ♠. Now E-W have committed
themselves to acting at the 3-level. Since you have no intention of
defending at this level, it is better to bid 3 ♠ and make them guess before
they have any more chance to exchange information. Of course, you
might also make the bid on a very weak hand, but one that did not have
enough offensive potential to justify an advance sacrifice of 4 ♠. Since
you could have anything, the 3 ♠ bid may goad them into stretching to
compete at the 4-level, and on this hand you will welcome that.

Both vul., South holds ♠ AQ10953 r K95 q K85 ♣ 9

West North East South

— — — 1♠
Pass 2♠ Pass ?

There is no reason to bid at this point. While you have every intention
of competing to 3 ♠ if the opponents get in the auction, you may have
already frozen them out. West, who was not strong enough to take
initial action, will think twice before stepping in at the 3-level after
you and your partner have exchanged information, and E-W may not
have a decent place to play the hand. Of course had West doubled
or overcalled, it would be another story. There would be a greater
likelihood of his having something else to say, so the 3 ♠ bid should
be played as preemptive and made on this hand to stop him from
conveniently describing his hand to his partner.
The type of action that must be avoided when you have the lower-
ranking suit is to fail to go directly to the 3-level when you have
the chance, allow the opponents to exchange information, and them
compete to the 3-level. For example:

182
 THE FIGHTING GROUND

Both vul., South holds ♠ 54 r K62 q KJ97 ♣ 9653

West North East South

— 1q Double ?

Are you willing to compete to the 3-level over the likely enemy two of
a major contract without any encouragement from partner? This is a
close decision that is up to you (personally, I would), but the important
point is that you must make that decision now with a 2 q or 3 q bid and
then stick with your decision. The auction to be avoided is:

West North East South

— 1q Double 2q
2♠ Pass Pass 3q

This is a losing sequence, yet many players fall into this trap. South
could hardly have been caught by surprise by the 2 ♠ bid, for it is
unlikely that E-W with fair high-card strength and a likely major-suit fit
will sell to 2 q . So, if South thought the hand was worth competing to
3 q on his own he should have done so immediately and put maximum
pressure on the opponents; if not, South should pass 2 ♠ and respect
his partner’s decision not to compete. Remember that North has heard
South’s 2 q bid and is quite capable of bidding 3 q himself if he thinks
it is the right action. North could well hold something like ♠ KQ10x
r Q10x q Axxx ♣ Qx. If this is North’s hand E-W have misjudged by
bidding 2 ♠, but only if South keeps quiet. If South bids 2 q and then 3 q
he is giving E-W a chance to escape if their initial 2 ♠ bid was wrong,
and making it easier for them to find the correct action over 3 q than if
South had bid it immediately. So either way, bidding 2 q and then 3 q
must be wrong.

♠ rq ♣ ♠ rq ♣

It is often necessary to raise partner or get your own suit into the auction
when competition starts on hands that normally wouldn’t call for these
actions. If the opponents are in the auction, you can assume it will be
a fight to the 3-level. In order to be prepared to win this fight, it is
important to lay the foundations early. Some examples:

183
 MATCHPOINTS

E-W vul., South holds ♠ AJ r J73 q AQJ42 ♣ 872

West North East South

— — — 1q
Pass 1r 1♠ ?

If East had passed, many players would prefer a 1NT or 2 q rebid on

South’s cards to a 2 r raise. This is largely a matter of partnership style,
depending on such things as how likely North is to have a weak 4-card
heart suit, whether or not the raise promises or implies four trumps,
whether North is likely to check back for 3-card support after a 1NT
rebid, etc. After the 1 ♠ overcall, however, a 2 r bid by South (or a
Support Double) is correct regardless of partnership style. If South does
anything else (pass, 1NT, or 2 q ), he is all too likely to hear a 2 ♠ bid
on his left. If this gets passed around to him he will have no idea what
to do now that the 3-level is approaching. 3 r will probably be right
if North has a 5-card suit and a fair hand, but it could be disastrous if
North has only four hearts. If South supports hearts immediately he can
confidently sell to 2 ♠, knowing that if North wasn’t willing to bid 3 r
then N-S don’t belong there.

Both vul., South holds ♠ J8 r 94 q QJ10653 ♣ K96

West North East South

— 1♠ Double ?

If East had passed, South would bid 1NT. The hand isn’t strong enough
for a 2 q response. After the double, South must bid 2 q . The bid no
longer shows the same strength that it would if East had passed. Now
if West bids the expected 2 r and this is passed around to South, he
will not feel compelled to bid on since he has already told his partner
about his diamond suit. If South passes over the double and West bids
2 r, South won’t have any idea what to do in the balancing seat. His
failure to prepare earlier puts him at a guess now that the critical 3-level
decision is at hand.

184
 THE FIGHTING GROUND

None vul., South holds ♠ KJ9862 r 73 q Q73 ♣ 92

West North East South

1♣ 1r 2♣ ?

South must bid 2 ♠. No matter that this is an overbid; the suit must be
bid now. If South passes West may very well bid 3 ♣ (particularly if he
has read this book!), and South will have egg on his face when it then
goes Pass–Pass to him.

E-W vul., South holds ♠ 632 r 75 q Q3 ♣ AQ10875

West North East South

1r 1♠ 2r ?

If East had passed, South might have tried 2 ♣, planning on following

with 2 ♠ if somebody took another bid, as somebody usually will. This
way, South gets his lead-director in and paints an accurate picture of
his hand so North will be well-placed to make the final decision should
E-W compete to three of a red suit. East’s 2 r bid changes the picture
considerably. No longer can South afford the luxury of a club bid. In
the first place the auction might die at 3 ♣, which would be bad if N-S
belong in spades as they probably do. Secondly, if E-W compete to
3 r South is not strong enough to contest to 3 ♠ on his own, and North
cannot help since he doesn’t know that South has spade support when
South bids 3 ♣. South must content himself with a 2 ♠ bid and hope
that his partner will find the winning auction if the opponents bid to the
3-level as they are likely to do. It is true that North won’t be able to
evaluate his club holding accurately, but at least he will have a fair idea
of South’s strength and spade support. You just can’t have everything,
particularly when the opponents make bids that gobble up your bidding
room. You simply have to do the best you can with what you have left.

♠ rq ♣ ♠ rq ♣

So far we have examined hands on which it is clear to drive to the 3-level

or describe your hand to partner so he can make the final decision.
What if it is not so clear, and it is your decision to make? The next

185
 MATCHPOINTS

chapter examines an extremely important concept that is the main basis

for making that decision.

186
 THE LAW OF TOTAL TRICKS

The problem of which side can make how many tricks with what trump
suit is the riddle that both pairs attempt to solve during the auction.
Unfortunately, the most accurate bidding system in the world will not
always lead to the right conclusion. Even if you could see partner’s hand
you would not always find the winning contract. Although you could
always get to the percentage contract, the location of a couple of key
enemy cards or the splits in important suits will in the end determine
how many tricks you take.
In competitive bidding, the lie of the cards works both ways. What
is good for you is often bad for the opponents, and vice versa. For
example, suppose your trump suit is:

North: r A Q 10 x
South: r J 9 8 x

If the r K is onside you will lose no tricks in the suit, while if it is

offside you will lose 1 trick. However, the location of the r K will also
make a difference to the opponents if they declare in some other suit.
Assuming that the suit splits 3–2, if the finesse is on for you then it is
off for the opponents and they will lose 2 tricks in the suit, while if it is
off for you then it is on for them and they have only 1 loser in the suit.
Note that there is a total of 2 tricks to be lost in the suit: either they have
2 losers and you have none or each side has 1. The location of the r K
determines which side loses how many tricks, but the total number of
losers between N-S and E-W remains constant.
Another example:

North: q A K x x
South: q Q x x x

If this is your trump suit you will have no trump losers if the suit splits
3–2, 1 loser if it splits 4–1, and 2 losers if it splits 5–0. Conversely, if
 MATCHPOINTS

this is a side suit for the opponents in some other trump contract then
they will have 2, 1, or 0 losers depending on how this suit is divided in
their hands. Once again, the total number of losers in the suit between
both pairs is constant, although the distribution of these losers depends
on how the suit is divided.
These examples seem to indicate that for an entire hand the total
number of losers for N-S and E-W may be a constant figure, even
though the distribution of these losers is unclear. If this is the case,
then the number of tricks available to N-S in their best trump suit plus
the number of tricks available to E-W in their best trump suit may be
constant. Let’s look at a full hand and see if this may be the case:

♠ K6
r K653
q AQ53
♣ J 10 9

♠ 854
r A9874
q 84
♣ A82

How will N-S fare in a heart contract? It depends a lot on the lie of the
opposing cards. Let’s suppose that everything works. Then N-S have
1 spade loser and 1 club loser, for 11 tricks. If this is the case, how will
E-W do in spades? They will lose 1 spade, 2 hearts, 2 diamonds (I’m
forgetting about the ruff), and 2 clubs, so they will take only 6 tricks.
Conversely, suppose hearts are 3–1 and all finesses lose for N-S.
Then N-S lose 2 spades, 1 heart, 1 diamond, and 2 clubs, so they take
7 tricks. E-W on the other hand lose only 1 heart, 1 diamond, and 1 club
for 10 tricks.
In both cases, the total number of tricks available to the two pairs
was 17. In the first case it was 11 + 6, in the second case 7 + 10. Try
arranging the E-W cards in various ways so that some finesses win
for N-S and some for E-W. Provided that E-W don’t have unusual
distribution such as a 5–2 club fit, the total number of tricks available
to the two pairs will remain at 17 regardless of how you distribute the
E-W cards.

188
 THE LAW OF TOTAL TRICKS

Now we notice something very interesting. N-S have 9 hearts

between them, and E-W have 8 spades. So the total number of trumps
held by the two pairs in their respective trump suits is also 17, the same
as the total number of tricks. Is this just a coincidence? It turns out
that this is a quite common occurrence. Examination of a large number
of hands has shown that the total number of tricks available to both
pairs (i.e., the tricks N-S can take if they declare in their best trump
suit plus the tricks E-W can take if they declare in their best trump suit)
averages out to be the total number of trumps held by the two pairs in
their respective trump suits. It quite often comes out exactly right, and
is seldom off by more than 1 trick. This remarkable fact is called the
Law of Total Tricks. It tends to break down at the 4- and 5-levels or
when one or both pairs have wild two-suit fits or a lot of distribution.
For the run-of-the-mill partscore hand, however, the Law of Total Tricks
is surprisingly accurate, and can be used very effectively to judge what
to do at the 3-level.
Let’s see how we can use the Law of Total Tricks to aid us in
competitive partscore decisions. First of all, look at what our goals
should be. Suppose the opponents are in 2 ♠, and we are considering
whether or not to compete to 3 r. If both contracts make, it is correct
to compete. If neither contract makes, it is correct to defend. What
if only one of the contracts will make and the other is down 1? This
is not so clear, but the scoring at matchpoints is such that it is usually
right to compete under these circumstances. The reason, of course, is
that a partscore scores more than +100 at contracts higher than two of
a minor. Consequently it is better to make a partscore than to defeat
the opponents 1 trick, and it is better to be set 1 trick than to allow
the opponents to make a partscore. The exception is when one or
both sides are vulnerable. If the opponents are vulnerable and you can
ascertain that they will be the ones who go down, then doubling them
and collecting +200 will be more profitable than making your partscore.
Conversely, if you are vulnerable and you believe that your contract will
be down 1 and the opponents will be sharp enough to double, then it will
be better to sell out. In practice this kind of pinpoint determination is
impossible, particularly about the doubles, so on balance it will pay to
bid on rather than sell if you think that one but not both contracts will
make. An added bonus is that the opponents may bid still one more,

189
 MATCHPOINTS

and if your initial judgment was correct this will definitely be favorable
for your side.
How can we incorporate the Law of Total Tricks into the decision-
making process? First of all, what about the opponents’ trump length?
On balance, you may assume that they have an 8-card fit. Occasionally
they may have a 7-card fit, but this is difficult to determine. Often they
will have a 9-card fit, and sometimes this can be diagnosed from the
bidding. Usually your side will also have an 8- or 9-card fit. We can
already see why it is almost always wrong to sell to a 2-level contract
when you can bid a higher-ranking 2-level contract. Even if both sides
only have 8-card fits the total number of trumps is 16, so the trick total
figures to be 16 also. Consequently it is likely that at least one if not
both sides can make their contract, so it is right to compete.
It is at the 3-level where the decisions get close. There are four
different possibilities, and we will examine them separately.

Case 1: You have an 8-card trump fit and the opponents have the higher-
ranking suit. For the sake of simplicity, let’s suppose for the rest
of this discussion that the suits are hearts and spades. This means
that the opponents have stopped in 2 ♠. You are considering
whether or not to compete to 3 r, and there are 16 combined
trumps. Therefore the opponents have done the right thing, since
the trick-total expectancy is also 16. Bidding 3 r would be a
1-trick “overbid” by trick-total analysis, and as we have seen it is
usually correct to make this overbid because one of the contracts
figures to make. Once again, vulnerability may be a deterrent
factor, even if a double is unlikely. For example, suppose both
sides are vulnerable. There is the possibility that the cards lie
very favorably for either you or your opponents, which means
that somebody may go for 200 even if they aren’t doubled. If
this is the case, it is better to sell out regardless of which pair
is unlucky. This doesn’t mean that selling out will necessarily
work; it simply means that with both sides vulnerable there is an
extra way to win by defending that wouldn’t exist if one or both
pairs weren’t vulnerable. Consequently, if the decision is really
close it is worth keeping an eye on the vulnerability. On the other
hand, it is possible that the opponents have a 9-card fit. In this
case, trick-total indicates that it is definitely correct to go to the

190
 THE LAW OF TOTAL TRICKS

3-level over their 2-level contract, for both contracts might make.
Furthermore even if bidding on is wrong you have another way
to win—the opponents may misjudge and bid one more. So all
other things being equal, it will usually pay off to bid on in this
situation.

16 Total Trumps (none vul.) 3 r vs. 2 ♠

N-S E-W N-S score for N-S score for
Tricks Tricks defending 2 ♠ declaring 3 r
7 9 −140 −100
8 8 −110 −50
9 7 +50 +140
10 6 +100 +170

Case 2: You have a 9-card heart fit, and the opponents are in 2 ♠. Now
the trick total is presumed to be 17, so it is clear to compete
to 3 r; there is a good chance that both contracts will make.
Furthermore, you now give the enemy a difficult decision as to
whether or not to bid on to 3 ♠. As we saw in several examples
from the previous chapter, hands that fall into this category
should usually be bid to the 3-level immediately, to make the
decision tougher for the opponents.

17 Total Trumps (none vul.) 3 r vs. 2 ♠

N-S E-W N-S score for N-S score for
Tricks Tricks defending 2 ♠ declaring 3 r
7 10 −170 −100
8 9 −140 −50
9 8 −110 +140
10 7 +50 +170

Case 3: You have an 8-card spade fit, and the opponents are in 3 r. The
presumed trick total is 16, and the enemy has already “overbid”
the trick total by one. Hence it is almost always incorrect to
191
 MATCHPOINTS

compete to 3 ♠, because it would be a 2-trick overbid by trick-

total standards. There is a good chance that both contracts will
go down 1 or that one of the contracts will be defeated 2 tricks; in
either case, bidding on is wrong. You should bid on only if you
have a two-suit fit, unexpected distribution, or reason to think that
the opponents have a 9-card fit, which would make your 3 ♠ call
only a 1-trick overbid.

16 Total Trumps (none vul.) 3 ♠ vs. 3 r

N-S E-W N-S score for N-S score for
Tricks Tricks defending 3 r declaring 3 ♠
7 9 −140 −100
8 8 +50 −50
9 7 +100 +140
10 6 +150 +170

Case 4: You have a 9-card spade fit, and the opponents are in 3 r. Once
again, the decision is close, since bidding on would be only a
1-trick overbid if the opponents have eight trumps. There is now
a better chance that they have nine trumps since they were willing
to compete to the 3-level. If so, bidding on will probably be right,
for then the trick total is 18, which means that both contracts
might well make. On the downside, you don’t have the extra
chance that you may push them one level higher, since they are
unlikely to compete to the 4-level without substantial distribution.

17 Total Trumps (none vul.) 3 ♠ vs. 3 r

N-S E-W N-S score for N-S score for
Tricks Tricks defending 3 r declaring 3 ♠
7 10 −170 −100
8 9 −140 −50
9 8 +50 +140
10 7 +100 +170

192
 THE LAW OF TOTAL TRICKS

All this can be summarized by a good rule of thumb: In a competitive

partscore auction, be willing to contract for the same number of tricks
as your side has trumps. In other words, you should generally compete
to the 3-level with a 9-card fit, but with an 8-card fit it is often correct
not to compete to the 3-level. This is just a general guideline and there
are plenty of exceptions, but it is surprising how often following this
rule leads to the correct action in a close competitive situation.
I must point out at this juncture that I have been discussing only
the decision of whether to bid or defend. There is, of course, a third
possibility, and that is to double the opponents. Doubling is really a
subset of defending, only you think that they are probably going down
or possibly that you will get a terrible result if they make it so you
might as well double even though they will make the contract over half
the time. This subject will be discussed in a later chapter. The best way
to look at a competitive partscore decision is to first determine whether
or not it is correct to bid, via the Law of Total Tricks. If you decide
that it is right to defend, only then do you consider whether or not to
double. The only time the doubling decision should come first is when
the double stands out, usually because you have a stack in the enemy
trump suit.

♠ rq ♣ ♠ rq ♣

Enough of this theory. Let’s look at some examples to see how the Law
of Total Tricks is applied in actual auctions.

None vul., South holds ♠ J73 r J7 q Q863 ♣ K973

West North East South

— 1♠ 2r 2♠
3r Pass Pass ?

South should pass. He knows that his side has only eight spades. How
does he know this? Because North would have bid 3 ♠ himself on
almost any hand with a 6-card spade suit, on exactly the same trick-
total reasoning. The opponents may or may not have a 9-card fit; that
is not clear. If they do then the 3 ♠ bid is only a 1-trick overbid and
may turn out right, but if the opponents have an 8-card fit the 3 ♠ bid is

193
 MATCHPOINTS

very likely to be a losing action. Even with neither side vul., which is
the vulnerability most conducive to aggressive partscore competition,
South is advised to pass. This will probably surprise a lot of players
who think that a 3 ♠ call is automatic in this situation, but the Law of
Total Tricks says otherwise.

E-W vul., South holds ♠ 97 r KJ43 q Q98 ♣ AK98

West North East South

— — — 1♣
1♠ Double ∗ 2♠ ?

South should bid 3 r. In the worst case he will be overbidding the

trick total by one, when North has four hearts and E-W have only eight
spades. As we have seen, this may not be bad. It should be noted that
this is not a constructive bid; competition is more important. If South
has a game-invitational hand in hearts he should just bid the game and
let the opponents guess what to do.

Both vul., South holds ♠ Q653 r AK62 q 8 ♣ A874

West North East South

— — 1q Double
Pass 1♠ 2q 2♠
3q Pass Pass ?

South should try 3 ♠. The bidding indicates that E-W have a 9-card
diamond fit, so North should, therefore, have at least a 4-card spade suit
since he probably has only three diamonds. Thus, South is overbidding
the trick total by at most 1 trick. The singleton diamond and the
excellent card placement (queen in the trump suit, aces and kings
outside), argue in favor of the 3 ♠ bid. It should be noted that if North
had held a 5-card spade suit he probably would have bid 3 ♠ himself
almost without looking at the rest of his hand, relying on trick total to
see him through. Also, South’s 2 ♠ bid does not show extra strength,
as it would if East had passed. The bid simply shows 4-card support,
which the takeout double doesn’t guarantee. This enables North to
make the proper competitive decision at the 3-level. If South had, say,

194
 THE LAW OF TOTAL TRICKS

♠ Qxxx r AQxx q xx ♣ KQx he should still bid 2 ♠ . The difference

is that he could then comfortably pass out 3 q , knowing then that a 3 ♠
bid would be too high. If South passes on this hand and West bids 3 q
as in the actual auction, South will have a very uncomfortable feeling
when it gets passed back to him, for both sides could have 9-card fits.
This shows why it is important that raises and free bids not show extra
values in competitive auctions.

Both vul., South holds ♠ K73 r QJ3 q K9842 ♣ 82

West North East South

— 1q Double ?

Many players would try a cute 1NT bid on this hand, planning to bid
3 q if E-W compete to two of a major or 3 ♣. This isn’t the worst bid
in the world, but it just doesn’t do the job. Trick-total analysis shows
that South must compete to 3 q anyway, and it is very unlikely that the
opponents will sell out to a lower contract if they have any strength
at all, so South should bid 3 q now. If South bids 1NT, West can get
his suit in at a safe level, and East can make the final decision at the
3-level. Over the 3 q bid, however, E-W don’t have that luxury. West is
in a bind, and he may sell out when it is his hand or bid at the wrong
time and take a minus score. If West does try three of a major, South
should bet that West has guessed wrong and double him. This kind of
bidding puts a lot of pressure on the opponents, and will induce them to
misjudge the auction quite often.

N-S vul., South holds ♠ 32 r KJ64 q 82 ♣ KQ654

West North East South

— 1♠ 2q Double ∗
Pass 2r Pass Pass
3q Pass Pass ?

South could speculate a double if he chose, although it would be pretty

chancy. A 3 r bid, however, is a definite no-no. With only eight trumps
and the opponents already at the 3-level, it is generally wrong to bid
on. On this auction, South can’t even be 100% sure that North has

195
 MATCHPOINTS

four hearts. Bidding on hands like this is a common error of aggressive

matchpoint players. The Law of Total Tricks says not to bid, and it is a
good law to obey.

Both vul., South holds ♠ Q53 r 875 q KQ3 ♣ AJ52

West North East South

— 2♠ Double ?

Over East’s double, South should bid 3 ♠ immediately. He will have

to bid it eventually according to the Law of Total Tricks, for West will
undoubtedly take the double out to his best suit. By bidding 3 ♠ now,
South makes the opponents guess immediately. Also, a delayed 3 ♠ call
is more likely to get doubled for the deadly −200, since the opponents
will be better able to judge the fit of the hand if they have a chance to
bid their suits. South can’t be sure that he can make it; in fact, he has
no idea whose hand it is. If he just follows the Law of Total Tricks, he
is likely to be right regardless of who can make what.

♠ rq ♣ ♠ rq ♣

Laws are made to be broken, and the Law of Total Tricks is no

exception. There are several factors that should sway the astute bidder
to overbid or underbid the total trick count.

1. Distribution. A singleton in the opponents’ suit is often sufficient

reason to stretch the Law a tiny bit. The reason is that they are
more likely to have a 9-card fit, so an overbid may not be that
much of a stretch. For example:

None vul., South holds ♠ AQ1096 r 3 q KQ83 ♣ Q103

West North East South

— — — 1♠
2r 2♠ 3r ?

Normally South should have a sixth spade for a 3 ♠ call in this sequence,
and the bid could easily turn out wrong. However, the singleton heart
suggests the overbid of 3 ♠. The opponents may well have a 9-card fit
196
 THE LAW OF TOTAL TRICKS

on this auction, and if North is looking at three spades and three hearts
he certainly won’t bid on, yet competing to 3 ♠ will probably turn out
best. If South’s small diamond were a small heart, a 3 ♠ call would
definitely be wrong.
On the other side of the coin, exceptionally flat distribution often
argues for conservatism. For example:

N-S vul., South holds ♠ K985 r 764 q A85 ♣ 852

West North East South

— 1♠ 2r 2♠
3r Pass Pass ?

South should pass. His fourth trump argues for a 3 ♠ call, but his total
lack of distribution says otherwise. He has a good clue that North
doesn’t have a singleton heart, since North might have bid 3 ♠ himself if
he did, and North also should have only five spades for the same reason.
If North has tripletons in both minors, prospects in 3 ♠ are rather bleak.
If South’s minor-suit distribution were 4–2 either way on this hand, the
3 ♠ call would be correct.

2. Two-suit fits. When you have a good fit for partner’s second suit,
it often pays to be a bit aggressive. The extra fit means that the
opponents are likely to have a two-suit fit also, and the second
suit is sort of another trump suit, so both you and your opponents
may take more tricks than the Law of Total Tricks would predict,
even on relatively balanced hands. For example:

E-W vul., South holds ♠ QJ83 r 98 q KJ76 ♣ 972

West North East South

— 1q 1r Double ∗
2r 2♠ 3r ?

North’s 2 ♠ bid doesn’t show any extra strength, of course, merely the
possession of four spades. Normally South would pass with an 8-card
fit and the opponents at the 3-level. However, the fit in diamonds,
North’s first suit, argues in favor of bidding on. A look at all four hands
illustrates the point:

197
 MATCHPOINTS

♠ AK54
r 532
E-W vul.
q AQ82
♣ 10 6
♠ 10 7 6 ♠ 92
N
r A74 r K Q J 10 6
W E
q 943 q 10 5
S
♣ KJ84 ♣ AQ53

♠ QJ83
r 98
q KJ76
♣ 972

No 9-card fits, no singletons, yet both sides can make 9 tricks. What
happened to the Law of Total Tricks? The answer is that both sides have
a good two-suit fit, so the hands will tend to make more than one would
normally expect. To illustrate how important the two-suit fit is, let’s
try interchanging the minor-suit holdings of the South and West hands.
N-S still make 3 ♠ with all the cards onside for them, but these same
finesses cause E-W to lose 2 tricks in each side suit and be down 2 in
3 r. This looks more like what the Law of Total Tricks would predict.
If the gods had been fairer with the distribution of the honors, then both
contracts would be down 1. With the new hand ( q xxx and ♣ KJxx),
South knows there is no double fit, so it is clearly correct to pass.

3. Location of intermediates. While aces and kings will score tricks

in any contract, queens and jacks are more effective trick-takers
on offense in long suits than in short suits. This is an important
principle of hand evaluation. We all know that ♠ Ax r KQJ10x
q Axxx ♣ xx is a stronger hand offensively than ♠ AJ r K10xxx
q Axxx ♣ Qx even though the “point count” is the same for both
hands. The reason is that queens and jacks are more likely to
score on the third and fourth rounds of suits, and this will only
be of help in trump suits or long side suits. If there is no third or
fourth round of a suit, the lower honors may not be of much value
on offense.
Conversely, queens and jacks in short suits may provide extra
tricks on defense that they don’t provide on offense. A holding
198
 THE LAW OF TOTAL TRICKS

of Qx opposite Jxx will produce at most one possibly useless

discard on offense, yet it is worth a full trick on defense if that is
the enemy trump suit or a side suit that splits 4–4 or 5–3. Queens
and jacks in long suits, however, are not likely to produce tricks
on defense, because the late rounds of the suit will be ruffed
by declarer. Thus, we say that the first hand in the previous
paragraph is offensively oriented, while the second hand with the
minor honors in the short suits is defensively oriented.

In competitive partscore bidding, we are concerned with both the

offensive and defensive potential of the hand. The Law of Total Tricks
gives us a good idea about who can make what, but this can be fine-
tuned by examining the location of the minor honors. If they are
concentrated in long suits, the trick total is likely to be higher than the
Law of Total Tricks suggests. If the queens and jacks are in the enemy
suit, however, this is an indication that the trick total may be lower than
we expect. For example:

None vul., South holds ♠ Q106 r A973 q AJ76 ♣ Q7

West North East South

— — — 1q
Pass 1r 1♠ 2r
2♠ Pass Pass ?

Presumably both sides have 8-card fits, since North didn’t try 3 r. With
neither side vulnerable and the opponents at the 2-level, it is usually
right to bid on, since 3 r would be a 1-trick “overbid” by the Law of
Total Tricks. The location of South’s queens should persuade him to
pass. These cards are all in short suits, so they are more likely to score
tricks on defense than on offense. Therefore, the trick-total count will
tend to be on the low side. The full hand:

199
 MATCHPOINTS

♠ 94
r K 10 6 4
q K98
♣ K842
♠ K52 ♠ AJ873
N
r Q82 r J5
W E
q Q 10 3 2 q 54
S
♣ J 10 5 ♣ A963

♠ Q 10 6
r A973
q AJ76
♣ Q7

N-S can make 8 tricks in hearts and E-W can make 7 tricks in spades,
one less than the Law of Total Tricks predicts, so selling out turns out
to be correct. The importance of the location of the minor honors can
be seen if we simply exchange South’s ♠ Q for West’s r Q, leaving
all distributions the same. Now hearts makes 9 tricks, while spades
makes 8 tricks, one more total trick than anticipated. This should make
a difference to South in his choice of actions. With ♠ 1062 r AQ93
q AJ76 ♣ Q7 the 3 r call is indicated.

200
 MATCHPOINT DOUBLES

The penalty double in a competitive partscore auction is one of

the most important actions for the successful matchpoint competitor.
Understanding when to double and when to sell out in a partscore battle
will swing many matchpoints for any duplicate player. The utopian goal
should be: Double when they are going down, don’t double when they
are making. This is analogous to the Rabbi’s Rule: When the king is
singleton, play the ace. It is true, of course, that the prime consideration
in doubling the opponents is whether or not you think they can make
their contract. However, often you are not very sure one way or the
other. In these situations, it is important to understand the matchpoint
odds you are getting on your double. Sometimes the double figures to
win far more matchpoints when it is right than it will lose when it is
wrong; at other times exactly the opposite is true. You want to double
frequently when there is much to gain and little to lose, but to avoid
close doubles when there is little to gain and a lot to lose.
First of all, let’s see how the matchpoint odds on a tight partscore
double differ from the IMP odds. We all know that it is wrong to double
the opponents into game at IMPs if it figures to be close. Suppose we
are not vulnerable against vulnerable opponents. We have bid to 3 r,
and they have outbid us to 3 ♠. Let’s further suppose that the contract
is 3 r at the other table, either making or down 1. We are considering
whether or not to make a close double of the 3 ♠ contract. Let us further
suppose that 3 ♠ is also either making or down 1. This last assumption
is consistent with our assessment that a double of 3 ♠ would be close.
If we double and are right, we go from +100 to +200. If 3 r makes
at the other table for +140, then we win 2 IMPs instead of losing 1 IMP
for a 3-IMP gain. At matchpoints, we gain a full matchpoint against the
other table. If 3 r is down 1 for −50, then the double takes us from plus
4 IMPs to plus 6 IMPs for a 2-IMP gain. At matchpoints, the double
makes no difference; we win the board anyway.
If we double and are wrong, we go from −140 to −730. If 3 r
makes, then we turn a 7-IMP loss into a 13-IMP loss for a 6-IMP cost.
 MATCHPOINTS

At matchpoints the cost is zero; the board was lost once the opponents
bid 3 ♠. If 3 r goes down, then we lose 12 IMPs for doubling instead
of 3 IMPs for passing for a 9-IMP loss. Once again, the double costs
nothing at matchpoints.
Putting all this together, we see that on the average we are risking
about three times as many IMPs by doubling as we stand to gain if we
are right. Consequently, we should be pretty sure about the double.
At matchpoints, however, the double gains a full matchpoint if it is
right against 3 r making at another table, and never costs if it is wrong
against other 3 r contracts. Therefore, if we judge that most of the field
will buy the hand our way for 3 r, which may make, then the double
is virtually a no-lose proposition that may gain a lot if it is right, so
we should double on almost any excuse. Another way of looking at
it is that we are headed for a very bad score if 3 ♠ makes whether we
double or not if the field buys the contract our way for 3 r. Therefore
the double will only turn a bad score into a zero if it is wrong, while it
will definitely net us a top if it is right.

Scenario 1: 3 ♠ and 3 r both make:

Score at other table: −140
Your score without X: −140
Your score with X: −730

IMPs without X −7 MPs without X 0

IMPs with X −13 MPs with X 0
Net IMP gain/loss −6 Net MP gain/loss 0

Scenario 2: 3 ♠ goes down and 3 r makes:

Score at other table: −140
Your score without X: +100
Your score with X: +200

IMPs without X −1 MPs without X 0

IMPs with X +2 MPs with X 1
Net IMP gain/loss +3 Net MP gain/loss 1

202
 MATCHPOINT DOUBLES

Scenario 3: 3 ♠ makes and 3 r goes down:

Score at other table: +100
Your score without X: −140
Your score with X: −730

IMPs without X −1 MPs without X 0

IMPs with X −12 MPs with X 0
Net IMP gain/loss −11 Net MP gain/loss 0

Scenario 4: 3 ♠ and 3 r both go down:

Score at other table: −100
Your score without X: +100
Your score with X: +200

IMPs without X +5 MPs without X 1

IMPs with X +7 MPs with X 1
Net IMP gain/loss +2 Net MP gain/loss 0

Does this mean that we can take a pot shot at the opponents every time
they outbid us, since we have everything to gain and nothing to lose?
Not at all. It is quite possible for the double to cost substantially when
it is wrong. The following occurrence in a Life Masters Pairs illustrates
this. With both sides vulnerable, our opponents had competed to 3 ♠
over our 3 r contract in an aggressive competitive sequence. I gave
some thought to doubling, but finally decided to sell out. They played
it well and made exactly 9 tricks, clearly all that were available, so I
naturally estimated an average-minus. Imagine my surprise when we
got 271⁄2 out of 29 matchpoints on the board! What had happened? One
look at the recap sheet told the story. The hearts split badly and so
8 tricks was the limit in our heart contract, and nearly everybody in the
room was doubled in 3 r or 4 r, going for −200 or −500. Obviously
I had made a good decision not to bid on, but even more important
was the decision not to double. If I had doubled we would have been
−730, losing to all those −200s and −500s instead of beating them, for
203
 MATCHPOINTS

virtually a full-board swing. This example is extreme, but it shows that

indiscriminate doubling in partscore battles can be very costly.

♠ rq ♣ ♠ rq ♣

In order to determine the potential gain or loss of a double, we must run

it through the matchpoint cost analysis. Let us suppose as before that we
are N-S and have bid to 3 r, the opponents have competed to 3 ♠, and we
are considering doubling. What are some likely results at other tables?
One possibility is that they also play 3 ♠, doubled or not. Against these
tables the gain and loss is easy to figure—you gain 1⁄2 matchpoint if
you double and are right, but lose 1⁄2 matchpoint if you double and are
wrong. Some of the other possibilities are more interesting:

1. E-W buy the hand for 2 ♠. This means that your 3 r bid
was somewhat of a push, not taken at other tables, and it was
successful in that you pushed the opponents to 3 ♠. In this case
the double gains nothing if it is right, for any plus score beats
−110, but the double costs 1⁄2 matchpoint if it is wrong, since
−140 would have tied the other table.

2. E-W bid up to 4 ♠. If you are considering doubling 3 ♠, then 4 ♠

will surely be doubled. If you are right you gain nothing, for
the pairs defending 4 ♠ will collect a larger number. If you are
wrong you lose nothing if exactly 9 tricks are the limit, since you
were dead anyway, but if 10 tricks are available the double costs
a full matchpoint, for the double will make you −630 or −930
depending on vulnerability, worse than −590 or −790.

3. N-S buys the hand for 3 r. This is likely to be a common result

if your decision is really close. Here the gain or loss from the
double depends a lot on what happens to 3 r. If 3 r makes, then
the double has nothing to lose if wrong, and may gain a full
matchpoint when it gets you a score greater than +140. This
will happen if you set them exactly 1 trick vulnerable or 2 tricks
non-vulnerable. On the other hand, suppose 3 r is going down.
Now the double gains nothing when it is right, but may lose a
full matchpoint when it is wrong if the penalty you avoided in

204
 MATCHPOINT DOUBLES

3 r, doubled or not, is more than the value of the opponents’

partscore.

4. N-S bid up to 4 r. If this makes, it won’t matter whether or not 3 ♠

is doubled unless the penalty against 3 ♠ is greater than the game
your way (which indicates that the double wasn’t so close). If
4 r is going down the double of 3 ♠ cannot gain, but it might lose
if 4 r is doubled and set more than the value of the opponents’
partscore.

So far, all we have seen is that sometimes tight partscore doubles have
more to gain than to lose, and sometimes they have more to lose than to
gain. The odds on the double are greatly affected by what is happening
at other tables. Let’s now examine the key factors that determine what
is likely to be going on around the room. What we are looking for
are factors that may cause the double to gain a full matchpoint against
another table if successful, or, conversely, cost a full matchpoint if
unsuccessful.

1. Vulnerability. This is very important when considering a close

partscore double. We have already seen that the primary gain for
a successful double comes when the increased penalty is greater
than the partscore you could have made, while the undoubled
penalty would be less. If our 3 r contract is making at other
tables and we double our opponents in 3 ♠, we only need to set
them 1 trick if they are vulnerable to outscore the 3 r contracts,
while a 2-trick set is necessary if they are not vulnerable.
Consequently, the double is more likely to gain when it is right
if the opponents are vulnerable. This type of successful double
is called “protecting your score.” You must double to protect the
+140 that you could have scored. Of course the opponents can
also see the vulnerability, so the downside of doubling vulnerable
opponents is that they are less likely to be out on a limb than if
they were not vulnerable, so they are more likely to make the
contract.
Not too many players realize that your own vulnerability
is also of prime importance when considering making a tight
double. It won’t matter if the double is successful, but what if it

205
 MATCHPOINTS

fails? The instinctive reaction is “If the double fails I get a zero,
so what difference does it make.” As we have seen, it can make
quite a difference. We are not so concerned with the matchpoint
score we get, but how many matchpoints our action will cost if it
is wrong. The greatest cost of doubling and being wrong comes
when pairs holding your cards go for numbers that are greater
than the partscore available to the opponents. Consequently, there
is potentially a much greater cost of doubling and being wrong
when you are vulnerable, for others may go −200 on your cards,
which your −140 would have beaten. If you are non-vulnerable
it is likely that −140 would be almost tied for bottom anyway, so
the double won’t cost as much if they make.
To summarize: Your most profitable partscore doubles occur
at favorable vulnerability, your least profitable at unfavorable
vulnerability, with equal vulnerability in between.

2. Balance of power. When your side holds more than half the
high-card strength, it is much safer to make a tight partscore
double. This is not only because high cards generate tricks, but
also because results at other tables will tend to favor the double.
Looking back to our 3 ♠ vs. 3 r example, if your side holds the
balance of power there figure to be a lot of pairs buying the hand
for 3 r, and they probably won’t get doubled. As we have seen,
this is the situation where the double gains the most if right and
loses the least if wrong. In other words, it may be “your hand,”
and you must double the opponents once they have outbid you.
Conversely, suppose the opponents hold the balance of power.
Now, many pairs your way may sell out to 2 ♠, and if they bid
3 r they may well get doubled because of the outstanding high-
card strength. If either of these results occurs at another table,
the double will not gain anything against that table if it is right,
but may cost a full matchpoint if it is wrong. To look at it another
way, you have pushed the opponents one trick higher and escaped
unscathed while doing so. It is likely to be their hand, so any
plus score should be good, and doubling risks turning a fair result
into a bottom if wrong. This is somewhat analogous to taking a
sacrifice against a game and pushing the opponents to the 5-level.
You would never double in this situation without having a sure

206
 MATCHPOINT DOUBLES

set, because just beating them will be a very good board. In a

partscore battle it is never clear who is pushing whom, but if the
opponents hold the majority of the high card points it is likely
that you are pushing them. Consequently, a double tends less to
be the indicated action.

3. The success of your contract. Suppose at another table your side

buys the hand for 3 r. If 3 r makes, then your double of 3 ♠ may
gain if it is right, but it will not cost anything to that table even
if it is wrong. Conversely, suppose 3 r is going down. Then
the double has nothing to gain if right, but everything to lose if
wrong. Consequently, you should be more inclined to double 3 ♠
if you liked your chances in 3 r, but less inclined to double if the
3 r contract seemed doubtful.

These factors are not necessarily independent. For example, the more
high-card strength you hold, the more likely you are to make your
contract. Never forget that the most important criterion of all for a
double is: Are they going down? If it is quite likely that they are, go
ahead and double regardless of potential gain and loss. You never know
what crazy things may be happening at other tables, so you might as
well collect as big a number as you can. Conversely, if it seems unlikely
that you will set them, then it is silly to double even if you anticipate
a bad board if they make. There is no reason to turn an average-minus
into a zero. Occasionally the field will surprise you, and your −140
becomes a good board in the most unexpected ways. These factors
should be taken into consideration only if it seems close as to whether
or not the opponents’ contract will make.

♠ rq ♣ ♠ rq ♣

There are two more points to consider. First of all, you should tend
to avoid doubling them if your double may tell them how to play the
hand. There is no more costly bid than a double that tells declarer how
to make a contract he would have otherwise gone down in. This is
more likely to happen at the game or slam level. Declarer can’t put too
much faith in the meaning of a tight partscore double, since you might
simply be doubling because you have been outbid and you think it is

207
 MATCHPOINTS

your hand. If the opponents have the preponderance of the high-card

strength, however, then declarer may deduce that your double is based
on a bad trump split and play the hand accordingly.
The other important point about doubling is that your own contract
comes first. If you believe that you can outbid your opponents and make
your contract or if the Law of Total Tricks suggests further competition,
then it is much better to bid on than to try a speculative double. The only
exception occurs when you are quite sure you can defeat the opposing
contract. In this case, if your score may be larger than what you could
get by bidding on you can forget the Law of Total Tricks—just double
and collect your number.

♠ rq ♣ ♠ rq ♣

In the following examples, there is a good bit of judgment involved

as to how likely various contracts are to make. If your assessment of
the situation differs substantially from mine, you may come up with a
different conclusion. Nevertheless, these examples will demonstrate
how to weigh the important factors to put the odds in our favor in
doubling situations.

E-W vul., South holds ♠ A53 r J64 q J9742 ♣ K3

West North East South

— 1♠ 2r 2♠
3r Pass Pass ?

This is a model matchpoint double. Trick-total argues against bidding

3 ♠. N-S figure to hold only eight spades since North failed to bid 3 ♠,
and E-W are likely to hold only eight hearts because North might have
pushed to 3 ♠ with a singleton heart. It appears close whether or not 3 r
will make. As for the other factors:

1. Vulnerability is perfect for the double. If you are right and

set them 1 trick you collect +200 for a top when +100 would
probably have been a bad score, while if you are wrong doubling
(as opposed to passing) does not figure to cost too much, since
N-S are not likely to go for more than 100 in a spade partscore.

208
 MATCHPOINT DOUBLES

2. N-S have the balance of strength since South has 9 HCP and
North has opened the bidding. This argues in favor of the double.

3. The 2 ♠ contract figured to make. This makes the double more

likely to gain when it is right and less likely to cost when it is
wrong.

The full hand illustrates the correctness of the double:

♠ K Q J 10 2
r 75
E-W vul.
q A65
♣ Q 10 4
♠ 874 ♠96
N
r 832 r A K Q 10 9
W E
q Q8 q K 10 3
S
♣ AJ872 ♣ 965

♠ A53
r J64
q J9742
♣ K3

Everybody’s actions are reasonable, yet both 3 r and 3 ♠ go down 1, as

the Law of Total Tricks suggests. You may claim that this is an unfair
example, because so much depends on those minor-suit tens, which
can’t really be considered in the bidding. If E-W had the ♣ 10 they
would make 3 r, while if N-S owned the q 10 then 3 ♠ would also make.
However, this is compensated for by the internal solidity of the trump
suits. For example, suppose North held the r Q instead of the ♠ Q, with
the distribution around the table otherwise the same. Now E-W must
lose an additional trick at hearts, and N-S may have an additional loser
at spades. As a percentage action the double stands out, but the bid
would not occur to many players, who would think their choice was
between pass and 3 ♠. On this hand the double will produce a top, pass
would be well below average since not all E-W pairs will compete to
3 r, and bidding 3 ♠ would be worst of all.

209
 MATCHPOINTS

Both vul., South holds ♠ K84 r J76 q K7 ♣ 97652

West North East South

— 1r 1♠ 2r
2♠ 3r Pass Pass
3♠ Pass Pass ?

South should pass. Obviously a 4 r bid is out of the question; North

would have bid game himself if it were close. Should South double?
It appears to be pretty much of a toss-up whether or not 3 ♠ will make.
What about the other critical factors?

1. Vulnerability. So-so. The opponents are vulnerable, which means

that the double is more likely to gain when it is right, but N-S are
also vulnerable, which means that other N-S pairs may be −200
in 3 r, doubled or not, so the double may cost substantially if it is
wrong.

2. Balance of power. Not clear, but South has no particular reason

to think that his side has more than half the high-card strength.
North’s 3 r bid doesn’t show extra strength—all it shows is a sixth
heart.

3. Success of 3 r. Again, not clear. It might make, but South has no

particular reason to think that it will.

All things considered, South is wisest to pass. It may be that E-W are
the ones who took the questionable push, and any plus score is likely
to be good. A typical hand for North is ♠ xx r AQxxxx q xx ♣ AQx.
Who can make what will depend completely on how the E-W cards
are divided. N-S could make anything from 10 tricks on a good day to
6 tricks on a bad day, while E-W might make from 11 tricks to 7 tricks
depending on the location of the key cards. On balance, doubling will
cost more when it is wrong than it will gain when it is right. Note how
the Law of Total Tricks comes up with 17 tricks, a very good estimate
of the combined trick-total for N-S and E-W.

210
 MATCHPOINT DOUBLES

None vul., South holds ♠ QJ943 r K1096 q Q4 ♣ 84

West North East South

1r Double 2r 2♠
3r Pass Pass ?

Although the double might be tempting, South should prefer 3 ♠. The

first priority is to bid your own contract, unless you are pretty sure you
can set the opponents more than the value of a partscore. South cannot
at all expect a two-trick set of 3 r; in fact, he might not beat it. The
Law of Total Tricks says to bid on, since N-S presumably have nine
spades between them. Admittedly North doesn’t absolutely guarantee
four spades with his takeout double, but he will have them far more
often than not. This makes a likely trick-total of 17, so with neither
side vulnerable bidding 3 ♠ is the indicated action. It would not be
correct for North to compete to 3 ♠ simply because he has four spades
and a singleton heart. South would certainly bid 2 ♠ with a 4-card spade
suit and four hearts, so from North’s point of view the trick-total might
easily be only 16. Therefore, it is South’s decision to make, and the fifth
spade points to the 3 ♠ bid. A typical North hand might be ♠ K10xx
r x q AKxx ♣ Qxxx. Opposite this hand the fate of 3 r is not clear—
down 1 is the probable result. 3 ♠ is very likely to make, so it is better
than defending 3 r doubled.
A small change in the hand or the conditions would make the double
a better choice. For example, make one of South’s small spades a
diamond, giving him ♠ QJ93 r K1096 q Q43 ♣ 84. Now the double
stands out. The trick-total is lower, and the hand certainly meets
the most important criterion for a double—South expects 3 r to go
down! Note that taking away one of South’s spades both improves the
defensive potential against 3 r, because the opponents are less likely to
hold a singleton spade, and decreases the offensive potential in 3 ♠, for
that ninth trump is a very valuable asset in the play. This is exactly what
the Law of Total Tricks anticipates.
Another change would be to make E-W vulnerable. Now the
arguments for the double are very strong in spite of the 5-card spade
suit. The double is simply a bet that they will not make it. If they go
down N-S get a top, while if they make E-W get a top. In this particular
situation N-S vulnerability does not matter, because South’s choices are

211
 MATCHPOINTS

doubling and bidding 3 ♠ rather than doubling or passing. The 3 ♠ bid

is likely to be an average result, while double is top or bottom. If South
thinks that 3 r will go down more often than not, then he should double.
It’s as simple as that.

N-S vul., South holds ♠ 86 r Q10853 q K72 ♣ J32

West North East South

— 1NT 2q 1 2r
2♠ Pass Pass ?
1. Spades and a minor

This is a frustrating hand. It seems as though South should take some

action holding 6 HCP opposite a strong notrump. Unfortunately, no
action makes much sense. North doesn’t have four hearts, or he would
have bid 3 r on a trick-total basis, and if he had three hearts and two
spades he might have tried 3 r anyway. Consequently, trick-total argues
against bidding on. What about doubling? The problem with that
is that South isn’t particularly optimistic about defeating 2 ♠; maybe
even money at best. The vulnerability argues against doubling, since
you would need a 2-trick set to compensate for a heart partial, and
that doesn’t seem very likely. The double can lose heavily at this
vulnerability if it is wrong, for other pairs holding the N-S cards may
go for −200 or more due to bad splits. The wisest course of action is to
pass and hope for some protection at other tables. You may be headed
for a below-average board, but that is still better than a bottom. The full
hand:

212
 MATCHPOINT DOUBLES

♠ Q95
r A6
N-S vul.
q AJ83
♣ KQ74
♠ 7432 ♠ A K J 10
N
r KJ97 r 42
W E
q5 q Q 10 9 6 4
S
♣ A 10 9 5 ♣ 86

♠ 86
r Q 10 8 5 3
q K72
♣ J32

Partner has full values, but you can’t do much damage to 2 ♠. 2 r, on the
other hand, will go down 2 if the defense is careful. Quietly selling out
to 2 ♠ will not lead to a bad score at all, but if South takes any further
action the only question is what number will he go for, −470 or −500.
This type of result is not at all uncommon when the opponents have a fit
and distribution, while your side has no great fit. Holding the majority
of the high cards doesn’t guarantee that it is your hand.

Both vul., South holds ♠ K72 r KJ104 q 94 ♣ Q652

West North East South

— 1♠ 2r ?

This time the Law of Total Tricks can be ignored. South simply expects
to beat 2 r, and that’s all that matters. Even if N-S hold eight spades and
E-W hold eight hearts, defending 2 r doubled figures to be the winning
action. South should make a Trap Pass, hoping for a reopening double
from partner. +200 beats any spade partscore, and South figures to get
at least that against 2 r doubled. If North holds a good enough hand to
make a spade game, N-S could easily score +800 on defense. It is true
that South would expect to make 2 ♠, but this is the one situation where
you don’t bid for your contract first; namely when you see a profitable
set.

213
 NUISANCE BIDS

The concept of a nuisance bid can best be understood by seeing the

effect it has on your own bidding. First, let’s look at:

None vul., South holds ♠ K32 r A9653 q 842 ♣ AQ

West North East South

— — — 1r
Pass 2♣ Pass ?

This is not a particularly comfortable rebid. 2 r on this weak suit is not

satisfactory, 2NT with an unbid suit completely unstopped is also an
unhappy choice, and anything else such as a 2 q rebid is likely to lead
to trouble later in the auction. Now, suppose East overcalls 2 q instead
of passing.

West North East South

— — — 1r
Pass 2♣ 2q ?

Suddenly, South’s problems are solved. He has a very easy and accurate
call available—a pass! This conveys the message that South has nothing
more to say, which is certainly true. In addition, South will be happy
to pass if North doubles. The interference was not a nuisance at all; in
fact, it made South’s bidding much easier.
Contrast with the following situation:
 MATCHPOINTS

None vul., South holds ♠ 94 r AQ973 q 8 ♣ KQJ32

West North East South

— — — 1r
Pass 1♠ Pass ?

Obviously South has no rebid problems at all. He has a very happy 2 ♣

rebid that goes a long way towards describing his hand perfectly. Now
let’s try it with East overcalling 2 q .

West North East South

— — — 1r
Pass 1♠ 2q ?

Suddenly, South is not so happy. This little overcall does a lot of

damage. South simply isn’t strong enough for a 3 ♣ bid, but if he passes
or bids 2 r the club suit is likely to be lost forever, particularly if West
now raises to 3 q .
It is interesting that in the first case the 2 q overcall solved South’s
problems, while in the second case it created problems where there were
none. This is what I call a nuisance bid. To some extent it has the same
effect as a preempt, although it is not a jump. A nuisance bid is simply
a bid that takes away from an opponent a bid he would like to be able
to make, such as the 2 ♣ rebid on the second hand.
Nuisance bids come in different shapes and sizes. The most
common form is a bid that blocks an opponent from bidding a suit
naturally that he would like to bid at a convenient level, but there are
several other possibilities, some of which can be quite subtle. Robbing
an opponent of a jump-raise (so that his raise may be competitive rather
than constructive), interfering with an artificial sequence and causing
confusion, or even stealing a cuebid from an opponent are a few of the
possibilities.
A player may make a bid in a competitive auction that appears
to serve no purpose. His side can’t buy the hand, there is no lead-
directional value to the bid, and a sacrifice is out of the question. All
the bid seems to do is give the opposition a fielder’s choice: they can
double and collect a number if it is right, or they can ignore the bid and
go after their own contract. Such a meaningless bid can still be correct
if it serves to hamper the opponents’ constructive auction in any way.

216
 NUISANCE BIDS

The wise player listens to the auction, anticipates the problems his left-
hand opponent is likely to face, and determines whether such a bid will
help or hinder the enemy. Also, if a player has a close choice of whether
or not to enter the auction or of what bid to make, he will do better to
choose the bid that will make life more difficult for the opponents by
stealing their bidding room.

N-S vul., South holds ♠ K9 r K84 q 95 ♣ KJ9754

West North East South

— Pass 1q ?

I don’t need to tell you the risks involved in a 2 ♣ overcall. At this

vulnerability it could be a disaster, even undoubled, and partner’s pass
indicates that the hand probably belongs to the opponents. Also there
is the danger that partner may play the overcaller for a better hand and
make a losing competitive decision later in the auction. Despite these
risks, I believe that is the winning action because it gobbles up so much
bidding space. It is true that most opponents have the Negative Double
available to help find 4–4 major-suit fits after interference, but there
may still be plenty of problems. Picture poor West holding something
like ♠ xxx r QJ9xx q Qxx ♣ Ax. If South doesn’t overcall, West has
an easy 1 r response and probably few rebid problems later. What
does he do over a 2 ♣ overcall? A 2 r bid risks getting the partnership
overboard, particularly if East has a minimum opening bid without heart
support. West could try a Negative Double, but we all know that East
will bid spades (partners always do in this situation), and then what?
2 q is all right on value, but it probably loses any heart fit forever. West
could pass and hope to recover later, but this involves some unpleasant
possibilities such as 2 ♣ getting passed out when E-W have a partscore
or a 3 ♣ raise passed around to West who now must make a guess at
the 3-level with nothing to go on. Personally I would bid 2 r on the
West hand since I believe it is best to be aggressive early in competitive
auctions, but I’m not particularly happy with the bid. Anybody who
says they are comfortable with their choice in this situation is probably
a liar. There is no doubt that this simple overcall can cause plenty of
problems for the opponents, enough to justify the risk.

217
 MATCHPOINTS

It is important to see that as the space consumed by the overcall

decreases, the nuisance value of the bid goes down, making it less
attractive. If the opening bid had been 1 r, only the spade suit would
be shut out by the 2 ♣ overcall, and even this might not be a serious
problem since many pairs play Flannery, so responder could assume
that his partner didn’t have four spades. Consequently, the risks of the
overcall probably aren’t compensated for by the nuisance value. If the
opening bid had been 1 ♠ the overcall wouldn’t shut out anything except
a 1NT response, which responder might be just as happy not making,
so there is no nuisance value at all and the 2 ♣ bid is clearly wrong.
In fact, the bid may actually ease responder’s problems, for it gives
him a Negative Double to play with that was otherwise unavailable.
For example, suppose responder had ♠ xxx r AQxx q KJxx ♣ xx.
Without the overcall responder must either bid 2 ♠, a slight underbid
that also loses a potential 4–4 heart fit if opener can’t move on, or bid
1NT, which risks both losing the spade fit (if opener passes) and getting
the partnership too high when responder jumps to 3 ♠ on the next round.
After the overcall he can make a Negative Double, convert 2 q to 2 ♠,
and he will have described his hand and permitted a stop at the 2-level.
Another consideration: Suppose West were a passed hand. Now
the 2 ♣ overcall over 1 q loses much of its nuisance value. The reason is
that West’s biggest problem hand, the 10-pointer with a 5-card major, is
no longer a great problem. West simply bids two of the major, and the
danger of getting too high is no longer present since West has already
limited his hand by his initial pass, so the bid is now non-forcing.
Once again, the loss of nuisance value now makes the already marginal
overcall no longer the percentage action.
The general principle illustrated in the preceding paragraph is that
the more the opponents know about each other’s hands, the less valuable
a nuisance bid becomes. For example:

None vul., South holds ♠ KJ972 r A9 q 107 ♣ 10975

West North East South

— — 1NT ?

After a normal notrump opening bid, a 2 ♠ overcall would be very

questionable. West knows approximately what his partner has, and

218
 NUISANCE BIDS

is therefore reasonably well placed to find a decent contract even

after the nuisance bid, so the opponents don’t figure to be too badly
hampered. In addition, West is in excellent position to double and
collect a number if South has picked the wrong time to try this bid.
Most experienced partnerships have conventional understandings such
as lebensohl that will help them cope with the 2 ♠ overcall if it gets in
their way (the bid would be more effective against a pair that does not
have these understandings). However, suppose the 1NT opening bid
were a “dynamic notrump” (part of the Romex system) showing a very
strong unbalanced hand. Now the 2 ♠ overcall will have much more
effect as a nuisance bid, since East hasn’t defined his shape yet. In
addition to causing problems for the opponents the bid will be hard to
double, since the opponents haven’t found out much about the hand.
Along the same lines, risky overcalls can be made freely over strong
artificial club openings since the shape of the hand is not defined, but
these same overcalls may be too dangerous over a standard 1 ♣ opening.
Conversely, more daring overcalls may be made against a standard
one of a major opening than against a Precision one of a major, since
the latter is very well-defined so the nuisance value of the overcall is
considerably less.

Both vul., South holds ♠ 8 r 982 q J1095 ♣ KQJ65

West North East South

— Pass Pass Pass
1q 1r Double ∗ ?

There is a lot to be said for a 2 ♣ bid by South. The lead-directional

value is obvious, and it is impossible to be dropped there since the
opponents have presumably located a 4–4 spade fit. There can be
little doubt that West’s next bid will be 2 ♠ (or 3 ♠ or 4 ♠ if his hand
is stronger), so 2 r by South has no more nuisance value than 2 ♣.
Since N-S have no intention of sticking their vulnerable necks on the
chopping block at the 3-level and since E-W will certainly compete to
at least 2 ♠, there is no real reason to raise hearts.
However, suppose on the same auction South holds ♠ J1095 r 982
q 8 ♣ KQJ65. Now a 2 r bid is better. The point is that with South
holding a singleton diamond it is quite likely that West has a long

219
 MATCHPOINTS

diamond suit that he will rebid in response to the Negative Double.

If South bids 2 ♣, West will rebid 2 q with a minimum opening and 3 q
with extras, and will have completed a good description of his hand.
Suppose instead that South makes the nuisance bid of 2 r. Things are
not so easy for West any more. What should he do with, say, ♠ Kxx
r xx q AQ109xx ♣ Ax? If he passes, the diamond suit may get lost,
particularly if North re-raises to 3 r, but if he bids 3 q his partner may
play him for a stronger hand and get overboard. If 3 q seems right,
then what should West rebid with a stronger hand such as ♠ AQx r xx
q AQJ10xx ♣ Ax? 3 q still seems right, but if he bids 3 q on both of
these hands East may have to guess whether to play his partner for the
stronger hand or the weaker one.

N-S vul., South holds ♠ 3 r K653 q KQJ3 ♣ Q843

West North East South

— — — 1q
Pass 2q Double ?

Offhand, it seems silly for South to bid 3 q . The opponents hold the
balance of strength and presumably at least nine spades, since North
failed to respond 1 ♠, so they can compete to 3 ♠ and might well have
a game. Why should South risk −200 if E-W decide unexpectedly to
double, when there doesn’t seem to be anything to gain? However,
close examination shows that there is a hidden gain to the 3 q bid. E-W
are close to a spade game, and must judge whether or not to bid it.
Even though South has no idea if this game will make, he wants to
make it as hard as possible for his opponents to assess their values. If
South passes, West can bid 2 ♠ on a moderate hand such as ♠ QJxxx
r QJx q xx ♣ xxx but jump to 3 ♠ on a somewhat stronger hand like
♠ QJxxx r QJx q xx ♣ Axx, and East is likely to judge correctly. If
South bids 3 q it is another story. West will be nervous about passing
on the first hand, for with the 5-card spade suit it is probably correct to
compete to 3 ♠ and he does not know if his partner has enough strength
or distribution to reopen. If he does bid 3 ♠ on this hand his partner may
play him for the stronger hand with the extra ace, on which a 3 ♠ bid is
also reasonable, and bid a game that doesn’t make. The 3 q call has the
nuisance value of denying West the option of voluntarily jumping to 3 ♠.

220
 NUISANCE BIDS

Consequently his 3 ♠ bid may be either competitive or constructive, and

it is very difficult for East to make a borderline decision about whether
or not to bid a game. If South passes West has a wider choice of actions
available, so East is more likely to go right since West’s hand will be
more closely defined.

None vul., South holds ♠ 32 r 1074 q A1096 ♣ QJ42

West North East South

— 3r 3♠ ?

What earthly good could a 4 r call do? E-W have already found their
spade fit and can certainly make game, which a 4 r won’t stop them
from bidding. It is possible that E-W can make 11 or 12 tricks in spades,
depending on whether or not North has the r A and how the E-W minor-
suit cards are distributed. South has no interest in sacrificing against an
E-W game or slam with neither vulnerable; in fact, 4 r might go down
too much if South bids it and E-W stop off to double. In spite of all
this, South should still bid 4 r. The reason is that it robs West of a
4 r cuebid. To illustrate this, let’s look at a couple of possible West
hands and see what his problems might be. First, suppose West holds
♠ Kxxx r xx q QJx ♣ Kxxx. West will bid 4 ♠ whether or not South
bids 4 r. No guarantees, but it should at least have a play. Now, let’s
add an ace to West’s hand, making it ♠ Kxxx r xx q QJx ♣ AKxx.
With this hand West has good reason to think of slam if his partner has
a strong overcall. Unfortunately if East has a minimal overcall such as
♠ AQJ10xx r xx q Ax ♣ Q10x even 5 ♠ might be in jeopardy. What
West would like to do is to suggest a spade slam without going above
the 4 ♠ safety level. If South passes West can bid 4 r, a cuebid that
carries exactly the desired message. However, if South bids 4 r West
can no longer make this cuebid. Consequently West is faced with the
choice of conservatively bidding 4 ♠ and possibly missing a slam, or
making a slam move above the game level and possibly getting to 5 ♠
down 1. It is not a happy choice, particularly since the 3 ♠ overcall has
a rather wide range. The 4 r bid by South has the nuisance value of
taking away West’s slam try below the game level.

221
 MATCHPOINTS

Both vul., South holds ♠ 3 r KJ10652 q AJ106 ♣ 106

West North East South

1♠ Pass 2♣ ?

South’s hand is below par for a vulnerable 2 r overcall against a two-

over-one response, and certainly risks going for a number in the face
of a misfit. In addition, it is unlikely that N-S will be able to buy the
contract. Does the 2 r overcall create enough of a nuisance to make it
worthwhile? A little thought will show that the 2 r bid probably has no
nuisance value at all; in fact, it is likely to help the opponents. South
has a singleton spade, so West may very well have a 6-card suit. If
South passes and West bids 2 ♠ East cannot be sure if this shows a
6-bagger, for West had to do something and 2 ♠ may have been his only
convenient rebid even if on a 5-card suit. However if South overcalls,
West doesn’t have to do anything. He can pass on a nondescript hand,
so his 2 ♠ rebid now definitely shows a 6-card suit. Consequently, the
2 r overcall is likely to help E-W rather than hinder them. But change
South’s hand to ♠ AJ106 r KJ10652 q 3 ♣ 106 and it is a different
picture. Now West is likely to have an easy 2 q rebid, and the 2 r
overcall will shut that out. In addition, if North has some heart support
a 3 r raise will probably be much more damaging than if South had a
singleton spade, for now E-W no longer have the higher-ranking suit
and may be faced with a very difficult decision.
It may seem like a little thing that the 2 r overcall in the previous
example is more effective when South has spades than when he has
diamonds, but these little things make the difference between winners
and losers over the course of a session. Anybody can make an obvious
overcall or avoid a ridiculous one. The player who consistently puts
the odds in his favor on the close decisions will in the long run have
an edge on his opponents. In this situation the 2 r overcall becomes a
better percentage action when it is more likely to damage the opponents
through its nuisance value, and this factor is sufficient to swing the
marginal overcall from a losing percentage action to a winning one.

222
 BALANCING

Balancing decisions present a unique problem in competitive bidding.

A player in the passout seat has the option of passing and ending
the auction right away, for better or worse. Since his partner has
done nothing but pass he knows less about the hand than in a normal
competitive partscore decision, so the total trick count is often a guess.
The decision of whether or not to balance is extremely critical,
and can swing many matchpoints. What to balance with can also be
important, but the big decision is usually whether or not to bid at all.
Paradoxically the conservative decision is usually to balance, while the
aggressive decision is to pass and end the auction. If you balance and
are wrong you will have several ways to recover. The opponents may
fail to collect a number that they have coming to them, they may be
pushed one level higher when they should have defended, or they may
fail to find a superior contract that became available when you entered
the auction. However, if you pass and are wrong it is all over—there
is no second way to win. Consequently, most close decisions in the
balancing seat should be resolved in favor of bidding.
There are four major types of balancing decisions, each of which
has its own characteristics. These are against an opening one of a suit
bid, against a preempt, after you have opened and an opponent has
overcalled, and after the opponents have found a low-level contract.
We shall examine each one separately.
 MATCHPOINTS

AFTER AN OPENING BID

The Law of Total Tricks tells us that it can virtually never be correct
to allow the opponents to buy the hand at the 1-level if they are in a
reasonable contract. Even if neither side can muster up an 8-card fit,
which is quite rare, the trick-total will still be 14, so it is a losing action
to sell out. When you pass out an opening one-bid, you are betting
that the opponents have made a mistake. It is unlikely that this mistake
is missing a game, for if they have a game usually either the opening
bidder would have opened 2 ♣ or 2NT or his partner would have found
a response. The more common mistake is that they have landed in the
wrong partscore, and if you bid you give them a chance to recover. This
is not so uncommon. How often have you opened 1 ♣ on something
like ♠ AQxx r KJx q Axx ♣ KQx, heard it go pass-pass, and prayed
that RHO would do something? Or haven’t you held a hand like ♠ —
r J10xxxx q Jxxxx ♣ xx and passed partner’s opening 1 ♠ bid in the
hope that LHO would reopen so you could get to a better contract? The
trick is to determine when the opponents have landed in the wrong suit.
There are three important factors to consider.
First of all, consider your overall strength. If your hand is strong
enough that there is a reasonable chance that your side holds the balance
of power then it is usually wrong to sell out unless you expect to
slaughter the opponents’ contract. Since there are many 12- or 13-point
hands on which your partner would not have acted, you should usually
bid with 9 or 10 points unless your distribution indicates that your
partner can’t have a good hand.
A second important consideration is length in the enemy trump suit.
If you have four or more trumps, you know that either partner is short
in trumps or the opponents are in a terrible contract. If partner is short
in trumps he would have found an overcall or takeout double with a
reasonably good hand, so you can safely assume that either he doesn’t
have a good hand or that the opponents are in a bad contract. Therefore,
you may pass moderate strength hands. Conversely if you have two or
fewer trumps partner may well have the balanced 13 or 14 count on
which he couldn’t find a convenient call because of his length in the
enemy suit. In this case, you should strain to reopen on almost any
hand unless you are extremely weak.

224
 BALANCING

A third consideration is shortness in an unbid major suit. If you

have a doubleton or especially a singleton, this is an indication to drop
the opponents in one of a minor if you have a close balancing decision
rather than stir up the animals. Who has the missing major? If it is
partner, his failure to overcall tells you that he does not have a good
hand, so the opponents clearly have the balance of power and may well
have a better contract. If partner does not have the missing major the
enemy does, so why give them a chance to improve their contract?
These factors are not independent. What it boils down to is the
longer you are in trumps and the shorter in unbid majors the more you
need in high cards to balance, while the shorter you are in trumps and
the longer in unbid majors the less you need in high cards and the more
you should strain to balance.
A few examples:

None vul., South holds ♠ 93 r KJ43 q Q1062 ♣ K42

West North East South

1q Pass Pass ?

Drop this one. You have only 9 HCP and the other factors (long
diamonds and short spades) argue for a pass.

N-S vul., South holds ♠ KQ93 r 105 q A963 ♣ 1087

West North East South

1♣ Pass Pass ?

It could be right to pass, but the strong spade suit argues for a 1 ♠
balance. If South’s spades and hearts were reversed a pass would be
more reasonable, because now the opponents might hold the ranking
major. Personally I would still balance, because I hate to stake the
entire board on this decision when I have so little information.

225
 MATCHPOINTS

Both vul., South holds ♠ Q64 r K64 q AJ43 ♣ J86

West North East South

1♠ Pass Pass ?

Bid 1NT. Who knows? When in doubt, bid rather than pass in a 1-level
balancing situation.

E-W vul., South holds ♠ 53 r A94 q AJ8 ♣ KQ1082

West North East South

1♣ Pass Pass ?

This is the type of hand on which to gamble a pass, in spite of the excess
strength. You have two ways to win, since the opponents are almost
certainly in the wrong suit. If it is not your hand you are quite happy to
sell out here rather than let the opponents improve their contract. Even
if it is your hand you may collect +200 against 1 ♣ and still not have a
game. The pass could easily backfire, but it is a good gamble. If the
clubs were shorter or if E-W were not vulnerable then the second way
to win by defeating 1 ♣ by more than the value of your partscore would
be less likely, so I would prefer a 1NT reopening.

♠ rq ♣ ♠ rq ♣

Once you have decided to balance, your action should be as natural

as possible depending on your distribution. If you have a decent suit,
bid it; if you have support for the other suits, make a takeout double;
and if you have a balanced hand, bid 1NT. Bidding your shape is
far more important than bidding your strength, since it is probably a
partscore hand and it is more important to concentrate on playing in
the right denomination than in getting to game. Don’t worry if your
hand isn’t perfect for the action. You may occasionally have to double
with less than ideal pattern, balance in a 4-card suit, or bid notrump
without a stopper in the enemy suit, but this is all part of the game.
The reason is that it is necessary to balance with hands on which you
would never consider taking direct action. Therefore, the requirements
for a balancing call cannot be as narrow as those for the same call in
direct position. There can never be much precision after a balancing

226
 BALANCING

action due to the extremely wide range of hands on which you must
balance, so accept this fact, make your most descriptive bid, and things
will usually work out reasonably well.

♠ rq ♣ ♠ rq ♣

There is one special problem involved in balancing that doesn’t occur

in normal auctions. If you hold ♠ KQ94 r — q A8543 ♣ K984 and
RHO opens 1 r you will, of course, make a takeout double. You won’t
like it if partner leaves it in, but this is such an infrequent occurrence that
it is not worth distorting your bidding to guard against it. However, if
you reopen with a double on this hand after 1 r-P-P, there is a significant
chance that partner has a heart stack behind the opening bidder and will
pass the double. Usually this won’t work out well unless you have at
least one heart to push through declarer. Consequently, it is best not to
double in the balancing seat on this type of hand. A balancing cuebid
should be used to show a void in the opponents’ suit and a hand on
which you can’t stand to have a double left in. Occasionally you may
miss a number, but you will come out ahead more often than not by
avoiding a double with a void in the balancing seat.

227
 MATCHPOINTS

AFTER A PREEMPT

While the same factors (strength, trump length, and shortness in

a higher-ranking suit) are still the critical factors when considering
whether or not to balance against a preempt, it is far more often correct
to sell out. There are two reasons for this. First of all, the Law of
Total Tricks isn’t necessarily hanging over your head the way it is at
the 1-level. If it goes 1 r-P-P to you, total tricks almost demands a bid
on any excuse. However if it goes 3 r-P-P to you, it may well be that
the opponents have already bid to or overbid the total-trick count, and
any further action by you would overshoot it still more. Secondly, the
partner of the preemptor knows exactly where he stands, and if you
guess wrong by bidding he will be sure to apply the ax. This is not true
at the 1-level; there you are likely to escape unscathed even when you
could be going for a number.
On the other side of the coin, there is an argument for bidding over
a preempt that doesn’t exist at the 1-level. If you drop the opponents
in one of a suit they may simply be playing in the wrong trump suit.
The preempting side usually belongs in the preemptor’s suit (although
perhaps not at the level of the preempt), so you have little fear of
pushing the opponents to a better contract by balancing. Consequently,
you should compete over a preempt if you feel you have a bigger plus
score coming to you than you would get on defense against the preempt
undoubled.

N-S vul., South holds ♠ Q10932 r 104 q A82 ♣ KJ3

West North East South

3r Pass Pass ?

A 3 ♠ bid could work out, but the odds are against it. South doesn’t
know much about the hand, but he can conclude that North probably
doesn’t have a good hand and four spades, since North failed to act
over 3 r. Consequently either N-S are outgunned or don’t have enough
spades to make 3 ♠ a winning action, so passing seems indicated. The
problem with bidding is that if it is wrong there is no recovery possible.
Passing can work out very well if West has picked a misfit on which
to preempt (note this is exactly the situation in which 3 ♠ will be

228
 BALANCING

worst, potentially a top-to-bottom swing), and the pass can still get a
reasonable score if 3 r is simply the normal level to which the hand
should be competed. If the opening bid had been 1 r a reopening would
be mandatory, and after a weak 2 r bid 2 ♠ is probably correct. This is
much closer, but the Law of Total Tricks indicates that reopening will
be right more often than not.

N-S vul., South holds ♠ 104 r KJ9 q AJ98 ♣ AQJ4

West North East South

3q Pass Pass ?

Passing could be right, of course, but South’s hand is just too strong to
risk selling out, and the penalty against 3 q undoubled might not even
compensate for a partscore. Personally, I would bid 3NT and hope for
the best. Even though this could go for a telephone number if East has a
good hand, I consider it a more conservative bid than the pass. Despite
my strong diamond holding, I can’t count on the opponents being in the
wrong suit after the preempt, as I might at the 1-level, so I must bid for
the plus score to which I think I am entitled.

♠ rq ♣ ♠ rq ♣

One more important point about balancing against a preempt is that you
should use the takeout double freely, even if somewhat off-shape. The
reason is that it is often right to defend when the opponents preempt,
and doubling leaves this option open while overcalling does not. For
example:

None vul., South holds ♠ AK105 r 95 q 93 ♣ KQ1032

West North East South

3r Pass Pass ?

If the opening bid had been 1 r, double would be a bad call. Partner
is likely to bid 2 q , and now what? It would be nice if partner were
planning to pass the double, but that just isn’t likely to happen. Either
1 ♠ or 2 ♣, depending on your mood and preference, is better than
doubling. After the 3 r opening, however, double is the best bid
229
 MATCHPOINTS

assuming you choose to balance. Partner may well leave it in, which
would probably be very good, or he could bid a black suit or 3NT. If he
goes all the way to 4 q he probably has a long suit, so things might turn
out all right anyway. There is no guarantee with any action, but double
is recommended because it is the most flexible bid and keeps all options
open, including defending.

230
 BALANCING

AFTER AN ENEMY OVERCALL

Suppose the bidding has gone:

West North East South

— — — 1r
2♣ Pass Pass ?

This time we do know something about North’s hand, based on what he

didn’t do. He certainly doesn’t have heart support unless he has a very
weak hand, or else he would have raised hearts to the appropriate level.
Also he doesn’t have the strength and/or suit for a new suit bid, and he
doesn’t have an appropriate hand for a Negative Double. He could have
a hand with which he wants to penalize the opponents, or he might have
some nondescript garbage that wasn’t suitable for any bid.
This time, the most important criterion is the number of cards held
in the overcaller’s suit. With two or fewer cards it is almost always
right to reopen. There is a good chance that partner is planning a penalty
pass, and even if he isn’t, total-trick analysis will show that some kind of
competition is probably better than selling out. Extra high-card strength
isn’t important. Even if you have a minimum, partner is bound to
have some cards or else the opponents probably would have bid more.
Furthermore, if you get in trouble it will be difficult for them to double
you because they don’t really have a very good idea of their combined
assets.
If you have three or more cards in the overcaller’s suit, you should
give serious thought to passing. A takeout double almost can’t be right
since there must be some suit for which you don’t have support. If you
don’t have the strength for a notrump reopening or an easy rebid of
your first suit or another suit, then passing is probably correct. Even if
partner has planned a penalty pass it would probably turn out all right,
for with your side having so many trumps the opponents undoubtedly
have a better contract, so the undoubled undertricks should amount to
a good score. If partner isn’t able to raise, make a Negative Double, or
bid a suit of his own he either is very weak or has at least three cards in
the opponent’s suit, in which case total tricks suggests defending.

231
 MATCHPOINTS

Both vul., South holds ♠ 103 r A10852 q KJ6 ♣ A62

West North East South

— — — 1r
1♠ Pass Pass ?

South should reopen with a double despite his minimum point count.
It is true that the hand probably belongs to E-W, but South just won’t
get rich if he sells to 1 ♠ with spade shortness. N-S probably have a
playable fit somewhere, and South must make some effort to get a plus
score either by buying the contract or pushing E-W up higher.

Both vul., South holds ♠ KJ975 r 75 q A92 ♣ AK6

West North East South

— — — 1♠
2q Pass Pass ?

South should sell out in spite of his extra strength. He just doesn’t have
any place to go. A takeout double with a doubleton heart doesn’t make
any sense. North is unlikely to have adequate spade support since he
didn’t raise, and North’s silence suggests that there is no future in the
hand. It would not be surprising if neither side had an 8-card fit, in
which case selling out is indicated by the Law of Total Tricks even at
this low level.

♠ rq ♣ ♠ rq ♣

The choice of reopening action should not be influenced much by the

possibility of partner’s having a penalty pass. It is necessary to be able
to handle a bid in your shortest suit if you reopen with a double. If a
penalty is missed that is just too bad, but it is more important to get to
the right contract. For example:

232
 BALANCING

E-W vul., South holds ♠ AQJ863 r 76 q A85 ♣ K8

West North East South

— — — 1♠
2r Pass Pass ?

South should bid 2 ♠. His partner might have been planning to sit for
2 r doubled, but if so N-S should still get to a reasonable contract. The
danger in doubling is that North will bid three of a minor when the best
contract is quite likely to be 2 ♠.

♠ rq ♣ ♠ rq ♣

If the opponents’ overcall is a preemptive jump, pretty much the same

criterion for reopening applies. The only difference is that with a
doubleton in the enemy suit and a minimum opening bid you might
tend to sell out. The reason is that the opponents may have already bid
to or surpassed the trick-total, and any action by your side could well be
overboard. Also, the partner of the preemptor may have a strong hand
and be in position to punish you if you step out of line. For example:

N-S vul., South holds ♠ K63 r AJ954 q 84 ♣ AJ9

West North East South

— — — 1r
3q Pass Pass ?

If West had overcalled 2 q South would have a routine reopening

double, but after the 3 q overcall it is probably right to pass. North
will strain to raise to 3 r on any decent hand with heart support, and
N-S certainly don’t have a game or North would have found some bid.
Therefore it is best to assume that the pushing to the 3-level has already
been done, so there is no reason to compete further.

233
 MATCHPOINTS

THE OPPONENTS HAVE ARRIVED

This is the most difficult balancing decision of all. The opponents have
bid to a low-level contract, accurately describing their hands to each
other, while you and your partner have remained silent. Now it is two
passes to you, and you must decide whether or not to balance. It is the
most dangerous type of balance, for both opponents know something
about their partner’s hand and they will be quick to double you if you
are wrong. Nevertheless, this type of balance is often necessary to get
matchpoints on a board.
The following is a classic example of such a balance:

None vul., South holds ♠ Q1052 r 54 q AJ95 ♣ Q63

West North East South

— — 1r Pass
2r Pass Pass ?

South should reopen with a double. He didn’t have the strength to

double the opening bid, although his shape is right. Now that the
opponents have subsided in 2 r, he must get in the auction, and the
double is the best descriptive bid. Note that South is in some sense
trusting his opponents to make his bid correct. Since East made no
move toward game E-W should have at most 23 or 24 HCP, which
means that North figures to have at least 8 or 9, probably more. Also
E-W probably have at least eight hearts between them, which increases
the likelihood of N-S having an 8-card fit and making it correct to bid by
total-trick analysis. Of course the reopening could backfire, particularly
if E-W have only a 7-card fit, but on balance it will pay off.
The main indicators for this type of balance are that the opponents
have found a fit, which suggests that your side also has a fit, and they
have stopped at the 2-level, where you may be able to outbid them
or push them one level higher. Vulnerability is a far more important
consideration here than in other balancing situations. The opponents
know where they are, so they will be able to double you if you step
the least bit out of line. Consequently you must be very cautious when
you are vulnerable, for down 1 doubled will be −200 and a bottom. If
you are not vulnerable a close balance is more attractive, for you still

234
 BALANCING

have an extra way to win—down one for −100 when their contract is
making.

E-W vul., South holds ♠ K985 r A9 q 1043 ♣ Q542

West North East South

1q Pass 1r Pass
2q Pass Pass ?

A reopening double could work out, but passing will be right more often
than not. The opponents do not necessarily have a fit, which means
that N-S might not have one either. A particular concern is South’s
doubleton heart. If the opponents don’t have an 8-card heart fit, as
suggested by their choosing diamonds rather than hearts, then North is
marked with at least four hearts. Consequently North can’t have both
four spades and a reasonable hand, or he would have made an initial
takeout double, so the odds swing against balancing. If West had raised
to 2 r, the balancing double would be much more attractive.

Both vul., South holds ♠ 109 r 93 q KJ52 ♣ KQ1085

West North East South

— — 1♠ Pass
2♠ Pass Pass ?

Should South back into the auction with the Unusual Notrump? It could
be right, but the vulnerability argues for selling out. If 3 ♣ or 3 q is going
down it is quite likely to be doubled, and a 1-trick set is −200 for a
bottom. If South were not vulnerable then 2NT would be the percentage
bid, for three of a minor down 1, doubled or not, could still be a good
result, giving South a second way to win. If the opponents were bidding
hearts and one of South’s suits were spades the balance would again be
more attractive, since it might only be necessary to contract for 8 tricks.
As it is, the odds seem to be slightly against balancing, so the pass is
recommended.

235
 THE LAST GUESS

It is apparent that if you can put your opponents in a position where

they must guess what to do, then you have a good chance of getting
a favorable result, for they may guess wrong. Conversely, if you are
the one who has to make the guess, then your prospects on the board
are not too good. It follows that it is to your advantage to maneuver
competitive auctions so it is the opponents and not your side who have
to make that last guess.
Let’s look at a couple of similar examples that illustrate this point:

N-S vul., South holds ♠ AKJ986 r A8 q 76 ♣ KJ4

West North East South

— — — 1♠
2q 2♠ 3q 4♠
?

West is thinking about taking a 5 q save. How do you feel sitting South?
Pretty good! You know exactly what to do if West bids 5 q . It certainly
isn’t right to bid 5 ♠; in fact, the 4 ♠ call was somewhat of a stretch. You
have put the opponents in the position of having to make the last guess
on the hand, namely whether or not to sacrifice. You don’t particularly
care which way they go since you have no idea whether or not 4 ♠ will
make, but you know that if they guess wrong you are headed for a very
good board. The key is that if they bid 5 q you don’t have any guess—it
is quite clear to double. Consequently, E-W have to make the last guess
on this hand.
Contrast with the following hand:
 MATCHPOINTS

N-S vul., South holds ♠ AQJ1072 r AQ2 q 3 ♣ Q72

West North East South

— — — 1♠
Pass 3♠ 1 4q 4♠
5q Pass Pass ?
1. Limit raise

E-W certainly seem to have taken a good sacrifice. 4 ♠ figures to be

cold, and it is extremely unlikely that 5 q can be set more than three
tricks. The question is: Should South bid 5 ♠? If he does, E-W will
have no problem passing. They won’t consider taking a 6 q sacrifice,
for fear that either it will go for too much or that N-S can’t make 5 ♠.
E-W will be happy to have pushed N-S to the 5-level. Consequently,
South must make the last guess on this hand, and it is a difficult one.
If South guesses wrong he will wind up with a very poor result, while
a right guess will probably only be a normal result. South might have
alleviated his problem by bidding 4 r rather than 4 ♠ to give his partner
the picture, but there would still be a problem. South is not in a
comfortable position. He must make the last guess on the hand, and
he must guess right to avoid a bad score. East probably has no idea
what South’s right action is and doesn’t really care what South does.
He is just hoping that South guesses wrong.

♠ rq ♣ ♠ rq ♣

Your goal in a competitive auction should be to maneuver the bidding

so that the opponents have to make that last guess. Naturally this will
not always be possible, particularly if the opponents are trying to do the
same thing. You know that you have achieved this goal when you have
bid the limit to which your side will compete, and you don’t know what
the right action for the opponents is. Since you have bid the limit of the
hand you do not have a guess if they bid on, except perhaps whether
or not to double. If it is clear to you either that the opponents should
defend or that they should bid on it will usually be clear to them also, so
they won’t really have a guess, but if you don’t know what their correct
action is the chances are that they have a really close decision and may
well guess wrong.

238
 THE LAST GUESS

Both vul., South holds ♠ 93 r AK10762 q Q103 ♣ A7

West North East South

— — — 1r
1♠ 2r Pass ?

South could conceivably pass, bid 3 r, or bid 4 r. If he passes there

is no question in his mind that it is correct for E-W to compete. If
he bids 4 r E-W will certainly do well to defend, probably doubled.
However, if South bids 3 r (non-invitational), he doesn’t really have any
idea what the correct E-W action is. In addition, South has no intention
of bidding higher. Consequently, the 3 r bid gives E-W the last guess,
and the guess won’t be easy. It is not surprising that this is consistent
with total-trick analysis.

E-W vul., South holds ♠ 7 r Q105 q AJ92 ♣ Q10976

West North East South

— 3r 3♠ ?

If South bids 4 r, he knows that it must be right for E-W to bid 4 ♠; this
contract figures to make. If South bids 6 r he can be pretty sure that
E-W should double; South has enough defense so that 6 ♠ will probably
go down. However, if South bids 5 r, he has no idea what the correct
E-W action is. 5 ♠ might make or it might not, and South really doesn’t
care. He has taken a good advance sacrifice at the highest level his side
will reach, and now E-W have to make the last guess.

N-S vul., South holds ♠ AQJ965 r 2 q A4 ♣ AQ83

West North East South

— — — 1♠
3r 4♠ 5r ?

There doesn’t figure to be a grand slam, for North would have bid
4 r rather than 4 ♠ if he had the necessary cards to make a grand. A
small slam, on the other hand, could be cold, reasonable, or hopeless
depending on what North has, and there is no way to find out. If South
meekly bids 5 ♠ this will undoubtedly end the auction, and South will

239
 MATCHPOINTS

have made the last guess. If N-S have a small slam, E-W must have a
good save. Therefore, South should bid 6 ♠. Now the last guess is thrust
back on E-W; they must guess whether or not South can make his slam.
South doesn’t know what action he is rooting for, but the problem is
back in the opponents’ court where it belongs.

E-W vul., South holds ♠ Q83 r 1094 q A5 ♣ KJ873

West North East South

1q 2♠ Double ?

South cannot tell if E-W have a game. This will depend somewhat
on whether or not E-W have a 4–4 heart fit, how the spades split, and
the lie of the club suit. A 4 ♠ call by South simply doesn’t give the
opponents any problem. They will just double for lack of anything else
to do, as they can’t very well be going to the 5-level with South holding
this many high cards, and they will probably be right. A pass by South
won’t give E-W a problem either. It is extremely unlikely that West
will pass the Negative Double, so he will just bid his hand. A 3 ♠ call,
however, is just right. South does not plan to compete to 4 ♠, so he has
no further problems. He does not know what the best action for the
opponents is, so E-W must now make the last guess.

♠ rq ♣ ♠ rq ♣

This last hand presents an interesting paradox. Suppose South bids

3 ♠ as recommended, and West does, in fact, bid 4 r. Should South
save? We have already determined that the answer is no. If South does
take the sacrifice, he will be thrusting the last guess back upon himself.
However, if 4 r makes then 4 ♠ would certainly be a good save, as it
can hardly be going down 4 tricks. So if South is not going to save, he
is betting that the opponents have done the wrong thing, and 4 r won’t
make. Well, if South thinks 4 r is going down, then is there any reason
that South shouldn’t back up his judgment with a double? The answer
in this case, which will surprise a lot of players, is no, there is no reason,
and South’s correct action is to double 4 r! This seems strange. South
wasn’t exactly bidding 3 ♠ to make, although on a good day it might,

240
 THE LAST GUESS

and the hand certainly seems to belong to E-W on high cards. Yet, it is
recommended that South double 4 r.
Let’s look at the matchpoint cost analysis. If South doubles 4 r
and is wrong he only costs 1⁄2 matchpoint to other tables that are in 4 r,
doubled or not, and costs nothing to any other result since 4 ♠ clearly
can’t go for 800. In other words, there is no possible score between
−620 and −790. Conversely, if 4 r is going down, then South gains
1⁄ matchpoint from other tables in 4 r by doubling. So far, this isn’t
2

quite enough justification for the double. South might think that 4 r
is slightly more likely to make than not and still not save for fear that
not enough tables will bid 4 r to make the sacrifice pay off even if it is
right. There is, however, a third possibility. The hand just might belong
to N-S for a spade partial. If this is the case, the double can gain a full
matchpoint if it is right against other pairs who buy the hand for 3 ♠,
making.
The full hand illustrates how this might be:

♠ A K 10 9 7 5
r 63
E-W vul.
q 872
♣ 54
♠ 62 ♠ J4
N
r AQ87 r KJ52
W E
q K J 10 6 4 q Q93
S
♣ A2 ♣ Q 10 9 6

♠ Q83
r 10 9 4
q A5
♣ KJ873

West North East South

1q 2♠ Double ∗ 3♠
4r Pass Pass Double
Pass Pass Pass

Everybody’s actions are quite reasonable. West was pushing slightly

for his 4 r call; had South passed he probably would have bid only 3 r.
It is not a bad bid, however, and it may give N-S the last guess as to

241
 MATCHPOINTS

whether or not to save. North has a good hand for South, and with a
winning club guess 3 ♠ is a make while 4 r is clearly down 1, so there
is likely to be a big swing on doubling 4 r.
Does this mean that every time we give the opponents the last guess
and they bid on we should double them? No, that would be ridiculous.
For example, on the earlier hand on which we took the premature 5 r
save we would not double if E-W bid 5 ♠, even if we felt that we had a
good save available in 6 r and we did not think that 5 ♠ was a favorite to
make, for there would be no reason to double. If 5 ♠ were going down
we would get a good board anyway, because not everybody will push
the opponents to the 5-level, while if 5 ♠ were making the double would
needlessly cost 1⁄2 matchpoint against other tables who bought the hand
for a spade contract at the 4- or 5-level. This type of double should only
be made when the opponents have outbid us by bidding a game, and
there is some chance that our contract would have made. On the hand
where the opponents bid 4 r over 3 ♠, had they bid 4 q instead a double
would not be recommended. There are several possible scores between
−130 and −710, so the double could cost a lot if wrong. Over the 4 r
bid, however, the double may gain a lot if right and lose little if wrong
once we decide not to sacrifice.

♠ rq ♣ ♠ rq ♣

The general idea in these “who knows who can make what” auctions
is to bid the limit of your hand quickly to make the opponents guess,
and if they outbid you make them pay for guessing wrong if you might
have made what you bid. This is particularly true when your last bid
is one under an enemy game contract, and they then bid the game.
In this type of situation a preemptor who has been raised is allowed
to violate his usual oath of silence and join in with a double or a bid
if his hand is exceptionally offensively or defensively oriented for his
preempt. A bid that is one level under the opponents’ potential game is
called a “one-under” bid, and allows these unusual competitive actions.
For example, on the previous hand North might well have doubled 4 r
himself after South bid 3 ♠ if the partnership style is extremely light
weak jump overcalls at favorable vulnerability, for he has extra defense
with his ♠ AK.

242
 THE LAST GUESS

Another example:

E-W vul., South holds ♠ 4 r AQJ10763 q A4 ♣ K84

West North East South

— Pass Pass 4r
4♠ Pass Pass ?

South’s 4 r bid is not a classic preempt, but it is a reasonable tactic

opposite a passed hand and certainly makes E-W guess. Once West
guesses to bid 4 ♠, however, South must assume that West has guessed
wrong and double him. There is too great a danger that 4 r will make
even though South couldn’t know if it would when he bid it, and 4 ♠
must be doubled to guard against this possibility. Note that North
couldn’t double on a scattered 9-count, for South might have a normal
4 r opening bid for all North knows. Sometimes the 4 ♠ contract will
make, but South can’t just sit there and pass, nor does he have any
reason to believe that 4 ♠ will make and take a 5 r save. If you are
going to adopt tactics such as opening 4 r in third seat on this type of
hand (winning tactics, in my opinion), then you must be prepared to
assume that the opponents have guessed wrong if they bid and double
them. The general principle is: Make them guess, and then make them
pay!

♠ rq ♣ ♠ rq ♣

It is not always necessary to bid immediately to the highest level you

are willing to reach. Sometimes you may be able to buy the contract
one level lower, even though you will go higher if necessary. Keep
in mind that this approach gives the opponents a chance to exchange
information that may help them when they have to make the final guess.
Also, make sure that you have already decided it is correct to compete
further if you take the slow approach. If you are not sure, it is best to
make your decision now one way or the other and then stick with it;
otherwise, you will be the one who is making the last guess.

243
 MATCHPOINTS

E-W vul., South holds ♠ KQ10632 r 84 q AJ7 ♣ K7

West North East South

— — — 1♠
2r 2♠ Pass ?

A 3 ♠ call by South certainly couldn’t be criticized, and if E-W go to

the 3-level South will certainly compete to 3 ♠ since he has six spades.
However, a good case can be made for passing. South knows that it
is correct for West to bid on, but West may not know that. There is
a reasonable chance that West will be afraid to compete to the 3-level
vulnerable, and N-S will then buy the hand for 2 ♠. In my opinion, this
possibility outweighs the danger of allowing E-W to exchange further
information.

E-W vul., South holds ♠ A765 r 86 q 753 ♣ Q654

West North East South

— 3♠ Double ?

E-W probably have a slam somewhere, and South is fully prepared to

take a save against it. But there is no rush in taking that save, for the
slam hasn’t been bid yet. A simple 4 ♠ bid by South may be sufficient
to deprive E-W of the bidding room they need to assess their combined
values, and they may make the mistake of doubling 4 ♠ or of stopping
at the 5-level. If E-W bid their slam, South will, of course, take the
save. This is not inconsistent bidding, for South has already made up
his mind as to what he will do. Consequently South will not be making
the last guess if E-W bid a slam, for South already knows what his right
action is. The only danger in this approach is that E-W may exchange
enough information to bid a grand, but this is unlikely and the possible
gain from bidding only 4 ♠ is well worth the risk.
Suppose, on the same auction, South had held ♠ K976 r 9 q A9874
♣ 864. Now it is not at all clear whether or not E-W have a slam, for
N-S might have two aces or North may have a diamond ruff coming.
South will certainly want to take a 5 ♠ save over 5 r, which E-W will
undoubtedly bid if South bids only 4 ♠. Consequently, South should
leap to 5 ♠ immediately and let E-W guess whether or not they have a
slam. Obviously if E-W then bid the slam South should not save; he

244
 THE LAST GUESS

must assume that his opponents have guessed wrong rather than risking
making the wrong last guess himself.

None vul., South holds ♠ Q963 r 32 q A75 ♣ QJ63

West North East South

— 2♠ 3r ?

This is not the sort of hand with which to fool around. If South bids
3 ♠ the opponents almost certainly will bid 4 r, right or wrong. South
simply must decide right now whether 4 ♠ will be a good save against
4 r. If he thinks so he should bid it now to make West guess if he
has a possible 5 r call, while if South doesn’t think 4 ♠ will be a good
save (either because 4 r might go down or because 4 ♠ will go for too
much) he should merely bid 3 ♠ and then bow out of the auction. Both
actions are reasonable. The one thing South must not do is bid 3 ♠ (or
pass) now and then take a 4 ♠ sacrifice when E-W bid 4 r. This allows
E-W to exchange maximum information and puts the last guess back
on South. Since South isn’t sure what the right guess is over 4 r, he
should make his decision early in the auction and stick with it. In this
way, he has put the last guess back in the hands of the opponents, where
it belongs. The time to wait and see if the opponents get to their best
contract before saving is when there is a good chance that they won’t.
On this hand there is little practical chance that the opponents will fail
to bid 4 r, so South should make an immediate decision and act on it.

245
 HIGH-LEVEL DECISIONS

Both partnerships have good distributional fits, and the bidding soon
soars up to the 4- or 5-level, where either or both pairs might or might
not be able to make their bids. At some point, you have to make the
crucial decision—to double, pass, or bid on. Many matchpoints can
ride on this decision. While there is no substitute for good judgment, it
is possible to load the dice so that you gain the most when your decision
is right and lose the least when it is wrong.
Suppose your side has a spade fit, the opponents have a diamond
fit, and they have competed to 5 q over your 4 ♠ bid. Unless one side
is taking a sacrifice that is down more than the value of game, it is
clearly correct to bid 5 ♠ if either 5 q or 5 ♠ makes, but to double if
neither does. At IMPs or rubber bridge the odds always favor bidding
on if the decision is close. If you bid when both contracts are down 1
you go from +100 to −100 (assuming neither vulnerable and that the
final contract will be doubled), for a 5-IMP loss. If you double when
they make and you are down 1 you get −550 instead of −100, costing
10 IMPs, and if it is your contract that is making you get +100 instead
of +650 for an 11-IMP swing. If both contracts happen to make, the
swing may get as high as 14 or 15 IMPs. Consequently the IMP odds
always favor bidding on if it is close, simply because you lose more by
defending and being wrong than by bidding and being wrong.
At matchpoints, the amount by which you are wrong is not
significant. Being −100 instead of +100 can be just as costly as being
−550 instead of −100. The object is simply to be right as often as
possible. Unfortunately total-trick analysis, so accurate at lower levels,
tends to be of little use at the 4-level or higher. Long trump holdings
matter, of course, but factors such as distribution and two-suit fits come
into play much more at the higher levels than in partscore competition.
If the decision is really close and you really have no idea who can
make what, it is usually better to bid on. By buying the contract you
give yourself two ways to win—either you might make or they might
make. If you defend, you are right only when both contracts are going
 MATCHPOINTS

down. Bidding on actually has a third way to win: even if both contracts
are going down, there is the possibility that the opponents will misjudge
the situation and bid still more. For example:

♠ 2
North dealer r 862
Both vul. q K 10 6 5 4
♣ Q J 10 9
♠ 9754 ♠ A Q J 10 8 3
N
r KQ3 r 975
W E
q 82 q —
S
♣ 7654 ♣ AK83

♠ K6
r A J 10 4
q AQJ973
♣ 2

West North East South

— Pass 1♠ 2q
2♠ 4q 1 4♠ 5q
Pass Pass 5♠ Double
Pass Pass Pass
1. Preemptive

This is a typical sequence on a distributional fitting hand. When East

bids 4 ♠ South doesn’t know if 4 ♠ will make, and he also doesn’t know
if 5 q will make. His 5 q bid is a good two-way shot. It might either
make or be a good save. On the actual hand both 4 ♠ and 5 q happen to
go down 1 trick, largely due to the lack of fit in the heart and club suits,
which can’t be diagnosed during the bidding. However, the auction isn’t
over yet. West has nothing more to say, and when it gets passed back
around to East he is in the same position that South was in when he bid
5 q . From East’s point of view 5 ♠ could well make, or 5 q might be
making, in which case 5 ♠ will certainly be a good save. His 5 ♠ call
turns out very badly, but it is not a terrible bid and certainly could have
worked. Finally, South knows what to do. He can’t be 100% sure of
beating 5 ♠, but it is still correct to double since 5 ♠ figures to go down
and 5 q might make. The interesting point about the hand is that South

248
 HIGH-LEVEL DECISIONS

was “wrong” in bidding 5 q , since both 4 ♠ and 5 q are down 1, yet the
bid worked out beautifully because the opponents made the mistake of
bidding one more. This is another example of both giving the opponents
the last guess and having an extra way to win. East guessed, and he
guessed wrong.
One can’t always buy the contract. The bidding has to stop
someplace. The point is that in a high-level competitive bidding
decision, if either the contract the opponents have just bid or the one
you are thinking about bidding might make, it is usually correct to bid
on, for you give yourself several ways to win. Once it is clear that you
can’t make your contract if you bid further, it is time to defend. On the
previous hand South thought he might make 5 q on a good day, but 6 q
is certainly out of the question. Consequently the 5 q bid is correct, but
bidding 6 q , while it could be right if E-W can make 5 ♠, is not a wise
action. It is seldom right in a “who knows who can make what” auction
to deliberately settle for a minus score. East’s 5 ♠ bid is justifiable only
on the grounds that it might make; if it could have no play it would
probably be better to defend. Once you have taken what is definitely a
sacrifice, you have committed yourself to the last guess.
If you choose to defend in a high-level competitive auction, it is
always correct to double rather than sell out undoubled if your previous
bid had been for a make rather than a save. For example, at neither
vulnerable you bid 4 r with the hope of making, and the opponents
now bid 4 ♠. A double, if wrong, as opposed to a pass, will only cost
1⁄ matchpoint against other tables in 4 ♠ , since there is no possible score
2

between −420 and -590 (presumably if you bid 4 r to make then others
holding your cards will not be going for −500 even at the 5-level).
However, the double gains 1⁄2 matchpoint against these same tables if
it is right. Since you have chosen to defend you must believe that 4 ♠ is
more likely to go down than not, for if you thought 4 ♠ was a favorite
to make you would have bid 5 r as a save, with the outside possibility
of making. Consequently, the odds favor the double. In addition, there
is the possibility that 4 ♠ was bid as a sacrifice against your 4 r game.
If this is the case not doubling would be very costly, particularly if
the sacrifice happens to go down more than the value of your game.
Allowing the opponents to get away with an undoubled save is very
expensive at matchpoints. This is a situation where the double stands

249
 MATCHPOINTS

to gain more when it is right than it will lose when it is wrong, and you
must believe it is a favorite to be right or you would bid on. Therefore,
doubling becomes very attractive.

None vul., South holds ♠ 2 r KQJ954 q A105 ♣ AQ9

West North East South

— — — 1r
1♠ 2r 2♠ 4r
4♠ Pass Pass ?

What should South do? When he bid 4 r it was with every intention
of making it, although there were no guarantees. Bidding 5 r in the
hope of making would be much too optimistic. Partner would need a
perfect hand and it would still probably be on a finesse. What about
the opposing 4 ♠ contract? If South feels 4 ♠ will make he should bid
5 r as a save, but I don’t think this is a good bid. There is no particular
reason to expect 4 ♠ to make, although it certainly might. If South bids
5 r he commits himself to a minus score when it might have been the
opponents who were saving, so there is no reason to take this action.
However, South must not pass simply because he doesn’t know who
can make what. It is essential to double! Since South is betting that
E-W can’t make 4 ♠ he might as well up the stakes, since there is no
likely score between −420 and −590. For all South knows, it is E-W
who are saving.

♠ rq ♣ ♠ rq ♣

The previous hand is a “double or save” situation. N-S can certainly

make at least 9 tricks in hearts, so South knows he has a good save if
E-W can make 4 ♠. Therefore, pass cannot be the best action. If the
opponents are making it is right to save, while if they are going down it
is right to double. You must do one or the other, particularly if it is at
all possible that you could have made your bid and it is the opponents
who are saving.
The exception to the “double or save” principle occurs when you
have already taken a sacrifice and the opponents have bid on. You
may be sure that a further sacrifice is still safe, but choose to defend

250
 HIGH-LEVEL DECISIONS

undoubled. The reason is that the further sacrifice might not gain much
when it is right, because some pairs may buy the hand your way at a
lower level. This situation can only occur at favorable vulnerability,
when a 3-trick set is a good save. At any other vulnerability either the
further sacrifice will be too expensive or you were only going down 1,
and so had some chance of making your bid. If the further sacrifice is
too expensive it is no longer a “double or save” situation, while if you
might have made your bid you must double, as we have seen.

E-W vul., South holds ♠ 7 r A865 q A652 ♣ QJ86

West North East South

1♠ 3q 4♠ 5q
5♠ Pass Pass ?

When South bid 5 q it was fully intended as a sacrifice—even the best

hand North can have for a weak jump overcall at favorable vulnerability
won’t produce 11 tricks. 6 q figures to be a good save versus a spade
game making at other tables, for N-S can certainly take 9 or 10 tricks.
What about 5 ♠? This is not so clear. If South feels that 5 ♠ is more
likely to go down than to make he might as well double, since it is
not likely that other N-S pairs will fail to find the save at the 5 q level.
Suppose South feels that 5 ♠ is 60% to make. Does this mean that he
should take the 6 q sacrifice? Not necessarily. It would be the correct
action if all other E-W pairs pushed to 5 ♠, but this will not happen.
Suppose at other tables half the E-W pairs double 5 q , and the other half
bid 5 ♠. If South takes the save he gains 1⁄2 matchpoint only from those
tables in 5 ♠ even if he is right, since the board has been lost to those
who bought it for 5 q . On a 12 top he will gain 3 matchpoints, getting 6
instead of 3. If he takes the save and is wrong he loses 1⁄2 matchpoint to
other tables in 5 ♠, and a full matchpoint to those who bought it for 5 q
doubled. This action will cost 9 matchpoints, for he will get a 0 instead
of 9 on a 12 top. Consequently he is giving 3-to-1 matchpoint odds
by saving, which is too much to give if 5 ♠ is 60% to make. It is also
incorrect to double, for the gain or loss in doubling is 1⁄2 matchpoint
to other tables in 5 ♠, so if 5 ♠ is a favorite to make double is also
a losing action. Consequently, it is correct to pass even though pass
cannot be the “perfect” bid. Either doubling or saving is the winning

251
 MATCHPOINTS

action, but passing is the percentage bid and will give South the highest
matchpoint expectation if this assessment of the hand is correct. This is
the exception to the “double or save” principle.

252
 HIGH-LEVEL DECISIONS

THE FORCING PASS

The concept of a forcing pass is confusing to many players. A forcing

pass is just a call showing no preference for defending but insufficient
values to bid on unilaterally in a high-level competitive auction where
the logic of the situation demands that the partnership either bid one
more or double. The most common case is when the opponents have
taken a sacrifice, but forcing passes can also crop up in the “who knows
who can make what” type of auctions, or even in “double or save”
positions. The important condition is that it cannot be correct for the
partnership to sell out without doubling. The forcing pass throws the
decision to the passer’s partner, and his decision should be final. If he
doubles, the initial passer must pass. If he were planning to bid on, he
should have done so immediately, thereby telling his partner that he has
extra offensive values, in case the key decision must be made at the next
higher level.
When is a pass forcing? A lot of players have trouble deciding
this, and any good partnership should work out precise agreements as
to which passes are forcing and which are not. It is most embarrassing
and quite costly to sell out undoubled when one partner thought he was
making a forcing pass and showing extra values while his partner did
not think the pass was forcing. A good general rule is that passes are
forcing when the partnership has found a fit and has shown at least
game-invitational values. This does not necessarily mean that the hand
belongs to the partnership. It is not uncommon to see a forcing pass
lead to a good sacrifice in the “who knows who can make what” type
of auction, even though the player taking the sacrifice and his partner
might not have thought it was a sacrifice at the time.
A forcing pass invites partner to bid on, so it must show a hand with
offensive orientation. Usually this will mean shortness in the enemy
trump suit if it is possible that the passer is short. If the passer has
already indicated a balanced hand, then his pass shows nothing wasted
in the enemy trump suit and invites partner to bid on if he has the
singleton. This is not mandatory, of course. With extra strength it
might be right to make a forcing pass with a doubleton in their suit,
while with minimum hand it could be correct to double with a singleton
simply because you judge that you can’t make anything higher, but still

253
 MATCHPOINTS

think you can beat the opponents’ contract. In general, however, length
in the opponents’ trump suit should be the key to the decision.
Consider the following common type of auction with none vul.:

West North East South

— — — 1r
1♠ 3r 1 4♠ ?
1. Limit raise

By our definition, if South passes it is forcing, since the partnership

has a fit and game-invitational values. It might be N-S who are saving
if they bid 5 r, but it cannot be correct to sell out undoubled. We can
assume that North doesn’t have a singleton spade since he didn’t bid
4 r, make a Splinter bid, make a fit-showing jump, or do any of the other
things he might have done with a singleton, depending on partnership
agreements.

a. ♠ 97 r AQJ53 q KQ3 ♣ 973. South must double. There is no

guarantee that 4 ♠ will go down, but South would rather defend
4 ♠ than play 5 r opposite a normal limit raise. A pass would
invite North to bid on with no wasted spade cards, which South
doesn’t want North to do.

b. ♠ 2 r AQ9764 q Q74 ♣ A87. This is a classic forcing pass. If

North has nothing wasted in spades it will probably be correct
to go on to 5 r. For example, if North holds ♠ xxx r KJxx
q Kx ♣ QJ10x then 5 r depends on the club finesse. However, if
the club finesse is off for N-S it will be on for E-W, so 4 ♠ will
probably make. Either way, it would be right to bid 5 r, which
might be a make and might be a good save. If North instead
holds ♠ Kxx r KJxx q xx ♣ QJ10x, he will double 4 ♠. This is
certainly the right action on the combined hands, since 5 r has no
play and 4 ♠ will probably go down.

254
 HIGH-LEVEL DECISIONS

c. ♠ 2 r AKQ832 q AJ8 ♣ K65. South should simply bid 5 r,

since this is what he thinks he can make. He will probably lose a
spade trick and one minor-suit trick opposite a normal limit raise
such as ♠ Qx r Jxxx q KQxx ♣ Qxx. There is no reason to pass
the decision to North since South knows where he wants to play
the hand, and the penalty against 4 ♠ is not likely to be sufficient.

d. ♠ 85 r AQJ943 q A5 ♣ AQ7. This is one hand on which South

should make a forcing pass despite his doubleton spade. If North
has no wasted spade cards he probably has enough strength in the
other suits to cover all of South’s losers, so all N-S will lose will
be two spade tricks. However, if North has some spade strength it
is probably right to defend. Either South will have an extra loser,
or the penalty against 4 ♠ just might be 500 if North produces a
trump trick.

e. ♠2r KQJ10742 q 65 ♣ KQ6. This hand is an exceptional case.

Despite the fact that North has shown game-invitational values,
South can be pretty sure that the hand belongs to E-W and that
he should bid 5 r as a sacrifice. In order to distinguish this from
a strong hand, South must pass first and then pull North’s double.
This type of sequence that should normally be avoided shows
a very offensively oriented hand with little defense, and alerts
partner not to be too fast to double if the opponents bid one
more. (Many pairs today play this “pass-then-pull” sequence as
showing a stronger hand than a direct 5-level bid and inviting
partner to bid slam with the right hand. Be sure to discuss this
sequence with your partner before you trot it out.)

f. ♠ — r AQJ643 q AQJ32 ♣ 93 South certainly will not defend

4 ♠, but he is worried about what to do over 5 ♠. His best bid
is 5 q . This is not meant as a slam try, as it would be in an
uncontested auction, although slam isn’t completely out of the
question. The bid simply shows a strong offensively oriented
hand with values in hearts and diamonds. Now if E-W compete
to 5 ♠, North will be well-placed to make the final decision. If
South had a real slam try he would just go ahead and bid the
slam, letting E-W guess whether or not to sacrifice.

255
 MATCHPOINTS

Let’s look at the view from the other side of the table.

None vul., South holds ♠ 863 r QJ82 q K6 ♣ A643

West North East South

— 1r 1♠ 3r 1
3♠ 4r 4♠ ?
1. Limit raise

South should make a forcing pass. This basically denies any wastage
in spades. South’s hand is limited and he has implied a balanced hand,
so that’s all the pass shows. It is up to the partner of the limited hand
to make the final decision here. If South instead held ♠ Q83 r J862
q K6 ♣ A643 he should double 4 ♠. North can still do what he wants,
of course, but South must show his spade wastage. There is no hand on
which South can bid 5 r, because he doesn’t know whether or not his
partner was stretching for the 4 r call. If North wanted South to make
the final decision, he should have bid something other than 4 r. For
example, none vul.,

West North East South

— 1r 1♠ 3r 1
3♠ 4q 4♠ ?
1. Limit raise

4 q is not a slam try, as it would be in an uncontested auction. North is

simply showing his values, preparing for a 4 ♠ bid by the opponents.
a. ♠ Q74 r J1087 q A3 ♣ KJ97. South should double. He has
most of his points in the black suits opposite North’s presumed
shortness, which makes his hand defensively oriented.

b. ♠ 764 r QJ87 q A3 ♣ K1097. South should pass. This is a

forcing pass. South is not sure what to do. His ♣ K argues for
defending, but the rest of his hand is offensively oriented. It will
be best if North makes the final decision.

c. ♠ 764 r KJ87 q AQ3 ♣ 1097. South should bid 5 r. This is

it—the big double fit. If North has interest in competing to the
5-level, as indicated by his 4 q call, then South has the perfect
256
 HIGH-LEVEL DECISIONS

hand for him so he should go straight to 5 r. If North knew that

he did not want to go to the 5-level opposite any limit raise he
would have bid 4 r rather than 4 q , and now South could only
pass or double.

257
COMPETITIVE BIDDING
REVIEW PROBLEMS

Problem 1

None vul., South holds ♠ 105 r AQ98 q 1064 ♣ AQJ3

West North East South

— — — 1♣
Pass 1r 2♠ ?
 MATCHPOINTS

Problem 1 Solution

None vul., South holds ♠ 105 r AQ98 q 1064 ♣ AQJ3

West North East South

— — — 1♣
Pass 1r 2♠ ?

3 r. It is imperative for South to show his heart support, even if it

involves a slight overbid. N-S have at least eight hearts, maybe nine,
and E-W probably have at least eight spades, so the Law of Total Tricks
argues for bidding. North should give South some leeway in this type
of auction. If South were worth a jump to 3 r, he would take the bull by
the horns and bid game himself.

260
COMPETITIVE BIDDING REVIEW PROBLEMS

Problem 2

Both vul., South holds ♠ AQ10974 r 985 q 95 ♣ 43

West North East South

1q 1r 2q ?

261
 MATCHPOINTS

Problem 2 Solution

Both vul., South holds ♠ AQ10974 r 985 q 95 ♣ 43

West North East South

1q 1r 2q ?

2 r. It would be nice to be able to introduce the spade suit, but it

simply can’t be done conveniently. If South bids 2 ♠ and the opponents
compete to 3 q , South won’t know whether or not to bid 3 r, since
he won’t know how many hearts North has. If South bids 2 r, North
can make the proper trick-total evaluation on the heart suit, which is
probably the most important basis for the competitive decision.

262
COMPETITIVE BIDDING REVIEW PROBLEMS

Problem 3

Both vul., South holds ♠ K9832 r 53 q K94 ♣ K94

West North East South

— 1♣ 1q 1♠
2q 2♠ 3q ?

263
 MATCHPOINTS

Problem 3 Solution

Both vul., South holds ♠ K9832 r 53 q K94 ♣ K94

West North East South

— 1♣ 1q 1♠
2q 2♠ 3q ?

Double. This is a good matchpoint double. South has extra strength,

good defense, and vulnerable opponents at the 3-level. If N-S are not
playing Support Doubles, there is a good chance that North has only
three spades, for he is likely to have several hearts since E-W haven’t
bid hearts. If North has four spades and a singleton diamond he will
probably pull the double; otherwise, the double figures to work out well.

264
COMPETITIVE BIDDING REVIEW PROBLEMS

Problem 4

N-S vul., South holds ♠ A85 r Q762 q Q84 ♣ J83

West North East South

— 1r 1♠ 2r
2♠ 3r 3♠ ?

265
 MATCHPOINTS

Problem 4 Solution

N-S vul., South holds ♠ A85 r Q762 q Q84 ♣ J83

West North East South

— 1r 1♠ 2r
2♠ 3r 3♠ ?

Pass. North’s 3 r bid is not invitational; it only shows a sixth heart

or a singleton, so South should not go to 4 r without an exceptional
hand. Even though N-S may have 10 trumps, this is not justification for
a 10-trick contract, as the Law of Total Tricks tends to break down at
higher levels. South should not double, for 3 ♠ is quite likely to make
and a 1-trick set still won’t beat +140. The vulnerability argues for
conservatism, so South should sell out and hope that others holding his
cards get into worse trouble by acting over 3 ♠.

266
COMPETITIVE BIDDING REVIEW PROBLEMS

Problem 5

None vul., South holds ♠ Q108654 r 106 q AJ3 ♣ A10

West North East South

1NT Pass 2♣ ∗ ?

267
 MATCHPOINTS

Problem 5 Solution

None vul., South holds ♠ Q108654 r 106 q AJ3 ♣ A10

West North East South

1NT Pass 2♣ ∗ ?

Pass. This is a marginal overcall at best, as it could well go for a

number and South doesn’t necessarily want a spade lead. What about
the nuisance value of shutting out the opponents’ potential heart fit?
A little thought shows that this nuisance value is imaginary. If West
has four hearts he can pass and see what East does. When East fails
to double 2 ♠, West can deduce that East has four hearts, else why did
he bid Stayman in the first place. Consequently, the 2 ♠ overcall won’t
hinder the opponents in their search for a heart fit. If the overcall were
directly behind the 1NT opener it would be a different story, for now
the enemy has had less opportunity to exchange information.

268
COMPETITIVE BIDDING REVIEW PROBLEMS

Problem 6

Both vul., South holds ♠ 84 r K642 q K754 ♣ 1062

West North East South

1♣ 1r Double ∗ ?

269
 MATCHPOINTS

Problem 6 Solution

Both vul., South holds ♠ 84 r K642 q K754 ♣ 1062

West North East South

1♣ 1r Double ∗ ?

3 r. South should go right to the 3-level without delay. He knows that

his side has at least nine hearts and that the opponents probably have
eight spades, so there is no reason to hold back. A lesser bid gives E-W
too much room to exchange information.

270
COMPETITIVE BIDDING REVIEW PROBLEMS

Problem 7

E-W vul., South holds ♠ 653 r K876 q A3 ♣ AQ108

West North East South

— 2r 2♠ 4r
4♠ Pass Pass ?

271
 MATCHPOINTS

Problem 7 Solution

E-W vul., South holds ♠ 653 r K876 q A3 ♣ AQ108

West North East South

— 2r 2♠ 4r
4♠ Pass Pass ?

Double. A classic “double or save” hand. N-S certainly have a good

save at 5 r if 4 ♠ is making. South has no particular reason to think that
4 ♠ will make, so taking the save is wrong. As long as South is betting
that 4 ♠ is going down he should double, for there is certainly some
chance that 4 r is making. The double has far more to gain when it is
right than it has to lose when it is wrong, and South is willing to bet that
it is right.

272
COMPETITIVE BIDDING REVIEW PROBLEMS

Problem 8

E-W vul., South holds ♠ AQ653 r J1076 q 72 ♣ 107

West North East South

1r Double Rdbl ?

273
 MATCHPOINTS

Problem 8 Solution

E-W vul., South holds ♠ AQ653 r J1076 q 72 ♣ 107

West North East South

1r Double Rdbl ?

3 ♠. South can assume a 9-card spade fit, and the opponents certainly
won’t sell to 2 ♠ after East’s redouble. Consequently, South should bid
the limit of his hand immediately. This is likely to produce a bonus if
East has a heart fit and tries 4 r, for South can now spring his trap and
double. Any lesser bid allows East to describe his hand at the 3-level,
and gives the opponents a chance to get off the hook.

274
COMPETITIVE BIDDING REVIEW PROBLEMS

Problem 9

Both vul., South holds ♠ KQ93 r 92 q 76532 ♣ A8

West North East South

1q Pass 1r Pass
2r Pass Pass ?

275
 MATCHPOINTS

Problem 9 Solution

Both vul., South holds ♠ KQ93 r 92 q 76532 ♣ A8

West North East South

1q Pass 1r Pass
2r Pass Pass ?

Pass. This is a close decision. The danger in bidding 2 ♠ is that South

has no place to run, and the opponents know enough about the hand to
penalize South when he is wrong. North’s silence and the vulnerability
argue in favor of passing.

276
COMPETITIVE BIDDING REVIEW PROBLEMS

Problem 10

N-S vul., South holds ♠ KJ75 r AJ10 q 9642 ♣ K7

West North East South

— — 1♣ Double
1r 2q 2r ?

277
 MATCHPOINTS

Problem 10 Solution

N-S vul., South holds ♠ KJ75 r AJ10 q 9642 ♣ K7

West North East South

— — 1♣ Double
1r 2q 2r ?

3 q . Never mind the minimal values and the strong major-suit holdings.
South has four diamonds, so he must compete to the 3-level opposite
his partner’s presumed 5-card suit. This bid does not show any extra
values—simply possession of a fourth diamond.

278
 Part IV

Defensive Bidding
 ENTERING THE AUCTION

It is very important to get into the auction if at all possible after

the opponents have opened the bidding. Players who have stiff
requirements for overcalls and takeout doubles are hurting their chances
in matchpoint competition. Look at it from the opening bidder’s point
of view: After you open the bidding, wouldn’t you prefer that the
opponents stay silent so you and your partner can exchange information
as you please? It is much more difficult to arrive at the best contract if
the enemy starts throwing bids in your face, gobbling up bidding room
and making the exchange of information more difficult. What is bad
for the opener’s side must be good for the other pair, and vice versa.
Consequently, it will generally be better to get into the auction than
pass quietly.
The main ways to get into the auction after a one-of-a-suit opening
bid are the simple overcall, the takeout double, and the 1NT overcall.
Preempts are also important, and will be discussed in a later chapter.
 MATCHPOINTS

OVERCALLS

The most important competing bid is the overcall. There are four
main reasons one might want to overcall: constructive bidding, lead-
direction, space consumption, and sacrificing.

1. For constructive purposes. This is the most common reason, and

certainly the most important. The opponents’ opening bid does
not give them a claim to the contract; it could well belong to you.
Consequently, an overcall is often simply an attempt to get to the
right contract. For example:

Both vul., South holds ♠ J8532 r A74 q K72 ♣ AQ

West North East South

— — 1r ?

South should overcall 1 ♠ despite the weak suit. If East had passed
South certainly would have opened 1 ♠, so why not bid it over the
opening bid? It will be noted that this overcall doesn’t satisfy any of
the other reasons for overcalling, but this doesn’t matter. South’s hand
is strong enough to compete and he holds the master suit, so now is the
time to act. Some players feel that they can pass now and balance later
if it is correct to do so. This is a losing tactic. In the first place the
1 ♠ overcall is very unlikely to be penalized, while a later balance at a
higher level after the opponents have exchanged information is much
more dangerous. Secondly, you lose the opportunity of having partner
raise spades and taking up enemy bidding room. Thirdly, if you balance
later you can’t be sure of finding a fit, while if you bid now partner will
raise only with appropriate trump support.
This type of overcall should not be overdone. With a weaker hand
such as ♠ Jxxxx r Axx q Kxx ♣ Qx South should pass. If North can’t
get into the auction there probably is no constructive future for the hand.
Also, if South holds a hand with a lower-ranking suit such as ♠ AQ
r Axx q Kxx ♣ Jxxxx a pass is advisable. A 2 ♣ overcall is much
more likely to be penalized than a 1 ♠ overcall, and since hearts outranks
clubs the chances of winning the partscore battle are small, while with
the spade suit there is a good chance of outbidding the opponents. These

282
 ENTERING THE AUCTION

two factors are sufficient to make the 2 ♣ overcall unattractive, while the
1 ♠ overcall on the same type of hand is correct.

2. Space consumption. This concept has already been discussed

under nuisance bids. It is a prime consideration when deciding
whether or not to make a marginal overcall. For example:

Both vul., South holds ♠ A95 r Q9873 q 92 ♣ KJ3

West North East South

— Pass 1q ?

A 1 r overcall would be pointless. The danger isn’t so much that it

may go for a number, but that partner may later over-compete or make
a disastrous heart opening lead. However, change a few suits to:

Both vul., South holds ♠ Q9873 r 92 q KJ3 ♣ A95

West North East South

— Pass 1♣ ?

Now, South should overcall 1 ♠. The disadvantages of the overcall

are still the same, but the gain in space consumption outweighs them.
It is true that most opponents have things such as Negative Doubles
available to help recover lost heart suits, but their bidding still won’t be
quite as accurate. This is particularly true if partner can muster up a
2 ♠ raise. Now the opponents will be floundering at the critical 3-level
before they have any idea what they have, and they are far more likely to
go wrong than if they had been left alone. This overcalling policy reaps
hidden matchpoints that super-sound overcallers don’t realize. They see
the occasional disaster, but they don’t understand that opponents will
bid less accurately when faced with space-consuming bids. The “lucky”
pair that gets many gifts from their opponents usually has taken actions
such as these to make life difficult for the opponents and so deserves
the gifts.

3. Lead-direction. The premium on overtricks at matchpoints

makes lead-direction overcalls far more important than at IMPs.
The overtrick saved by the right lead can be worth several

283
 MATCHPOINTS

matchpoints. Consequently, it is often worth sticking your neck

out to get a lead-director in. For example:

Both vul., South holds ♠ A7 r 872 q KQ10765 ♣ 75

West North East South

1♠ Pass 2♣ ?

The 2 q overcall has little value other than as a lead-director, and the bid
could certainly go for a number. Despite this, I still think the overcall is
a winning action. If West becomes declarer, as is likely, a diamond lead
is the only lead South’s hand can stand, and the occasional catastrophe
is more than compensated for by the times the opponents simply bid
to their normal contract and the diamond lead saves a trick. However,
with ♠ 72 r A87 q KQ10765 ♣ 75 South should pass. Now South
doesn’t particularly mind a heart lead; in fact, it might turn out better.
If the opponents’ black suits are solid and partner has something like
r KJxx, it is the heart lead, and not the diamond, that will cut down the
overtricks in a spade contract. On the first hand, the diamond lead may
be necessary both to set up a diamond trick and avoid blowing a heart
trick, so the gain may be substantial. On the hand with the r A the risk
of the overcall is not worthwhile, for it might not gain anything even if
it is not penalized.

4. Sacrificing. A fourth reason for overcalling is to set up a possible

sacrifice. For example:

E-W vul., South holds ♠ Q106432 r 9 q KJ6 ♣ 542

West North East South

1♣ Pass 2r 1 ?
1. Game-forcing.

South should stick his nose in with a 2 ♠ overcall, despite the risks. The
opponents have already announced game-going strength, so 2 ♠ would
have to go for 800 to be too expensive. This isn’t very likely, and even if
the penalty is available the opponents may not be in a position to collect
it. If North has a spade fit there is a good possibility of a successful 4 ♠
save over 4 r (or even a 6 ♠ save over 6 r), so the possible gain justifies
284
 ENTERING THE AUCTION

the risk. At any other vulnerability the 2 ♠ overcall would be downright

reckless. The likelihood of a good save would be greatly diminished,
and the danger of going for too much at 2 ♠ and getting caught would
increase dramatically.

♠ rq ♣ ♠ rq ♣

It should be noted that many overcalls satisfy more than one of these
reasons. For example:

E-W vul., South holds ♠ KQ10643 r 105 q K2 ♣ A73

West North East South

— — 1♣ 1♠

The obvious 1 ♠ overcall might work for any of the four reasons,
depending on what happens later in the auction. The important point
is that you should have at least one of these for any overcall, otherwise
there is no purpose in the bid. For example:

E-W vul., South holds ♠ A75 r K8 q Q8763 ♣ 1092

West North East South

— — 1♣ ?

There is no reason to bid 1 q just to hear yourself speak. Admittedly the

bid is not likely to come to harm, but it can’t gain anything. If partner
can’t bid on his own the hand belongs to E-W, no space is consumed,
South doesn’t particularly want diamond lead, and South’s hand is not
oriented toward a sacrifice. Overcalling on hands such as these may
cause partner to misjudge a competitive auction later on. However, any
change in the hand may make the overcall fit one of the reasons. Make
the hand stronger, such as ♠ A75 r K8 q Q8763 ♣ A92, and now the
overcall is correct for constructive value. Make your suit the higher-
ranking suit, as in ♠ Q8763 r K8 q A75 ♣ 1092, and the overcall
becomes right for space-consuming purposes. Strengthen the suit,
such as ♠ 875 r K8 q AQ763 ♣ 1092, and the overcall is worthwhile
for lead-directional value. Give South better distribution, like ♠ A7

285
 MATCHPOINTS

r K865 q Q87632 ♣ 9, and now the overcall is reasonable due to the

sacrificial possibilities.

♠ rq ♣ ♠ rq ♣

The most important thing for the partner of the overcaller to do if he has
trump support is to raise immediately to whatever level he thinks best
and then stay out of the auction. This is particularly true if the partner
of the opening bidder acts. The proper level is usually determined by
the Law of Total Tricks. Preemptive jump raises of overcalls are a must
for any pair wishing to put maximum pressure on the opponents. A
convention such as the Rosenkranz Redouble (redouble of a Negative
Double is automatic with a top honor in the suit) is simply ineffective—
it gives the opponents too much room. Only when the partner of the
overcaller is strong enough to invite a game should he go slowly and
start with a cuebid.

None vul., in each case the auction has gone:

West North East South

1♣ 1♠ Double ∗ ?

a. ♠ Q53 r 83 q Q10763 ♣ 652. Bid 2 ♠. So what if it is under

strength. Trick-total says to compete to 2 ♠, and this bid won’t
make life any easier for the opponents.

b. ♠ Q53 r 83 q K10763 ♣ A52. Still just a 2 ♠ bid. The hand isn’t

quite strong enough for a cuebid. We may have a tough decision
later at the 3-level, but that is for later. Now, we let the opponents
guess first.

c. ♠Q653 r 83 q Q10763 ♣ 52. Bid 3 ♠. A perfect trick-total bid.

This will make life very difficult for E-W.

d. ♠ Q653 r A3 q K10763 ♣ 52. Still bid 3 ♠ . Again put the

pressure on the opponents, making them guess. This time if they
guess to bid at the 4-level we will bet that they guessed wrong
and double to make them pay.

286
 ENTERING THE AUCTION

e. ♠ Q653 r A3 q AJ763 ♣ 52. This hand should start slowly with

a cuebid. The hand figures to belong to N-S for at least 3 ♠, and a
game is quite possible, so it is wrong to preempt. South wants to
discover if North has any extra strength, and he is not too worried
about the opponents. If the overcalled suit had been hearts and
South had held ♠ A3 r Q653 q AJ763 ♣ 52 it would not be
unreasonable to go right to 4 r as a two-way shot: it might make,
or the opponents might have an accident competing to 4 ♠ over
the 4 r bid. Since N-S hold the spade suit, this kind of preemption
is not necessary.

f. ♠ Q6532 r 3 q K10763 ♣ 52. Simply bid 4 ♠ and let the

opponents guess. It might make, it might be a good save, and
E-W might now have an accident by being forced to guess at this
high level.

287
 MATCHPOINTS

THE TAKEOUT DOUBLE

The idea behind the takeout double is entirely different from the
overcall. When you overcall you are suggesting that your suit is likely
to be the best trump suit. When you make a takeout double, you are
suggesting that partner pick the best trump suit. Since he may pick any
of the other three suits, you must be prepared to handle a bid in any of
them. This means that you should usually have at least 3-card support
for each unbid suit in order to make a takeout double. Keep in mind
that the opponents are likely to preempt vigorously against a takeout
double if they have a fit, since your side still hasn’t located a trump
suit. Partner will strain to bid his 5-card suits at the 2- or even 3-level
in competition, particularly a suit that outranks the enemy suit. If you
don’t have support for any suit he might bid, it could be disastrous.

None vul., South holds ♠ K6 r AQ108 q 1075 ♣ KQ62

West North East South

— — 1q ?

South must either pass or overcall 1 r if it suits his fancy, but a double
is out of the question. North is just too likely to bid spades at any level
in competition, and this will probably turn out badly. If the opening bid
had been 1 ♠, however, South would have a fine takeout double. Change
South’s hand to ♠ KQ62 r AQ108 q 1075 ♣ K6 and now the takeout
double of a 1 q opening bid becomes a reasonable gamble despite the
doubleton club. Partner will not strain as hard to bid a lower-ranking
suit in a competitive auction, since it means going one level higher, so if
he bids clubs over a diamond raise he will certainly have five and maybe
six. The double could backfire, but with both majors and opening-bid
strength it is worth getting into the auction.

♠ rq ♣ ♠ rq ♣

The question of whether to double or overcall has always plagued

players. Strength is not an important factor in making this decision. As
we have seen, in competitive auctions it is more important to find the
right trump suit and determine the extent of fit than it is to determine

288
 ENTERING THE AUCTION

strength. The Law of Total Tricks will see us through the 3-level if we
can find our best trump suit.

None vul., South holds ♠ J762 r AQJ32 q 8 ♣ K85

West North East South

— — 1q ?

Despite the strong heart suit and the minimal strength of South’s hand,
I prefer a takeout double to a 1 r overcall. The hand is quite suitable
as a dummy for either spades or clubs, so South should let his partner
pick the suit. The fourth spade is the key to this decision. If South
overcalls 1 r any 4–4 spade fit will almost certainly be lost forever, for
North will need five spades to voluntarily bid the suit, particularly in
competition. Switch the black suits so that South has ♠ K85 r AQJ32
q 8 ♣ J762 and now a 1 r overcall is preferable. The main danger in
doubling is that North might have four spades and three hearts, and then
the partnership will probably wind up in the 4–3 fit rather than the 5–3
fit. As a general rule, when the opponents open one of a minor and you
have a 5-card major, it is right to double if you have four cards in the
other major but better to overcall with fewer than four.

N-S vul., South holds ♠ KQ4 r AQ4 q 3 ♣ AQ6543

West North East South

— — 1q ?

A simple 2 ♣ overcall is best. A double could land N-S in an awkward

4–3 major-suit fit when the hand belongs in clubs. The plan is to
overcall first and then reopen with a double if the opponents bid more
diamonds. This is much more efficient than doubling first and then
bidding clubs. Which auction would you rather have:

289
 MATCHPOINTS

a. West North East South

— — 1q 2♣
3q Pass Pass ?

b. West North East South

— — 1q Double
3q Pass Pass ?

In the first auction South can comfortably reopen with a double and feel
that he has accurately described his hand. All contracts are still open,
including defending 3 q doubled. In the second auction, however, South
is in a bind. If he doubles again North will bid 3 ♠ on something like
♠ Jxxx r xx q Qxx ♣ Jxxx and the hand will not play very well, while
4 ♣ is comfortable. On the other hand, if South bids 4 ♣ he may find
his partner with ♠ J10xxx r xxx q Qxx ♣ Jx. Now the hand belongs
in spades, but 3 ♠ is probably the limit of the hand. By overcalling first
and then doubling South can have his cake and eat it too—he will get
to the right contract opposite both of these hands.

♠ rq ♣ ♠ rq ♣

There are two types of hands on which it is correct to make a takeout

double without support for all unbid suits. One is a hand so strong that
you fear you may miss a game if partner cannot muster up a response
to your overcall. For example:

N-S vul., South holds ♠ AKJ1094 r A8 q K9 ♣ AJ8

West North East South

— — 1q ?

South should double. This is just a bit too strong to risk having a 1 ♠
overcall get passed out. South can bid spades over whatever North bids.
If the hand were slightly weaker, say no q K, then a 1 ♠ overcall would
be preferable.
The other type of hand is one in which you have two suits and can
convert to your longest suit at the same level if partner bids the suit you
don’t have. For example:
290
 ENTERING THE AUCTION

None vul., South holds ♠ A876 r 3 q AQJ865 ♣ Q7

West North East South

— — 1r ?

A 2 q overcall risks losing a spade fit, while a double risks partner

bidding clubs. This is all right on this hand, for you can convert clubs to
diamonds at any level he bids, and you will have described a hand like
this. This is called “Equal-Level Conversion,” and is the one sequence
where a double followed by a new suit doesn’t necessarily show a
powerhouse. Note that the order of suits is very important. If you had
instead held ♠ A876 r 3 q Q7 ♣ AQJ865 you would have to overcall
with 2 ♣ and try to get back to spades later if it is right. You couldn’t
risk a double, for you would have to bid clubs at the next higher level if
partner bid diamonds, and the hand is not nearly strong enough for this
action.

291
 MATCHPOINTS

THE 1NT OVERCALL

The 1NT overcall covers many hands that don’t qualify for an overcall
because there is no good suit and aren’t suitable for a takeout double
because of distribution, yet are strong enough to act. All the 1NT
overcall shows is something resembling an opening 1NT bid with a
stopper in the enemy suit. Many players choose to pass with these
hands. Either they think they are being clever by “trapping,” or they
are afraid of going for a number in 1NT. I think they are making a big
mistake. If you pass on this sort of hand you are going to have to guess
what to do later in the auction at a higher level, after the opponents
have described their hands to each other. However if you make your
1NT overcall you have told your story. You can then sit back and relax,
watching the opponents guess while your partner knows exactly what
you have. For example:

Both vul., South holds ♠ K4 r A765 q K94 ♣ AJ73

West North East South

— — 1r Pass
1♠ Pass 2♠ ?

This is the type of bind that players who pass on this hand out of fear
or cleverness find themselves getting into. What should South do now?
Any action is extremely dangerous, but a pass is equally dangerous, for
the opponents may have stolen you blind. It is far better to get the 1NT
overcall in on the first round of bidding and avoid a guess like this later.
True, you would rather have some spot cards in hearts, a little more
strength, or a source of tricks, but you aren’t always dealt the perfect
hand for any bid.

♠ rq ♣ ♠ rq ♣

292
 ENTERING THE AUCTION

Let’s look at a few auction-entry decisions:

None vul., South holds ♠ Q73 r AQ63 q Q109 ♣ K86

West North East South

— — 1♣ ?

I recommend a takeout double. The hand is unfortunately too weak for

a 1NT overcall, which would otherwise be ideal. If you pass you will
just have the same problem at a higher level later in the auction. A 1 r
overcall would be favored by many, but it is a losing proposition. First
of all, the hand may play well in diamonds or spades if partner owns
either of these suits, which is the message conveyed by the double but
not the overcall. Secondly, you should avoid overcalling on 4-card suits
if there is a reasonable alternative. The reason isn’t that it is dangerous
at the 1-level, for you seldom go for a number there. The real danger is
that partner will play you for a 5-card suit and overbid the trick-total if
he has support.

None vul., South holds ♠ AQ8 r J8763 q AJ9 ♣ 86

West North East South

— — 1♣ ?

In the long run a 1 r overcall will work out best, even on this ragged
suit. If partner has heart support it will be as good a trump suit as any,
and if not he can bid his own suit if he has anything to say. A takeout
double risks landing in the wrong suit, and passing isn’t the road to
matchpoints.

E-W vul., South holds ♠ A86 r A7 q KQ1098 ♣ Q87

West North East South

— — 1q ?

Don’t get clever with a Trap Pass. Get your 1NT overcall in there now.
The opponents aren’t likely to stay in diamonds, so by passing you give
them freedom to find out where they really do belong and at the same
time make it harder for partner to enter the auction if he has a long suit.

293
 MATCHPOINTS

None vul., South holds ♠ AKJ9 r 84 q A864 ♣ 1053

West North East South

— — 1q ?

This is the perfect hand for the 4-card overcall. You want a spade lead,
the bid consumes space, and the overcall has constructive value. Also
notice that this is exactly the type of hand that may play well in a 4–3
fit—good trumps and no intermediates in the side suit. In fact, you
would probably choose to open the bidding 1 ♠ on this hand as dealer if
system permitted, for the same reasons.

None vul., South holds ♠ A865 r 102 q K86 ♣ AQ94

West North East South

— — 1q ?

Sometimes we just have to pass. This isn’t meant as a Trap. South

would like to take action, but nothing is at all suitable. The doubleton
heart rules out a takeout double, the spade suit is too weak to overcall on
a 4-bagger, and the hand is not strong enough for a 1NT overcall. South
doesn’t like it, as he may be faced with a difficult guess next round, but
there simply is no good bid available.

294
 ENTERING THE AUCTION

OVERCALLING ARTIFICIAL BIDS

When the opponents open with a strong, artificial bid such as 2 ♣ or a

Precision 1 ♣, the gain from getting into the bidding is much greater.
Anything that can disrupt the auction will be beneficial, since you are
taking the opponents out of carefully prepared sequences and forcing
them into natural auctions without the benefit of having named their
long suit. In addition it is very difficult for them to double you, since
they don’t know how well their hands fit. Constructive bidding is not
too important since it probably isn’t your hand, but lead-direction and
especially space consumption have a very great value.

None vul., South holds ♠ KQ102 r 63 q 1076 ♣ J752

West North East South

— — 1♣ 1 ?
1. 16+ artificial

A 1 ♠ overcall is automatic, although it would be quite reckless over a

natural club. The bid has both space-consumption and lead-directional
value, and even if partner overbids the trick-total the opponents may not
know it. Partner should go a little easy on his jump raises, for he knows
that you may have a hand like this.

E-W vul., South holds ♠ 92 r QJ1076 q K763 ♣ 63

West North East South

— — 1♣ 1 ?
1. 16+ artificial

I recommend a preemptive jump overcall of 2 r. You are unlikely to get

caught even if you are going for your life, and this bid will do far more
damage to the enemy than a mere 1 r overcall.

295
 MATCHPOINTS

Both vul., South holds ♠ KJ8 r A5 q Q9874 ♣ A87

West North East South

— — 1♣ 1 ?
1. 16+ artificial

Over a natural 1 ♣ opening South should overcall 1 q for constructive

purposes. Over the strong and artificial bid the 1 q overcall serves little
purpose since it consumes no space. South does not particularly want
a diamond lead, and the hand is unlikely to belong to N-S, so South
should pass.

296
 PREEMPTS

The preempt is one of the most valuable tools of the winning matchpoint
player who realizes the importance of making life difficult for the
opponents. Any pair playing in a tournament can get to a reasonable
contract most of the time if left on their own. A preempt creates
problems for the best of experts. It is quite difficult to arrive confidently
at the best contract when you have to start searching at the 3-level.
Players who say that preempts don’t interfere with their bidding
accuracy are just kidding themselves.
Suppose you have opened 3 r. There are only three auctions that
you don’t want to hear. They are as follows:

1. West North East South

— — — 3r
Double Pass Pass Pass

This is dangerous, but it might not be too bad. East may have had
no satisfactory bid and gambled the pass as the least of evils, so
your contract might well make. At least if the hearts are stacked
they are in front of you.

2. West North East South

— — — 3r
Pass Pass Pass

This is actually worse than the first sequence. It is possible that

you have stolen the hand from the opponents, but it is more likely
that they didn’t bid because they didn’t have anything. If this is
the case you can only hope that 3 r is your best contract. If it isn’t
there is no way to recover. In addition, if the hand is a misfit, as it
so often is when a preempt gets dropped, you will probably wind
up with the worst of it.
 MATCHPOINTS

3. West North East South

— — — 3r
Pass Pass Double Pass
Pass Pass

This is the really bad one. If the heart stack is behind you, you
will go for a number. That is one of the risks you take when you
preempt.

Any other auction figures to be favorable for your side. If partner bids
something he does so knowing what you have, so whether he is bidding
for a make or a sacrifice you are happy. If the opponents go after their
own contract you are also pleased. Not only have you taken away their
bidding room making it less likely that they will find the right spot, but
you have described your hand to your partner so you can relax while
he takes the proper action, which might be anything from punishing the
opponents for stepping in at the wrong time to taking a sacrifice against
a slam.
In order to determine whether or not to preempt on a given hand, we
must see if we have a reasonable chance of landing on our feet if one of
the auctions we don’t want to hear occurs. There are several factors that
determine the likely outcome of a preempt, and they must all be taken
into account when deciding whether or not to make a marginal preempt.
First of all, let’s look at the model preempt: None vul., South
deals and holds ♠ 3 r KQJ10843 q 96 ♣ 982. This is the perfect 3 r
opening bid. South has a very good heart suit, no defense at all, and
no possibility of playing anywhere else. Even on the auctions South
doesn’t want to hear, he figures to come out reasonably well.
Naturally you can’t wait until you pick up this hand before you
preempt. If you do you are not preempting nearly often enough, and
consequently not giving your opponents enough problems. However,
every preempt should approximate this hand to some extent. We will
examine the factors that load the dice for or against a preempt.

1. Internal trump strength. This is probably the most important

factor of all. If you are left playing the contract, that queen, jack,
or ten in the trump suit will often mean the difference between
making and going down; or if you are doubled between going for

298
 PREEMPTS

that extra undertrick that costs more than the opponents’ game.
QJ109xxx is a much better suit for preemptive purposes than
AQ9xxxx. If the suit breaks terribly you could lose three trump
tricks with the second trump suit, but there will be only two trump
losers with the first suit.

2. Possession of the ace of trumps. Oddly enough, this is a liability

rather than an asset for preemptive purposes. The reason is that
the ace is a big factor on defense. If you wind up playing the hand
in your preempt, which is the dangerous situation, the success of
your preempt will often be determined by how well pairs at other
tables holding your opponents’ cards will do if they buy the hand.
That ace in your suit might mean ace and a ruff at other tables,
which is why it is a liability rather than an asset. You would rather
have KQ10xxxx than AQ10xxxx when you preempt.

3. High cards outside your suit. These are definite liabilities. Once
again, if you wind up playing the hand you want E-W contracts
to be making. That side queen-doubleton, of no value to you for
offense, just might be the card that sets E-W contracts at other
tables.

4. Possible alternative contracts. If you have support for another

suit, it is possible that your side belongs in that suit. Obviously
it will be very difficult to get there after you preempt, so this side
support is a liability. It is very unfortunate when you are dropped
in a preempt, even undoubled, and find yourself in the wrong
trump suit. Major-suit support is the most dangerous of all. A
3 q opening on ♠ x r xx q KQ109xx ♣ xxxx with neither vul.
is quite reasonable, but with ♠ xxxx r xx q KQ109xx ♣ x it is
very risky. There is too great a danger that you belong in spades.
Obviously the longer your preempt suit is, the more likely it is to
be the best trump suit.

There are other important factors that determine the success of a

preempt that have nothing to do with the hand. They are:

1. Vulnerability. This is a very important consideration. The big

danger with a preempt is that it will be set for more than the value
299
 MATCHPOINTS

of an opposing game or partscore. At unfavorable vulnerability

this is a distinct possibility since if the opponents double they
need only a 2-trick set to beat their game, so you must toe the
line very closely. At favorable vulnerability they need a 4-trick
set and may not be able to tell that it is available even if it is, so
you can be very frisky with preempts. You can be looser with
your preempts with none vul. than with both vul. It won’t matter
if the opponents have game, but if the hand is a partscore hand
you are far more likely to be wrong when you are dropped in
your preempt with both vul. than with none vul., for you might
go down −200 undoubled. Consequently, the preferred order of
vulnerability for preempting is:

a. Not vul. vs. vul.

b. None vul.
c. Both vul.
d. Vul. vs. not vul.

2. Position. This can have quite a bit of influence on the success

or failure of a preempt. If you are in first seat nobody knows
anything, so it is anyone’s guess. Second seat is a different story.
If RHO dealt and passed, there is a greater chance that the hand
belongs to your side, as your partner and LHO are equally likely
to hold the strong hand. Furthermore, the opponents are less
likely to have an accident because they have already exchanged
some information—RHO doesn’t have an opening bid and his
partner knows this. Consequently, the gains from the preempt
diminish and the losses increase, so you should toe the line pretty
carefully. Third seat is entirely different. Now partner is a
passed hand, so if you have a preemptive type of hand you don’t
have to worry about the hand belonging to your side. LHO is
marked with the strong hand, so you can be very frisky with your
preempts since it is unlikely that you will be dropped. No longer
are you concerned if it goes all pass, for that probably means that
you have stolen the contract.

3. Exchange of information. The more information the opponents

have had a chance to exchange, the less effective your preempts

300
 PREEMPTS

are likely to be. They will be better placed to penalize you if it

is right, and more likely to get to the right contract otherwise.
This is a big factor when considering whether or not to make
a preemptive jump overcall. If RHO opens 1NT the preempt
loses much of its power since he has already described his hand
pretty accurately, so a jump overcall had better be pretty sound.
A preempt is more effective over a one of a minor opening than
over a one of a major opening, because the major-suit bid is better
defined, usually a 5-card suit, while the minor-suit opening might
be anything. Against a strong and artificial 1 ♣ opening bid it
is open season for preemption. The 1 ♣ bid says nothing at all
about distribution, so the preempt is likely to be most effective.
In addition there is now a very high probability that the hand
belongs to the opponents, so it is unlikely to go all pass and you
probably do not have a bad result if it does.
♠ rq ♣ ♠ rq ♣

Let’s now look at a few examples and decide whether or not the odds
favor a preempt. Keep in mind that preemptive styles vary greatly.
A player who tends toward sounder preempts may think the odds are
against preempting on a given hand, while a player who favors frequent
preempts may think the odds are right for the preempt. Nevertheless,
any player considering a preempt should run through the important
factors, check out the plusses and minuses, and use these to draw his
own conclusions.

None vul., South holds ♠ Q83 r AJ109765 q 7 ♣ 98

West North East South

— — Pass ?

Fair intermediates in hearts, r A is bad, outside queen is bad, spade

holding for potential support is bad, vulnerability is good, second seat
is bad. All in all it looks like too many minuses, so South should pass.
I consider myself an aggressive preemptor, but this hand just has too
much going against it. Remove one of the minus factors (say make
South dealer) and it becomes a marginal preempt. Remove a second
minus factor and the preempt is clear-cut.
301
 MATCHPOINTS

Both vul., South holds ♠ 6 r 106 q 6432 ♣ AQJ975

West North East South

— — — ?

The only bad features are the vulnerability (which isn’t terrible), the
♣ A, and 6-card suit. I consider it correct to open 3 ♣ . In my opinion,
there is more to gain than to lose. If that side 4-card suit were spades I
would definitely pass, and if it were hearts I would consider it a close
decision.

E-W vul., South holds ♠ 74 r KQ1065 q K742 ♣ 105

West North East South

— Pass Pass ?

This hand has just enough going for it to justify a weak 2 r bid, despite
the 5-card suit. The position (third seat) and the vulnerability are
perfect, and the suit is reasonable for a 5-bagger. The bid could turn
out badly, but it will gain more often than not.

E-W vul., South holds ♠ A1087652 r 8 q Q4 ♣ K75

West North East South

— — — ?

Despite the good position and vulnerability, this hand should be passed.
This is the worst type of suit on which to preempt—the ace and no
intermediates. In addition, South has way too much stuff outside. This
is the kind of hand on which if you open 3 ♠ and it goes all pass you
just know that you are headed for a bad result even before you see the
dummy. On a normal preempt you expect to have a chance even when
you are dropped there, but not on this hand.

302
 PREEMPTS

N-S vul., South holds ♠ 2 r 102 q KQ109754 ♣ 987

West North East South

— — Pass ?

Despite the vulnerability and position, I would chance a 3 q opening.

The hand is otherwise absolutely perfect for a preempt. It could go for
a number, but often the opponents won’t be able to catch you even when
you are in trouble. If the hand had so much as one flaw, I would refrain
from preempting at this vulnerability in second position.

None vul., South holds ♠ 7 r 10843 q K2 ♣ AQ10965

West North East South

— Pass 1q ?

I would try a 3 ♣ bid despite the marginal trump holding and the outside
q K. Partner is a passed hand, so we don’t have to worry about messing
his bidding up. If we don’t get stuck here the 3 ♣ bid will give the
opponents plenty of headaches. Had the opening bid been 1 ♠ the 3 ♣ bid
would be less effective, for it would be more likely that the opponents
had already found their trump suit.

♠ rq ♣ ♠ rq ♣

Assuming you have chosen to preempt on a hand, the next question is:
How high do you go? The old rule of one, two, and three (be at most
down 1 at unfavorable vulnerability, 2 at equal vulnerability, and 3 at
favorable vulnerability) has long been discarded by winning players as
too conservative. If you want to make life difficult for the opponents
you have to take some chances yourself.
A better philosophy is: Preempt as high as you dare, given the
conditions (vulnerability, seat, strength of suit, etc.). Occasionally you
will go for a number, but the damage you do to the opponents will more
than make up for it. An extra level of preemption really does make a
big difference. Look at it from the other side:

303
 MATCHPOINTS

N-S vul., South holds ♠ 87 r AQ q K96 ♣ AQJ1087

West North East South

— — 2♠ ?

No problem at all. South has a clear-cut 3 ♣ overcall, with reasonable

expectation of getting to the right contract. If notrump is right North
will probably bid it, and if North has a long heart suit, making that the
right strain, he can introduce it safely. Furthermore, if North doesn’t
have anything 3 ♣ is probably where we belong. However, let’s try the
same hand on a different bidding sequence:

West North East South

— — 3♠ ?

Not so happy now! South is too strong to pass, a 3NT bid without a
spade stopper would be a wild gamble, and a takeout double risks a 4 r
response on an inadequate suit. It looks like South is stuck with a 4 ♣
overcall as the least of evils, but this is no bargain. 3NT is lost forever,
4 ♣ might be too high, and partner with short clubs and a long heart suit
may be afraid to introduce his suit for fear of a misfit. Whatever South
does he is quite likely to wind up in the wrong contract, and may land in
a ludicrous one. No question about it, the 3 ♠ opening bid on this hand
is far more effective than the 2 ♠ opening bid. In keeping with Newton’s
law of gravitation, it is not unreasonable to say that the accuracy of the
opponents’ bidding varies inversely with the square of the level of the
preempt.
There is, of course, a major drawback to preempting too high. As
we have seen, the auctions that we tend to like the least when we
preempt are those that leave us playing in the preempt, doubled or not.
The higher we preempt, the more likely we are to buy the contract. A
weak 2 q opening bid may not damage the opponents too much, but we
are unlikely to get dropped there. A 4 ♠ opening bid, on the other hand,
is quite likely to end the auction. Consequently, we’d better be happy
to play it there. It should be noted that the rank of the preemptor’s
suit is important when considering how high to go. A 3 ♠ opening
bid is a much “higher” preempt than a 3 ♣ opening bid even though
they are on the same level, because the 3 ♠ opening bid wipes out the
entire 3-level while the 3 ♣ opening leaves the opponents room to play

304
 PREEMPTS

a 3-level contract. Therefore, a 3 ♠ opening bid is more likely to buy

the hand than a 3 ♣ opener, so you must be a bit more prepared to play
it there.
A good rule of thumb is: Count the number of cards in your suit,
subtract 4, and preempt at that level. We can see how this works by
looking at some model preempts. In each case South deals, none vul.

♠6 r KQJ842 q 1065 ♣ 1086. A model 2 r opening bid.

♠6 r KQJ10842 q 65 ♣ 1086. A model 3 r opening bid.
♠6 r KQJ108432 q 65 ♣ 86. A model 4 r opening bid.

Naturally this rule shouldn’t be followed blindly; it is just a guideline.

Other factors such as vulnerability, strength of suit, rank of suit, etc.
play an important part when considering how high to go.

E-W vul., South holds ♠ 7 r 83 q KQ10975 ♣ 10975

West North East South

— — — ?

Even if you are playing weak 2 q bids, it would be cowardly not to

open 3 q on this hand. A 2 q bid just doesn’t do enough damage, and
with this hand that is perfect for preempting, South should open 3 q
even though he has only a 6-card suit. Look at it using the Law of Total
Tricks. North’s average number of diamonds is 2, giving the partnership
8 diamonds. If E-W have an 8- or 9-card major-suit fit, as is likely, N-S
want to compete to 3 q anyway, so why not bid it now. However, with
♠ KQ10975 r 83 q 7 ♣ 10975 I would settle for a weak 2 ♠ bid. The
problem with opening 3 ♠ is that it may be too effective. The opponents
have no bidding room left at all, so they are quite likely to drop you
in 3 ♠. With only a 6-card suit, you can’t be very confident that this is
a good result. If partner has 2 spades and it is a partscore competitive
hand, the Law of Total Tricks indicates that you probably shouldn’t be
competing to 3 ♠ over their 3-level contract. If I were in third seat,
however, I would open 3 ♠. There would be so many plus factors that
the high preempt would figure to gain most of the time.

305
 MATCHPOINTS

Both vul., South holds ♠ AQ1085432 r 8 q 92 ♣ 95

West North East South

— — 1q ?

I would bid only 3 ♠ despite having an 8-card suit. The hand is

otherwise featureless, and it does contain the ♠ A—a definite liability
for a high preempt since you are quite likely to buy it for 4 ♠ doubled
if you bid it. The problem with 4 ♠ is that it doesn’t give the opponents
much choice. They pretty well have to double or pass, and this could
easily turn out badly for N-S. A 3 ♠ bid will do plenty of damage, and
there is much less risk of being stuck with the contract and being wrong.
However, with ♠ 8 r AQ1085432 q 92 ♣ 95 I would bid 4 r over 1 q .
Now the opponents may miss a spade fit or bid 4 ♠ when partner has a
spade stack. A mere 3 r call makes it a bit too easy for them to discover
their spade fit or lack of it.

306
 PREEMPTS

RESPONDING TO PREEMPTS

When your partner preempts, you are in complete control of the

auction. You know approximately what he has, and can accurately
assess both the offensive and defensive potential of the combined hands.
Consequently, it is up to you to make the decision for the partnership.
Partner will not bid again. Only in special situations (such as the “one-
under” auctions discussed earlier) will the preemptor even consider
taking action, so it is up to you to place the contract.
The most important principle to follow is: Be direct. Decide where
and how high you wish to go, and bid it immediately. The opponents
can’t always tell if you are bidding for a make or an advance save if you
go right to game. With E-W vul., the auction:

West North East South

— 2♠ Double 4♠

Would be correct with either ♠ A92 r K9 q KQ92 ♣ AJ98 or ♠ A1086

r 5 q 10872 ♣ K762, and if the opponents misjudge which hand you
have they will be headed for a very bad result, for they will either let
you take an undoubled save or go for a telephone number. You don’t
have to worry about fooling partner, for he has no further say in the
proceedings.
It usually does not pay to be tricky with psyches, false notrump
bids, or the wait-in-the-bushes approach opposite a preempt. These
tactics usually help the opponents. To see this, consider the following
example:

N-S vul., South holds ♠ KQ62 r AJ75 q 3 ♣ J1063

West North East South

3q Double 4NT ?

Obviously East is fooling around with his 4NT bid, and you should be
grateful to him for doing so. Now you can double showing strength.
When the bidding continues pass-pass-5 q as it almost certainly will,
you can now pass. After your double of 4NT this is forcing, so partner
will place you with high cards and an offensively oriented hand but no
good suit to bid, which is exactly what you have. With this knowledge

307
 MATCHPOINTS

he has a good chance of making the right decision, whether it be double,

bid five of a major, or bid a slam. However, suppose the bidding had
instead gone:

West North East South

3q Double 5q ?

Suddenly a nice descriptive sequence is no longer available. There is no

logical reason why a pass by you would be forcing, for East might be
bidding to make and you could hold a yarborough. You are definitely
too strong to sell out, so what should you do? A bid of five of a
major is a shot in the dark. It could be right, but you have no reason
to be confident of making 11 tricks in your best trump suit and, more
important, no way to find your best trump suit. Consequently, you are
pretty well forced to double. Partner will have to pass on most hands,
including many that will produce 11 or 12 tricks in a major, and the
penalty against 5 q will probably not be sufficient compensation. This
example shows how much easier it is to bid over fool-around bids than
over direct no-nonsense bids in this situation.

♠ rq ♣ ♠ rq ♣

If you choose to raise partner’s preempt, how high should you go? This
was discussed in the chapter on the last guess. You should go to the
level where you don’t know what the correct action for the opponents
is. For example:

E-W vul., South holds ♠ 93 r K543 q AJ10 ♣ 9875

West North East South

— 3r 3♠ ?

E-W are almost certainly cold for at least 10 tricks in spades. You could
conceivably go for 800 in 5 r, but it will be a good save far more often.
While there are E-W hands that will produce a slam, the odds seem to be
against it; in fact 5 ♠ just might be going down if your side can score one
heart trick and two diamond tricks. So what should you do? If you bid
4 r you know it is correct for E-W to bid 4 ♠. They will undoubtedly do
so, and you haven’t accomplished a thing except to give them a chance
308
 PREEMPTS

to exchange some information to make a later decision easier. If you bid

5 r you don’t really know what E-W should be doing. If 5 ♠ is down or
if 5 r is −800 they should double, but there is certainly a good chance
that 5 ♠ will make and 5 r will cost 500. If you bid 6 r you know that it
is correct for E-W to double. 6 ♠ is probably down, and 6 r could easily
be set 4 tricks. Consequently, it is correct to bid 5 r. Since you don’t
know what they should do neither will they, so their next bid will be the
last guess, which is what you want.

None vul., South holds ♠ A103 r K54 q Q54 ♣ K763

West North East South

— 2r Double ?

Bid 3 r. Not only is this bid indicated by total tricks, but it puts the most
pressure on the opponents without sticking your neck out too far. You
don’t know whether you are rooting for them to bid, pass, or double, so
how can they know what to do? If you bid 4 r it is too easy for them
just to double for lack of anything better to do, and you probably won’t
like it. Conversely, if you pass the opponents will probably subside in
2 ♠ or three of a minor. Now you will be compelled either to sell out,
which is probably wrong on trick total, or to bid 3 r after the opponents
have exchanged more information and increased the risk of a double.

♠ rq ♣ ♠ rq ♣

There is one type of hand on which a “wait and see” approach is correct.
If you know that you have a good sacrifice at any level, it may be a
good idea to let the opponents bid to what they think they can make
unimpeded, and then save over that. For example:

E-W vul., South holds ♠ K954 r A q 975432 ♣ 96

West North East South

1r 3♠ 4♣ ?

You figure to be safe at the 5-level, if necessary. The problem with

bidding any number of spades at this point is that the opponents just
might bash into a slam. Then what would you do? If you save you
309
 MATCHPOINTS

might find that the slam is off two cashing aces, while if you don’t save
the slam could well be cold. You have to make the last guess, which is
what you are trying avoid. Your best bet is to pass and let the opponents
find what they can make in peace. They will not leap to a slam trying to
stampede you into a save, for they have no way to know you have any
interest in saving. Consequently they will bid the hand as accurately as
they can, so you can trust their bidding and eventually take a save over
whatever contract they reach.

310
 PENALTY DOUBLES

The hand clearly belongs to the opponents and they have arrived in their
contract, but you have reason to believe they will go down. Should
you double? This is often a very difficult problem. There is more
involved than just the probability of setting the contract, although that is
obviously the number-one consideration. As always, we are concerned
about the potential matchpoint gain if the double is successful vs. the
potential matchpoint loss if it fails.
The double can affect the probability of the contract’s success. The
contract may be more likely to make because of the double, more likely
to go down because of the double, or the double might not make any
difference.

N-S vul., South holds ♠ 953 r A875 q — ♣ 987653

West North East South

1♠ Pass 3♠ 1 Pass
4NT ∗ Pass 5r ∗ Pass
6♠ Pass Pass ?
1. Limit raise

If South doubles, it is a Lightner double calling for an unusual lead.

On this auction the double should show a void, so North will lead his
longest side suit. That suit is likely to be diamonds, so the double
improves South’s chances of beating the contract.

E-W vul., South holds ♠ QJ109 r 982 q A8 ♣ 5432

West North East South

— — 1♠ Pass
2♠ Pass 4♠ ?

A double by South will not affect the chances of 4 ♠ making. If partner

has a trick it will go down; if not, it will make.
 MATCHPOINTS

None vul., South holds ♠ 72 r QJ97 q 52 ♣ Q9753

West North East South

— — 1r Pass
3q 1 Pass 3r Pass
4r Pass 6r ?
1. Game-forcing

On this hand, a double could improve the chances of the contract

making. Without the double declarer would lose two trump tricks. If
South doubles, however, declarer knows about the heart stack, and if
he can now play the hearts so as to lose only 1 trick he will make an
otherwise unmakeable contract.

♠ rq ♣ ♠ rq ♣

There is another way for a double to affect the outcome. If the

opponents have a possible alternative contract the double may warn
them of a bad trump split and drive them to the other contract. For
example:

Both vul., South holds ♠ 62 r J10985 q Q98 ♣ 642

West North East South

2NT Pass 3q ∗ Pass
3r Pass 4NT Pass
6r Pass Pass ?

South can certainly defeat 6 r, but it is possible that 6NT will make.
South must take into account both the probability of 6NT making and
the likelihood of the opponents running to 6NT when he considers
doubling.
It is also possible for a speculative double to drive the opponents to
a worse contract. For example:

312
 PENALTY DOUBLES

Both vul., South holds ♠ QJ103 r 65 q A73 ♣ KJ42

West North East South

— — 1r Pass
1♠ Pass 2♠ Pass
3r Pass 4r ?

If South chooses to double 4 r, E-W may think that it is the hearts that
are breaking badly and run to 4 ♠, which South would enjoy.

♠ rq ♣ ♠ rq ♣

If you are considering doubling a voluntarily bid game or slam, it is

important to determine if the contract is likely to be reached at other
tables. If the contract figures to be a normal one, then you have even-
money odds on your double—you simply gain 1⁄2 matchpoint from every
other table if you are right and lose 1⁄2 matchpoint to every other table if
you are wrong. Consequently, all that is needed is that you be a favorite
to defeat the contract after the double. This means that you can risk
helping declarer play the hand if the contract is still more likely than
not to go down after declarer has this help.

None vul., South holds ♠ A6 r Q108 q 10963 ♣ 9764

West North East South

— — 1r Pass
3r 1 Pass 4NT ∗ Pass
5q ∗ Pass 6r ?
1. Limit raise

Should South double? If he doesn’t double he is almost sure to defeat

the slam; declarer probably won’t pick up the trump suit even if he can.
If South does double, however, he will tell declarer how to make the
contract if dummy has AJxx or KJxx in trumps. Declarer is more
likely to have the trump strength, so the slam will still go down more
often than not. Let’s suppose that the slam will go down all the time
if not doubled, 70% of the time if doubled. On this strong auction, it
seems clear that the slam is easy to bid and much of the field will be
there. Assume that the slam isn’t doubled at other tables. If you don’t

313
 MATCHPOINTS

double, you will get an average for +50. If you double, 70% of the
time you get a top for +100, and 30% of the time you get a bottom for
−1210. Double is obviously the percentage action, even though it may
get you a very bad board. When this hand came up, the slam was bid at
12 out of 13 tables. The two enterprising players who doubled collected
+100 and 111⁄2 matchpoints on a 12 top, while the remaining ten who
didn’t double had to settle for 51⁄2 matchpoints, not realizing that they
had passed up a golden opportunity.
If the contract may not be reached at other tables, the odds on
information-giving doubles change radically. On the previous hand,
suppose the auction had instead been:

West North East South

— — 1r Pass
3r 1 Pass 4♣ ∗ Pass
4q ∗ Pass 4r Pass
5♣ ∗ Pass 5q ∗ Pass
5r Pass 6r ?
1. Limit raise

They finally got there, but what a struggle! Since each opponent signed
off at some point in the auction, this is a marginal slam at best. It
seems reasonable to assume that at least half of the other pairs won’t
bid the slam. Let’s look at the matchpoint cost analysis of the double
under this new assumption. If we double and are right (i.e., we beat
the slam anyway), the double gains 1⁄2 matchpoint from other tables in
the slam. It gains nothing against tables in game, for they were beaten
regardless. If we double and are wrong (i.e., the double tells declarer
how to make the slam, which he otherwise would not have done), we
lose 1⁄2 matchpoint to other tables in slam. In addition, we now lose a
full matchpoint to other tables in game, for we would have beaten them
had we kept quiet. Consequently, if we assume that half the field will
bid the slam we must set them at least 75% of the time for the double
to be correct, because the double gives up 3-to-1 matchpoint odds. Not
doubling gets us 9 on a 12 top, while doubling gets us either a 0 or a 12,
so we are risking 9 matchpoints to gain 3. Under these assumptions,
the double is no longer a percentage bid. It should be noted that the
double would be a terrible bid at IMPs or rubber bridge regardless of

314
 PENALTY DOUBLES

the auction. When it is right it gains 50 points or 2 IMPs, but when it

is wrong it costs 1260 points (from +50 to −1210) for a 15-IMP loss.
At matchpoints this type of double can be quite correct since frequency
rather than size of the swing is what matters.

♠ rq ♣ ♠ rq ♣

If the opponents are in a normal contract, it is necessary to be a

favorite to defeat them before you double. Even if the double improves
the defense’s chances, it is not a good double unless those defensive
chances go over the 50% mark.

None vul., South holds ♠ 72 r AQ1098 q 732 ♣ K72

West North East South

1♣ Pass 1r Pass
1NT Pass 3NT ?

South obviously wants a heart lead, and there doesn’t appear to be much
hope for the defense without one. Unfortunately, South has no reason
to be confident that a heart lead will set the contract, although it will
certainly improve the chances of a set. The auction is strong, 3NT is
probably an easy contract to bid, and it will be reached at just about
every table. If South doubles he gains 1⁄2 matchpoint from every table
if he is right, and loses 1⁄2 matchpoint to every table if he is wrong, so
it’s a 50–50 proposition. Since South can’t expect to be a favorite to
defeat 3NT even with the heart lead, the double is incorrect. Note that
at IMPs the double would be much better, for it loses 150 points or
4 IMPs (−400 to −550) when wrong, but gains 500 points or 11 IMPs
(−400 to +100) when right. These figures have to be shaded due to the
possibility of a redouble and/or overtricks, but the odds on the double
are still much better than at matchpoints.
Make the auction on the last hand a bit slower:

315
 MATCHPOINTS

West North East South

1♣ Pass 1r Pass
1NT Pass 2NT Pass
3NT Pass Pass ?

Now the odds on the double are much better. In the first place, the
opponents do not have excess strength since East only invited, so the
chances of beating the contract are better than before. More important
from a matchpoint angle, E-W have stretched to a thin game that will
not be bid at some tables. If South doubles and is wrong he loses
1⁄ matchpoint only to those tables in game, as the hand was already lost
2

to those not in game. If South doubles and is right (the lead is necessary
to defeat 3NT), he gains 1⁄2 matchpoint from other tables in game and
also gains one full matchpoint from each table not in game. In this case
if the double improves South’s chances of beating the contract by 25
or 30% the double is correct, even if South is still a slight underdog to
defeat 3NT.

♠ rq ♣ ♠ rq ♣

If the opponents have more than one possible game or slam contract
available, it is important to decide if there are other higher-scoring
contracts. You will get a bottom regardless if you double and are wrong,
but the double will not be too costly if the opponents are in their highest-
scoring contract, for you will lose 1⁄2 matchpoint only to other tables in
the same contract. As an extreme example, suppose you double 7NT.
This is a top-scoring contract, so you only gain or lose 1⁄2 matchpoint
against other tables in 7NT; therefore it is an even-money proposition
assuming the double will not affect the play. However if you double 7 ♠
there is more to lose when wrong. If both 7 ♠ and 7NT make you lose
a full matchpoint to other tables in 7NT as well as the 1⁄2 matchpoint to
tables in 7 ♠. This is because 7 ♠ is not a top-scoring contract. A more
practical example:

316
 PENALTY DOUBLES

E-W vul., South holds ♠ A6542 r 9753 q KQ10 ♣ 9

West North East South

1♠ 3♣ 3q Pass
4q Pass 5q ?

South is certainly a favorite to defeat 5 q . He has excellent chances

of scoring one spade trick and two diamond tricks, and even if one
of these tricks evaporates partner may produce a trick. Nevertheless,
South should refrain from doubling. Either E-W have gone completely
off the walls, in which case it won’t matter much what South does since
an undoubled penalty will defeat N-S scores at other tables, or E-W
have most of the outstanding high-card strength. If the latter is the
case, it would be hard to imagine E-W not having a good play for 9
or 10 tricks at notrump, which is likely to be a more popular contract
than 5 q . Consequently, it would be very costly to double, be wrong,
changing your score from −600 to −750, and lose a full matchpoint to
every −630. This is one double that has far more to lose than to gain.

Both vul., South holds ♠ K6 r QJ109 q KJ8 ♣ 7652

West North East South

— — 1q Pass
1r Pass 1NT Pass
2♣ 1 Pass 2r Pass
2♠ Pass 2NT Pass
3r Pass 4r ?
1. New Minor Forcing

Normally this is a good situation for doubling a voluntarily bid game.

The opponents staggered into game, they are getting a bad trump split,
and South’s kings lie over the opening bidder. While 4 r is a favorite
to go down in my opinion, a double is still not recommended. The
reason is that the opponents have an alternative possible contract, 3NT,
which may be found at other tables, and it is probably a higher-scoring
contract on this hand. It would be a matchpoint catastrophe to double
4 r, have them make it for −790, and then find you could have had a
very good board by passing because the field was in 3NT making four.
If the auction had instead gone:

317
 MATCHPOINTS

West North East South

— — 1q Pass
1r Pass 2r Pass
3r Pass 4r ?

The double would be substantially better. The chances of beating 4 r are

about the same, but the cost of being wrong by doubling has decreased.
If E-W had a reasonable auction it is clear that the field will be in hearts,
so if you think they are more likely than not to go down you should
double. It should be noted that it does not matter what percentage
of the field bids the game on this hand, for the double will not affect
the comparison with tables not in game. The double simply gains
1⁄ matchpoint from other tables in 4 r if right, and loses 1⁄ matchpoint to
2 2

them if wrong. If the double might help declarer in the play this analysis
wouldn’t be so simple, as we have seen, but on this hand declarer can
do nothing about the bad heart split so the double is unlikely to affect
his chances of making 4 r.
Conversely, if the opponents are in a top-scoring contract, a double
may be correct even if they are slight favorites to make. The reason
is that you get a very bad board if they make it anyway, so you might
as well assume that they are going down and collect as big a number
as possible. This type of double should only be made when you
have reason to believe that the alternative contract will also go down;
otherwise the double doesn’t stand to gain when it is right.

Both vul., South holds ♠ QJ92 r A8 q 962 ♣ K732

West North East South

— — 1NT Pass
2♣ Pass 2♠ Pass
3♠ Pass 3NT ?

Just your luck! The opponents looked like they were headed for 4 ♠, and
all of a sudden they may have stopped off in 3NT. This is very likely to
be the best-scoring contract if it makes, as it is hard to imagine spades
taking one more trick than notrump considering your spade holding.
In addition 4 ♠ is quite likely to go down, for you are looking at at
least 3 tricks and partner is marked with about 5 or 6 points by West’s

318
 PENALTY DOUBLES

non-forcing 3 ♠ call. Consequently, you figure to lose to other tables in

4 ♠ or in a partscore if 3NT makes, and the double if wrong will cost
1⁄ matchpoint to only those few pairs who also reject the 4–4 spade fit.
2

Now let’s look at the bright side. Suppose the double is right and 3NT
is going down. If this is the case you will gain at least 1⁄2 matchpoint
from all other tables in game, whether it be 3NT or 4 ♠. 4 ♠ contracts
will probably escape undoubled because defenders won’t risk telling
declarer about the spade stack. As an added bonus, the double might
scare some unsophisticated players back into 4 ♠, which you would
love. Since this double has so much more to gain than to lose, I think it
is correct, even though I rate 3NT a slight favorite to make on the given
information.

♠ rq ♣ ♠ rq ♣

On occasion, your opponents will have a seemingly absurd auction or

land in a contract that appears ridiculously high from looking at your
hand. Should you double? Many people say no. Their reasoning is:
If the contract is that absurd nobody else will be there, so the double
has nothing to gain. This appears to make sense, but the argument can
be put the other way: If nobody else is there the double has nothing
to lose, so why not double? The important question is: What makes
the contract absurd? If you conclude that if they are going down then
nobody else will be there while if they are making they will have some
company, then it is correct to refrain from doubling. For example:

E-W vul., South holds ♠ 9542 r A8 q A84 ♣ 8764

West North East South

1♣ 2q 2r 5q
6r Pass Pass ?

South’s 5 q call was quite enterprising, but it certainly seems to have

paid off. West was deprived of room to investigate slam possibilities,
so he bashed into a slam that is probably off two cashing aces. The
question is: Should South rub salt into the wound with a double?
Either the q A cashes or it doesn’t, so the play will not be affected
by a double. What about other tables? They are not likely to face

319
 MATCHPOINTS

such fierce competition, and so they should be able to find out if they
are off two cashing aces and stop short of slam. Therefore if the q A
cashes doubling won’t gain anything because nobody else will be in
slam. However, if the q A doesn’t cash there may well be some other
pairs finding their way to slam on this hand. The double can never gain
anything if it is right, but it will cost 1⁄2 matchpoint to other pairs in slam
if it is wrong. Consequently, South should pass, even though 6 r is a
big favorite to go down.
Don’t get into the habit of not doubling when you think the
opponents have gone crazy. If they got so high it may be a trouble hand,
and others may have problems also. If this is the case, it is important to
collect as big a number as you possibly can. Keep in mind that if the
opponents land in a crazy contract and make it your double will only
make your zero a little rounder. Be sure that they don’t have a likely
runout that can make and that there is no great risk of your double
telling them how to make the hand, but for the most part it is quite
correct to double what is under your nose.

N-S vul., South holds ♠ 10942 r AQ102 q K4 ♣ 652

West North East South

1r Pass 2q Pass
3r Pass 4NT Pass
5q Pass 6r ?

Double away. Sure, they might make it if East has three hearts and
enough entries to take three heart finesses and doesn’t have a diamond
loser, but all this is extremely unlikely. It is important to double, for
many pairs holding the E-W cards will be going minus due to the bad
heart split, and you must collect as big a number as possible.

320
 PENALTY DOUBLES

Both vul., South holds ♠ A8 r 73 q KQJ6 ♣ J10432

West North East South

1r Pass 1♠ Pass
3r Pass 4NT Pass
5r Pass 6NT ?

What are you doing with all these cards? Don’t worry about it—just
double and lead the q K. If you are an extremely unlucky bridge player,
the full hand will be:
♠ J542
r 965
Both vul.
q 9843
♣ 96
♠ KQ963
N
r2
W E
♠ 10 7 q A75
S
♣ KQ87
r A K Q J 10 8 4
q 10 2
♣ A5
♠ A8
r 73
q KQJ6
♣ J 10 4 3 2

East wins the opening diamond lead and runs his hearts, keeping
♣ KQxx and a spade honor. You are squeezed in three suits, and must
give declarer his twelfth trick. Unlucky, but don’t blame the double—
blame West for holding seven hearts and the q 10, or East for holding a
4-card club suit, or perhaps the gods of bridge for dealing you this hand
in the first place. On this hand 6NT will make whether you double or
not, so the cost of doubling is relatively small since the opponents are
certainly in the top-scoring contract. It is far more likely that 6NT will
go down, so you must collect the most you can to guard against all the
other E-W pairs who may be finding their way into trouble.

♠ rq ♣ ♠ rq ♣

321
 MATCHPOINTS

If partner doubles the opponents at a game contract, it is almost never

right to pull unless you think you have some chance of making what
you bid. Even if you have stretched considerably for a previous bid
and he may be basing his double partly on values you have shown and
don’t have, you are better off passing and taking your lumps. The only
exception occurs when you were planning on taking a sacrifice and he
doubled before you had a chance to save. Even then, think twice before
pulling. It is just possible that partner knows what he is doing. If you
were not planning on taking a sacrifice then the reason must be that
the save figures to go down too much, and this estimate hasn’t been
changed by partner’s double. Avoid traps such as the following:

None vul., South holds ♠ QJ1095 r 6 q K83 ♣ J954

West North East South

— — 1♣ 1♠
2r 2♠ 3r Pass
4r Double Pass ?

Many players will make the mistake of bidding 4 ♠ here. They would
not have taken the save if partner hadn’t doubled simply because 4 ♠
is too likely to be −500. Once partner doubles, they say to themselves
“I don’t have the values for my overcall and partner had only a 2 ♠
bid, so how can we beat them? I must pull to protect partner, who is
playing me for more than I have.” This is faulty reasoning. Perhaps
South should have passed or made a weak jump overcall rather than
overcalling 1 ♠, but that is water over the dam. If South felt that 4 ♠
would be set 4 tricks before the double, prospects are even worse now
that partner presumably has values in hearts. Bidding 4 ♠ even if it is
“right” (i.e., −500 instead of −590) won’t salvage many matchpoints,
but if 4 r happens to be going down the 4 ♠ bid will be a complete top-
to-bottom swing. Partner is allowed to know what he is doing. The full
hand:

322
 PENALTY DOUBLES

♠ K83
r K Q 10
None vul.
q 10 9 7 6 2
♣ 10 3
♠ A2 ♠ 764
N
r A9742 r J853
W E
q J54 q AQ
S
♣ K87 ♣ AQ62

♠ Q J 10 9 5
r6
q K83
♣ J954

As you can see, everybody has their bids except South, and 4 r is a
normal contract. Partner was simply betting that 4 r would go down,
and he was right. If you bid 4 ♠ you will be −500 against a game that
isn’t making—not a very impressive result. To make matters worse
you would probably have had a top, for not many will be doubling 4 r.
This sort of error is one of the most costly mistakes one can make at
matchpoints.

323
 SACRIFICE BIDDING

The goal on most bridge hands is to achieve the greatest possible plus
score. There are, however, some hands on which a plus score is very
unlikely. On these hands the goal may change to getting the smallest
possible minus score. One way to do this can be to take a successful
sacrifice against the enemy game or slam.
Sacrifices are very dangerous bids to make. For a sacrifice to be
successful, three conditions must be met:

1. The enemy contract must make.

2. The sacrifice must not be too expensive.
3. The field must get to the game or slam against
which you are sacrificing.

This is quite a parlay. All through this book we have been searching for
bids that have more than one way to win, and a sacrifice has three ways
to lose! If any part of the parlay fails your sacrifice will net a very poor
matchpoint score, although the cost of being wrong may vary. If the
field fails to get to the contract your sacrifice costs nothing (if anything
it gains a bit), but you remain with a poor matchpoint score and lose
out on any possible chance of declarer’s going down in his contract.
If the sacrifice is too expensive you lose 1⁄2 matchpoint to every other
table, assuming you would have tied the normal result of game bid and
made had you not saved. The most serious error occurs when the enemy
contract isn’t making, which means you have taken a phantom sacrifice.
This not only costs 1⁄2 matchpoint to other tables at which the same
contract is reached, but also a full matchpoint to those tables where
the pairs holding your opponents’ cards stop short of game. To put it
another way, you turn a well-above-average score into a bottom.
There is a good general rule about sacrifices that will avoid the
parlay problem. You should never sacrifice unless you are virtually
100% sure of two of the three necessary conditions. In other words, if
you think (but are not sure) that the enemy contract will make and you
 MATCHPOINTS

think (but are not sure) that your save will not be too expensive, then you
should not take a sacrifice. The problem is that the conditions are not
necessarily independent. The onside finesse that allows the opponents
to make their contract just might be the offside finesse that defeats your
save one trick too many.

None vul., South holds ♠ A843 r 43 q QJ98 ♣ KQ9

West North East South

1r 2♠ 4r ?

Should South save? It is probable that 4 r is making, but by no means

certain. It would not be unreasonable to score 1 spade, 1 diamond, and
2 clubs on defense. It is also probable that N-S can score 8 tricks in
spades, but this is also nowhere near certain. N-S could easily lose
2 tricks in each side suit if North doesn’t have a singleton. South can’t
even be all that sure that the game will be bid at all the other tables, but
this isn’t quite so important—on a distributional fitting hand you can
assume that other tables will get to game even without 26 HCP. A look
at all four hands shows the dangers involved in taking a sacrifice that
bets on a parlay.

♠ KQJ752
r 95
None vul.
q 10 6
♣ 10 4 2
♠ 10 9 ♠ 6
N
r A Q 10 8 6 r KJ72
W E
q K73 q A542
S
♣ AJ5 ♣ 8763

♠ A843
r 43
q QJ98
♣ KQ9

4 r makes, but 4 ♠ goes down 3 for −500 if the defenders attack clubs
in time, which shouldn’t be too hard. Suppose the E-W minor-suit aces
were interchanged. Now 4 ♠ goes down 2, but 4 r also may go down.

326
 SACRIFICE BIDDING

A slight change in the hand or the conditions would make the

sacrifice a reasonable bid. If N-S were at favorable vulnerability or if
South were a bit more distributional, say ♠ A843 r 4 q QJ983 ♣ KQ9
South could be quite confident that 4 ♠ would not be too expensive a
save, although 4 r still might go down. Take away South’s diamond
holding, leaving him ♠ A843 r 43 q 9842 ♣ KQ9, and now South can
be virtually sure that the 4 ♠ sacrifice won’t be a phantom, even though it
might be too expensive. The idea is to have all but one of the conditions
nailed down tight. Then you can examine the one condition in question,
and if you think the odds are right go ahead and save.

♠ rq ♣ ♠ rq ♣

Sacrificing against slams is a very tricky proposition. Usually the save

won’t be too expensive, or you wouldn’t be thinking about saving in
the first place. The big problem is that many other pairs may not bid
the slam, so that even if the save is a success you won’t get many
matchpoints. If there is any kind of chance that the slam can be beaten
it is better to defend unless you are quite sure that the field will bid
the slam. Incidentally, potential slam sacrifices are often an exception
to the “double or save” principle. You don’t double because you think
slam is a favorite to make. However, you don’t save even though you
think the save will be right more often than not because the gain when
the save is right is small if most of the field doesn’t bid the slam, while
the cost of saving if the slam is going down is huge.
Many players will take a marginal save against a slam if they think
their save will go down less than the value of an enemy game. This
approach has some validity, for the save now has a second way to win—
some pairs may sell out to the game. However, in a good field this is a
losing approach. The danger is that other pairs holding your cards may
buy the contract one level lower, in which case the save gains nothing
when it is right (since you will go for an extra trick) but loses a full
matchpoint to these pairs when it is wrong.

327
 MATCHPOINTS

E-W vul., South holds ♠ 108763 r 9 q QJ32 ♣ A97

West North East South

1r 3♠ 4r 4♠
6r Pass Pass ?

South has reasonable defensive prospects against the slam. North might
have a singleton club, or E-W may have a third-round diamond loser
that can’t be disposed of. Some players will take the save anyway,
figuring that there might be 9 tricks available in spades so the save
would go down −500, less than the value of game. However, had the
auction gone:

West North East South

1r 3♠ 4r 4♠
5r Pass Pass ?

These same players will pass even though they know they almost
certainly have a good save available at 5 ♠. Their reasoning would be
that they don’t want to push the opponents to slam. Personally, I think
this is exactly backward. South should take the 5 ♠ save that he knows
is correct, but if the opponents bid 6 r South should take his chances
on defense. The 6 ♠ save just has too many ways to lose, particularly
since South has good defensive prospects against 6 r. The slam might
go down, other N-S pairs may buy the hand for 5 ♠ doubled, and the 6 ♠
save might still be −800. The 5 ♠ save on the other hand is a definite
improvement over defending 5 r, and if it pushes the opponents to slam,
which probably won’t happen, they haven’t made it yet.
There are some added bonuses that can make sacrificing more
profitable. They are:

1. The opponents might take the push and bid one more. You know
you have done well with your sacrifice when this happens, for you
now have a free shot at a plus score. This is another reason why
the most important consideration when pondering a sacrifice is
whether or not it is a phantom. If you are absolutely sure that the
opponents are making, it may be worth risking a save being too
expensive if you think there is a chance of pushing them higher.

328
 SACRIFICE BIDDING

E-W vul., South holds ♠ 103 r KJ108 q A43 ♣ 8753

West North East South

1♠ 3q 4♠ ?

South cannot be sure of making 8 tricks in diamonds. It looks as

though 4 ♠ is cold, South may have some chance against 5 ♠, and he
will welcome a 6 ♠ bid. Consequently, 5 q is a reasonable call. If South
were sure that 5 q would be doubled he might not bid it, for it could
easily go for −800 if partner has the wrong hand. However, the extra
way to win—the opponents take the push to 5 ♠ and go down—makes
the 5 q sacrifice a worthwhile gamble.

2. The opponents may not double. It can happen that you know you
are taking a sacrifice, but E-W don’t know enough about the hand
to be sure of this. If this is the case, you may escape undoubled
for a very good board.

E-W vul., South holds ♠ QJ4 r 86 q KQ32 ♣ QJ82

West North East South

— 2♠ 3r ?

4 ♠ will certainly be a good save against 4 r if 4 r is making, but it is

not clear that 4 r will make. South could try to buy the hand for 3 ♠, but
this may goad West into a marginal 4 r call and then South would have
no idea what to do. A better idea is to bid 4 ♠ at once. South knows
that he is taking a save, but E-W may not know this. If neither of them
has extra strength it will be very difficult for them to double, for South
could easily be bidding to make. The full hand:

329
 MATCHPOINTS

♠ A K 10 9 5 3
r 53
E-W vul.
q 876
♣ 96
♠ 86 ♠ 72
N
r K74 r A Q J 10 9 2
W E
q A95 q J 10 4
S
♣ 10 7 5 4 3 ♣ AK

♠ QJ4
r 86
q KQ32
♣ QJ82

If South passes West might or might not bid game, while if South bids
3 ♠ West will be more inclined to try 4 r. It turns out that 4 r makes
because of West’s q 9, and 4 ♠ goes down 3 because of that same card.
However, if South jumps to 4 ♠ it is very hard for either East or West
to double, for they have no reason to think that South isn’t bidding to
make. If South slips it by, 4 ♠ undoubled will be −150, which even
beats those defending a heart partial for −170.
3. The save might make. This, of course is the best kind of sacrifice
to take. The extra way to win, the possibility of making, radically
improves the odds on the sacrifice. For example:

None vul., South holds ♠ AJ6 r K2 q J1087 ♣ KQ82

West North East South

— 3♠ 4r ?

South has pretty good defensive prospects against 4 r, so the save is not
at all clear. If South chooses not to save, he should of course double,
for this is a perfect “double or save” situation. The deciding factor in
favor of the save is that 4 ♠ just could be a make if North has a singleton
diamond. This possibility, coupled with the chance that 4 r makes and
4 ♠ is a good save, tips the balance in favor of the 4 ♠ call. If South’s
save had no chance to make, then it would be better to defend.
These factors can give extra incentive toward sacrificing. However,
most of the time your sacrifice will simply be doubled and you will
330
 SACRIFICE BIDDING

get a minus score. A successful sacrifice usually nets only a slightly-

above-average matchpoint score, while an unsuccessful one virtually
always gets a bottom. The dice are loaded against the saver, because
he is betting on a parlay. Consequently it is generally best to avoid
marginal sacrifices; stick to those that are clearly correct. A bad
sacrifice, particularly if it is a phantom, is one of the most costly bids
at matchpoints. The saver voluntarily takes the last guess upon himself,
so he’d better be right.

331
DEFENSIVE BIDDING
REVIEW PROBLEMS

Problem 1

None vul., South holds ♠ K r QJ10983 q 754 ♣ AKJ

West North East South

1♠ Pass 2q ?
 MATCHPOINTS

Problem 1 Solution

None vul., South holds ♠ K r QJ10983 q 754 ♣ AKJ

West North East South

1♠ Pass 2q ?

Pass. A 2 r overcall is safe enough, but what does it accomplish? No

space is consumed, N-S are unlikely to buy the contract, and South
is not interested in a sacrifice. As for lead direction the overcall has
negative value, for South wants a club lead against a spade contract.

334
DEFENSIVE BIDDING REVIEW PROBLEMS

Problem 2

None vul., South holds ♠ 3 r 63 q A732 ♣ KQ10432

West North East South

— Pass 1q ?

335
 MATCHPOINTS

Problem 2 Solution

None vul., South holds ♠ 3 r 63 q A732 ♣ KQ10432

West North East South

— Pass 1q ?

3 ♣. With partner a passed hand and none vul., this preempt is worth
the gamble. It could go for a number, of course, but if E-W have
to search for their best contract the difference in space consumption
between 2 ♣ and 3 ♣ is enormous. A 2 ♣ overcall may be no more than a
mild annoyance, while the 3 ♣ bid will make accurate exploration very
difficult for the opponents.

336
DEFENSIVE BIDDING REVIEW PROBLEMS

Problem 3

E-W vul., South holds ♠ KQ109432 r J87 q 752 ♣ —

West North East South

1r Pass 2♣ 3♠
4NT ∗ Pass 5r ∗ Pass
5NT ∗ Pass 6q ∗ Pass
6r Pass Pass ?

337
 MATCHPOINTS

Problem 3 Solution

E-W vul., South holds ♠ KQ109432 r J87 q 752 ♣ —

West North East South

1r Pass 2♣ 3♠
4NT ∗ Pass 5r ∗ Pass
5NT ∗ Pass 6q ∗ Pass
6r Pass Pass ?

Pass. This is not the time for a Lightner double. The opponents may
well have a successful runout to 6NT. Even if they stay in 6 r South
can’t be sure of beating the slam with a club lead. If the slam still makes
after the ruff and the same number of tricks were available in notrump,
then the double becomes very costly, because the opponents are not in
the top-scoring contract. Lastly, South can’t even be positive that he
wants a club lead (imagine North with Qx of hearts, short spades, and
a natural club trick, for example).

338
DEFENSIVE BIDDING REVIEW PROBLEMS

Problem 4

E-W vul., South holds ♠ 4 r K96 q AQ10532 ♣ Q105

West North East South

— — 1♠ 2q
2♠ 3q 4♠ ?

339
 MATCHPOINTS

Problem 4 Solution

E-W vul., South holds ♠ 4 r K96 q AQ10532 ♣ Q105

West North East South

— — 1♠ 2q
2♠ 3q 4♠ ?

Pass. South probably can take 8 tricks in diamonds, and E-W probably
can make 4 ♠, but South is far from certain about either of these
conditions. There are unlikely to be any of the bonus factors for saving
(5 q can’t be making and E-W will almost certainly double), so the
sacrifice doesn’t figure to be the percentage bid.

340
DEFENSIVE BIDDING REVIEW PROBLEMS

Problem 5

N-S vul., South holds ♠ KQ9 r J10753 q A ♣ AQ52

West North East South

— — 1q ?

341
 MATCHPOINTS

Problem 5 Solution

N-S vul., South holds ♠ KQ9 r J10753 q A ♣ AQ52

West North East South

— — 1q ?

1 r. This will make future bidding easier than if South starts with
a takeout double. If the opponents are in 3 q by the time it comes
around to South, as they often will be, South will be better placed after
having overcalled, as he can then double and keep all his options open.
Interestingly enough, if we take away one of South’s aces the double
becomes a better bid. South is only planning to take one bid on the
hand, and with a weak heart suit and good support for all other suits the
double is more descriptive.

342
DEFENSIVE BIDDING REVIEW PROBLEMS

Problem 6

E-W vul., South holds ♠ 2 r 95 q KJ9532 ♣ A763

West North East South

— 3♣ 3♠ ?

343
 MATCHPOINTS

Problem 6 Solution

E-W vul., South holds ♠ 2 r 95 q KJ9532 ♣ A763

West North East South

— 3♣ 3♠ ?

6 ♣. E-W can certainly make at least 5 ♠, and it is hard to imagine that

South can buy the hand for 5 ♣. Furthermore N-S can probably take
9 tricks in clubs, so 6 ♣ figures to be a good save. South doesn’t know
what E-W should do over 6 ♣, so he should bid it planning to sell out
if E-W bid the slam. This approach will make the E-W slam decision
most difficult. In addition, if E-W happen to have enough distribution
to make a grand then this 6 ♣ bid makes it virtually impossible for them
to get there.

344
DEFENSIVE BIDDING REVIEW PROBLEMS

Problem 7

Both vul., South holds ♠ 62 r 5 q K98 ♣ AQ109643

West North East South

1q Pass 1NT ?

345
 MATCHPOINTS

Problem 7 Solution

Both vul., South holds ♠ 62 r 5 q K98 ♣ AQ109643

West North East South

1q Pass 1NT ?

2 ♣. There just is no good reason to preempt. East has described his

hand and E-W have already determined that they don’t have a major-
suit fit, so the preempt accomplishes little and risks going for a number.
The 2 ♣ overcall probably won’t do much good in the auction, but it’s
pretty safe and will get partner off to the right lead if West declares a
diamond partial, as is likely.

346
DEFENSIVE BIDDING REVIEW PROBLEMS

Problem 8

N-S vul., South holds ♠ 102 r J954 q 32 ♣ AKQJ7

West North East South

1q Pass 2q 1 Pass
2♠ Pass 2NT Pass
3NT Pass Pass ?
1. 10+ points

347
 MATCHPOINTS

Problem 8 Solution

N-S vul., South holds ♠ 102 r J954 q 32 ♣ AKQJ7

West North East South

1q Pass 2q 1 Pass
2♠ Pass 2NT Pass
3NT Pass Pass ?
1. 10+ points

Pass. South knows that E-W have a runout, and East is likely to
diagnose the weakness and run to 4 q if South doubles. The opponents
have apparently bid a marginal game from the wrong side, so South
should take his profit. However if South weren’t on lead (say West
had rebid 2NT), then the double would be correct. The odds shift
completely. While E-W might still run, the double that alerts North
to make an unusual lead greatly increases South’s chances of defeating
the contract.

348
DEFENSIVE BIDDING REVIEW PROBLEMS

Problem 9

N-S vul., South holds ♠ AQJ965 r 7 q KJ2 ♣ QJ6

West North East South

— — — 1♠
2r 2♠ 4r ?

349
 MATCHPOINTS

Problem 9 Solution

N-S vul., South holds ♠ AQJ965 r 7 q KJ2 ♣ QJ6

West North East South

— — — 1♠
2r 2♠ 4r ?

4 ♠. It looks like a pretty marginal sacrifice. South has plenty of defense

so he can’t be sure that 4 r is making, and 4 ♠ could easily go down two
if things lie badly. Nevertheless, 4 ♠ is the percentage bid because it
has all the plus factors in its favor. First of all, it is not out of the
question that 4 ♠ will make. Secondly, even if 4 ♠ is going down E-W
may not be sure of this and may fail to double or better yet take a 5 r
save themselves. There are so many ways for the 4 ♠ bid to work out
well that it must be made.

350
DEFENSIVE BIDDING REVIEW PROBLEMS

Problem 10

N-S vul., South holds ♠ KJ85 r A74 q A5 ♣ 8642

West North East South

— — 1♠ Pass
2♠ Pass 4♠ ?

351
 MATCHPOINTS

Problem 10 Solution

N-S vul., South holds ♠ KJ85 r A74 q A5 ♣ 8642

West North East South

— — 1♠ Pass
2♠ Pass 4♠ ?

Double. This sounds like a standard auction to get to a normal contract,

and South figures to beat 4 ♠ more often than not even after he tips
off his trump holding. If dummy puts down ♠ Q9x and declarer picks
up the trump suit for 1 loser which he wouldn’t have done without the
double, that is just too bad. The double will be right more often than not,
and that is all that counts if the field figures to be in the same contract.

352
 Part V

The Play
 OPENING LEADS

Opening lead strategy is not much different at matchpoints than at

IMPs. The opening leader has little enough information to go on, so
he usually just takes his best shot. There are certain situations that do
call for special tactics at matchpoints, and the winning player learns to
recognize them and act accordingly.
The primary difference in the play between matchpoints and IMPs
is the importance of overtricks. Overtricks are relatively unimportant at
IMPs, so you go all out to beat the contract. At matchpoints, this isn’t
the case. Allowing declarer to make an overtrick by your opening lead
can be just as costly as allowing him to make the contract. The idea
is to determine which hands call for the all-out effort and which hands
require caution to hold in the overtricks.
There are three types of contracts the opponents might arrive at.
First, there are normal contracts, likely to be reached at most tables.
Against this sort of contract overtricks are just as important as setting
the contract. Second, there are abnormal contracts, such as a marginal
slam or a doubled contract. Against these setting the contract has the
highest priority, so you should make what you think is the best lead to
defeat the contract without regard to possible overtricks. Third, there
are alternative contracts. These are contracts that may not be popular,
but that can compete with other possible contracts. A 4–3 major-suit
game or 3NT when a good major-suit fit is available are examples of
alternative contracts. Another is a contract that is likely to be played
from the other side at most tables. These are the trickiest of all. It may
be necessary to defeat the contract to get any matchpoints. On the other
hand, an overtrick can swing a full matchpoint against other tables; for
example it might take you from −630 to −660 when others your way
are −650. It is important to try to judge from the auction which type of
contract you are up against.
A normal contract is one that figures to be reached at most tables.
It can usually be recognized by the simplicity and directness of the
auction. If the bidding is something like:
 MATCHPOINTS

West North East South

— — — 1NT
Pass 3NT Pass Pass
Pass
or:
West North East South
— — — 1♠
Pass 2♠ Pass 4♠
Pass Pass Pass

you can be pretty confident that the auction will be echoed around
the room. More complex auctions can also lead to obviously normal
contracts. For example:

West North East South

— — — 1r
Pass 2♣ Pass 2q
Pass 2♠ Pass 2NT
Pass 3NT Pass Pass
Pass

The auction may be different at other tables, but the final contract is
likely to be the same simply because the opponents don’t have any place
else to go. Of course it might be played from the other side, which can
make a difference, but 3NT it will be around the room.
If the contract is the same at other tables you gain 1⁄2 matchpoint
against every table for each trick your opening lead gains, and lose
1⁄ matchpoint for each trick your opening lead loses, assuming that the
2

other table results are within one trick of yours. This is true whether the
contract or an overtrick is at stake. Consequently, your goal on opening
lead is not necessarily to set the contract, but to take as many tricks as
you can. This often means a more passive lead than usual. For example:

356
 OPENING LEADS

None vul., West holds ♠ KJ8 r 1092 q A63 ♣ 9543

West North East South

— — — 1NT
Pass 3NT Pass Pass
Pass

At IMPs you might get brilliant and try the ♠ J. The best chance to beat
the contract appears to be to find partner with length and strength in
spades, and overtricks are relatively unimportant. At matchpoints you
can’t afford this luxury. A spade lead is more likely to cost a trick
than to gain one, and each overtrick will swing several matchpoints. A
passive heart or club lead is recommended.

Both vul., West holds ♠ Q83 r J107 q J109 ♣ KJ72

West North East South

— — — 1r
Pass 3r 1 Pass 4q
Pass 4r Pass Pass
Pass
1. Limit raise

At IMPs a good case can be made for leading a low club. The idea
is to hope to cash three club tricks and get an uppercut on the fourth
round. With the opponents trying for slam it will take a lot of luck to
beat 4 r, but the possibility exists. This may be a winner at IMPs, but at
matchpoints forget the brilliance and lead the normal q J. It is extremely
improbable that you can set 4 r whatever you lead, and the club lead is
much more likely to cost the defense a trick than to gain one, for partner
can’t hold much in high cards. Note that even though 4 r might not be a
universal contract since some pairs may reach slam, you should defend
as though it is a normal contract. Your opening lead won’t affect the
pairs in slam. They will beat you or lose to you at their table whatever
you do. Consequently, your only real competitors are those defending
the same 4 r contract, so lead accordingly.

♠ rq ♣ ♠ rq ♣

357
 MATCHPOINTS

It is not always correct to make a passive lead against a normal contract

at matchpoints. If the bidding indicates that an aggressive lead is called
for, go ahead and make it.

E-W vul., West holds ♠ 1087 r A53 q 10983 ♣ KJ3

West North East South

— — — 1♠
Pass 2r Pass 2♠
Pass 4♠ Pass Pass
Pass

It appears as though the major suits lie favorably for declarer, and
dummy’s hearts will soon be established for minor-suit discards. This
is not the time to be passive. A club lead is correct both at IMPs and
matchpoints. In addition to being the best lead to set the contract, it is
also the best lead to hold down the overtricks by establishing club tricks
before West’s r A is knocked out. If the club lead blows a club trick
there is a good chance that declarer would have been able to discard the
potential club loser on dummy’s hearts anyway, so the lead might not
cost even if it gives declarer a free finesse.

♠ rq ♣ ♠ rq ♣

Another type of lead that is common at matchpoints is laying down

an ace for a possible cashout. Often the cashout will merely hold in
the overtricks, unimportant at IMPs but critical at matchpoints. For
example:

358
 OPENING LEADS

N-S vul., West holds ♠ AQ10 r 932 q J108 ♣ 9743

West North East South

— — — 1r
Pass 3♣ 1 Pass 3r
Pass 4r Pass Pass
Pass
1. Game-forcing

With this strong auction and West’s weakness in hearts and clubs, it
is likely that the only tricks for the defense will be in diamonds and
spades. At IMPs the q J lead is clear. It is hard to imagine beating the
contract if partner doesn’t own the q A, so this lead gives the defense
the option of cashing diamond or spade tricks, whichever are available.
At matchpoints, however, the ♠ A is a better shot. East is just as likely
to hold the ♠ K as the q A, and if he does, failing to cash the spades will
cost several overtricks. Even if East doesn’t hold the ♠ K the cashout
could well be right. Dummy may have the ♠ K in which case no harm
and possibly some good is done by the lead, and if East has no entry at
all (perhaps his high card is the onside r K), then the spade lead may
be necessary to hold declarer to 6, which is quite important since many
other N-S pairs also figure to arrive in 4 r. The spade lead costs only
when South owns the ♠ K and East has an entry. In all other cases the
lead either gains or breaks even. Consequently, the spade lead will be
right more often than not, even though it virtually gives up on setting
the contract.

♠ rq ♣ ♠ rq ♣

When the opponents arrive in an abnormal contract, it is usually right

to go all out to beat them. Overtricks are relatively unimportant. The
exception to this is the major vs. notrump decisions, where overtricks
are extremely important.

359
 MATCHPOINTS

Both vul., West holds ♠ A975 r J103 q 9865 ♣ KJ

West North East South

— 1♣ Pass 1q
Pass 3♣ Pass 3q
Pass 3♠ Pass 4NT ∗
Pass 5r ∗ Pass 6NT
Pass Pass Pass

Not a very impressive auction, but it is likely that they have hit the
jackpot. Your club holding is a complete disaster for the defense. At
IMPs I would try a low spade lead, as this seems to be the best shot
if not the only one. The opponents’ contract appears unusual enough
that the same lead should be made matchpoints. The overtrick probably
won’t matter much, but setting the contract will be a full-board swing
if the field stays out of slam or plays 6 q .
Often it is a good idea to determine why the opponents have reached
the unusual contract when a better alternative contract seems more
likely. You can base your opening lead on the assumption that the
opponents are right, for if they are wrong your opening lead won’t
matter. For example:

Both vul., West holds ♠ AQ96 r 9654 q 52 ♣ QJ9

West North East South

— — — 1q
Pass 1r Pass 3q
Pass 4q Pass 5q
Pass Pass Pass

Not a particularly scientific auction. You would think they would try
a little harder to get to 3NT. Perhaps they know what they are doing,
perhaps not, but it is clear that if N-S hold the ♠ K 3NT is likely to score
well above 5 q , and the field will be there. Therefore, you should base
your opening lead on the assumption that N-S don’t hold the ♠ K, for
if they do you probably get a good board regardless of what you lead.
The ♠ A is the best lead, with the idea of cashing as many spade tricks
as possible. If the lead blows a trick, it won’t cost many matchpoints.

360
 OPENING LEADS

♠ rq ♣ ♠ rq ♣

The most difficult contracts to lead against are those that may not be
normal but are competitive with other contracts. The most common of
these is 3NT when the field figures to be in four of a major, or vice
versa. Every trick can swing a full matchpoint against another table,
and you don’t really know what your goal is. What you must do is look
at your hand, try to determine if the opponents have done well or not in
their choice of contracts, and choose your opening lead accordingly.

E-W vul., West holds ♠ QJ97 r 1094 q KJ43 ♣ 85

West North East South

— — — 1NT
Pass 2q 1 Pass 2♠
Pass 3♠ Pass 3NT
Pass Pass Pass
1. Forcing Stayman

Just as the opponents were headed for your spade suit, they both
guessed to play notrump. From your point of view it looks like they
guessed right. To make matters worse, most of the field will probably go
with the 4–4 major-suit fit. Notrump will have to take one trick less than
spades for spades to be the right contract for N-S, which seems unlikely.
However, you must base your opening lead on the assumption that this
is the case, for you will be headed for a bad result regardless of your
lead if notrump always scores at least as many tricks as spades. How
can spades be better? Only if the opponents have a weak side suit, and
you can see that the most likely weak suit is diamonds. Consequently,
you should make the aggressive lead of a diamond. If partner doesn’t
have a diamond card the lead may cost a trick, but if this is the case
you are probably headed for a bad score whatever you lead. If partner
does have a diamond card it may be necessary to lead a diamond now,
and the lead may be the difference between getting one more trick on
defense than those defending spades and getting the same number of
tricks, so the diamond lead could swing the full board. Of course, if the
opponents had landed in 4 ♠ you wouldn’t touch a diamond lead with
a ten-foot pole. Then you are happy with their contract, so you avoid
blowing a trick.
361
 MATCHPOINTS

Both vul., West holds ♠ Q3 r A98 q QJ754 ♣ 754

West North East South

— — — 1NT
Pass 2q 1 Pass 2r
Pass 2NT Pass 3r
Pass 4r Pass Pass
Pass
1. Forcing Stayman

In spite of the opening 1NT bid, the opponents had the tools to find a
5–3 heart fit. At many other tables N-S pairs will be playing standard
Stayman and Transfer bids, and the auction is likely to go 1NT-2 ♣;
2 r-3NT. Against 3NT, other Wests will lead a diamond. If East has a
diamond honor this lead will be a success, and you will not do well
defending 4 r regardless of what you lead, for 3NT is likely to go
down or be held to 9 tricks if declarer doesn’t have time to use his
heart suit. You must assume that the diamond lead against 3NT will
not be successful in order for you to have a chance. Therefore, stay
away from the diamond lead against 4 r, because based on the necessary
assumption you have made, it will not work well. Instead lead a passive
club, wait for your tricks, and hope that the notrumpers will be able to
score the same number of tricks as your opponents take in hearts.

None vul., West holds ♠ KJ954 r 962 q K2 ♣ A73

West North East South

— — — 1NT 1
Pass 2q 2 Pass 2r
Pass 3q Pass 3r
Pass 4r Pass Pass
Pass
1. 12–14 HCP
2. Transfer

Normally you might be inclined to lead a spade on this kind of auction.

If you don’t establish a spade trick fast it may go away on a minor-suit
winner in dummy, so even if the spade lead rides to the ♠ AQ it may
not cost a trick. The problem with a spade lead on this hand is that you

362
 OPENING LEADS

may have been fixed by the bidding. At most tables South will open
one of a minor, West will overcall in spades, and North will be declarer
in 4 r with a spade lead from the other side. Consequently, if you lead a
spade you get an average when it is right and bottom when it is wrong,
as you are competing with spade leads from your partner’s hand. It
makes more sense to assume that the spade lead is not necessary, so
lead a trump. This could conceivably do some good if South has to ruff
diamonds in the short hand, and at least you get away from the average-
or-worse position you would be in if you tried to compete with the other
spade leaders.

E-W vul., West holds ♠ 62 r K10653 q Q2 ♣ K1085

West North East South

— — — 1q
Pass 1r Pass 1NT 1
Pass 3NT Pass Pass
Pass
1. 15–17 HCP

Your opponents are playing weak notrumps, which accounts for the
auction. The contract appears normal, but the information available to
the opening leader is not. It is easy to project the likely auction at other
tables: 1NT-2 ♣; 2 q -3NT. On this auction most Wests will lead a heart.
You know that dummy will show up with four hearts, so a heart lead
will probably be bad for the defense. Having been warned off the heart
lead, should you try a club or a spade? I think you should lead a spade.
The important point is that you are already ahead of the field by not
blowing a trick with a heart lead, so you should play completely safe
rather than go for the kill. If the club lead is right you will probably still
do well for having avoided the heart lead, but if the club lead gives up a
trick you are back down to average.

363
 MATCHPOINTS

Both vul., West holds ♠ 54 r 8762 q KQ8 ♣ A742

West North East South

— — — 1r
Pass 2♠ 1 Pass 3♣
Pass 4♠ Pass 4NT
Pass 5r Pass 6NT
Pass Pass Pass
1. Game-forcing

This is a very unusual problem. How could you not lead the q K? A little
thought will show that the lead might be very costly. North has shown
a long solid spade suit, and South’s hearts are likely to run considering
your anemic heart holding. It is not hard to imagine that N-S have
7 spades, 5 hearts, and 1 diamond off the top, so it may be essential to
cash the ♣ A. An important consideration is that a popular contract is
likely to be 6 ♠ by the North hand, and other Easts won’t know which
minor to lead. Consequently there may be several −1460s your way, so
there will be a big swing between −1440 and −1470. It is not often that
you virtually give up on trying to beat a slam even at matchpoints, but
this is one time when the overtrick and the alternative possible contract
really make a big difference. The full hand:

♠ AKQJ963
r A9
Both vul.
q 643
♣ 6
♠ 54 ♠ 87
N
r 8762 r 10 4
W E
q KQ8 q J 10 9 5 2
S
♣ A742 ♣ J953

♠ 10 2
r KQJ53
q A7
♣ K Q 10 8

364
 OPENING LEADS

The slam is easy to reach, and many pairs played 6 ♠ making 7 after
East’s natural q J lead. The swing between cashing the ♣ A and losing
it against 6NT was from a well-above-average score to a tie for bottom.

365
 DECLARER PLAY

In my younger days I once asked an expert how a hand should be played

at matchpoints. His simple but illuminating answer was: “For all the
tricks you can take.” This is not always correct, but it is not far from the
truth. It is often even correct to risk your contract for the possibility of
extra overtricks.
When the dummy comes down the first thing a declarer should do
at matchpoints is assess his contract and the effect the opening lead has
on his chances. If he is in a routine contract with a normal opening
lead, he naturally plays to maximize his expected number of tricks
taken. If the lead was unusually favorable, he will play conservatively
to attempt to retain his advantage. If the lead was unusually devastating,
he may make very daring plays to get back even with the field. If there
are alternative possible contracts, he should estimate the likely results
in those contracts and take whatever action is necessary to beat those
results if possible.
 MATCHPOINTS

♠ AQ64
r A J 10 9
Both vul.
q 874
♣ Q2

♠ K J 10 8 2
r 43
q AK5
♣ K98

West North East South

— — — 1♠
Pass 2NT 1 Pass 3♠ 2
Pass 4♠ Pass Pass
Pass
1. Forcing raise
2. Good balanced minimum

Opening lead: q Q.

You win the q Q lead, draw two rounds of trumps ending in your hand
(they split 2–2), and lead a heart to the r J which loses. East returns a
diamond, you win, and West plays the q 10. Now what?
Let’s assess your position. The 4 ♠ contract was easy to bid, and
the field should all get there from the South side one way or another.
Furthermore, it appears as though West started with q QJ10, so the
lead was normal, as was the diamond continuation. Therefore, every
other South player will face this exact problem. There are two possible
lines of play. One is to play West for the missing heart honor and take
another finesse. If the finesse succeeds South will make 5, but if it loses
he will go down unless East has the ♣ A and no more diamonds. The
alternative line is to play East for the missing heart honor by leading a
heart to the r A and taking a ruffing finesse. This will make 5 if East
started with both honors, and guarantees the contract because even if the
ruffing finesse loses South will have pitched his diamond. The second
line of play is obviously best at IMPs, but what about at matchpoints?
Restricted Choice tells us that West is a 2-to-1 favorite to have the
missing heart honor. If you don’t understand Restricted Choice, look
at it in the following way: Initially, there was a 25% chance that West

368
 DECLARER PLAY

had both honors, a 25% chance that East had both honors, and a 50%
chance that the honors were split. In other words, it was initially twice
as likely that the honors were split as that East had both honors. The
fact that we have lost to an honor in the East hand (it doesn’t matter
which one) doesn’t change this figure. All we know is that West didn’t
start with both honors.
Now, let’s look at the matchpoint cost analysis. Whichever line of
play we take, if it is wrong we will cost ourselves 1⁄2 matchpoint to every
other table regardless of what they do. If the other table goes right we
get a loss instead of a tie, while if the other table is wrong we get a tie
instead of a win. Consequently, it is correct to make the play that is
most likely to be right. In this case this means taking the second heart
finesse, despite the risk to the contract.

♠ rq ♣ ♠ rq ♣

Even if you are faced with a mildly unfriendly lead, it is usually correct
to make the percentage play for as many tricks as possible if you are
in a normal contract. On the previous hand, suppose the opening lead
has been the q 2 to East’s q Q, and East had returned the q 10. This is
not good for South. West has apparently found the killing lead from
q Jxxx, and the lead certainly will not be found at every other table. If
a diamond is not led, South can take both heart finesses without risking
his contract. It might seem as though you should therefore take the anti-
percentage ruffing finesse and go for a top or a bottom, since you can at
best tie those tables that don’t get a diamond lead if you play the same
way they do and the second finesse wins, while you will lose to them
(down 1 while they make 4), if the second finesse loses. Let’s analyze
it:
We already know that the double finesse is best against tables that
receive the diamond lead. Against tables that don’t get the diamond lead
you lose 1⁄2 matchpoint if you take the ruffing finesse and are wrong
(from a tie to a loss), while you gain a full matchpoint if the ruffing
finesse is right (you make 5 instead of going down 1, while they make
exactly 4). So you are getting a 2-to-1 matchpoint payoff against tables
not getting the diamond lead if you take the ruffing finesse, but giving
2-to-1 odds by making the anti-percentage play, so it’s exactly break-
even. One-third of the time you gain one matchpoint, two-thirds of the
369
 MATCHPOINTS

time you lose 1⁄2 matchpoints. Consequently, the double finesse is still
the best play at matchpoints, because some other tables will receive the
same diamond lead.
Let’s change the hand slightly, sticking with the same theme:

♠ AQ64
r A Q 10 2
Both vul.
q 874
♣ Q2

♠ K J 10 8 2
r 43
q AK5
♣ K98

Again, you are in 4 ♠ with the q Q lead. You win, draw trumps in
two rounds, and lead a heart to the r 10 which loses to the r J. Back
comes a diamond. Do you take another heart finesse or play to ruff out
r Kxx? Restricted Choice is not applicable here—East would always
play the r J from the r KJ, so the r K is equally likely to be with either
defender. If you take the finesse, a 50% play, you will lose 1⁄2 matchpoint
to every other table when it fails. If it wins you will not necessarily gain
1⁄ matchpoint, since if West has r Kxx the play of r A and ruff a heart
2

works as well as the finesse. Suppose you play the r A and ruff a small
one. You will surely gain 1⁄2 matchpoint from every other table when
the r K is offside, and if it is onside you will still not lose if West has
r Kxx. Thus the finesse costs half the time, while the rest of the time
it either gains or breaks even, while rejecting the finesse gains half the
time, and either costs or breaks even the other half. So the r A is the
better play.
How about with the q 2 lead? Now the lead is not necessarily
normal. Those declarers not receiving a diamond lead will certainly
take two heart finesses, the percentage play in hearts, as they are still
safe even if both finesses lose. What should you do when your r 10
loses to the r J and the q 10 comes back? If you finesse you will tie
declarers who did not get a diamond lead half the time and lose to them
half the time. Suppose you play a heart to the r A and ruff a heart
instead. You will then tie or beat those other declarers half the time—

370
 DECLARER PLAY

whenever East has the r K or West has r Kxx—and lose to them only
if West has r Kxxx(x) of hearts. Thus r A and ruff a heart is the better
play.
The preceding analysis is quite complex, and it would be difficult
to go through all these possibilities at the table. Still, this hand does
shed some light on when declarer should make his best play for the
most tricks and when he should hedge and play moderately safely. In
general you should look for the play that has the highest matchpoint
expectation, even though this isn’t necessarily the best play for the
most tricks. For example, in the hand where dummy had r AQ10x,
the second finesse was the best play for the most tricks, but playing for
the drop was the best matchpoint play.

♠ rq ♣ ♠ rq ♣

Should you ever take an anti-percentage play in a normal contract?

Usually not, but occasionally it may be necessary if you are badly fixed
by the opening lead. For example:

371
 MATCHPOINTS

♠ KQ82
r A86
E-W vul.
q KJ9
♣ J 10 4

♠ A J 10 9 4
r K3
q A54
♣ Q53

West North East South

— — — 1♠
Pass 2NT 1 Pass 3♠ 2
Pass 4♠ Pass Pass
Pass
1. Forcing raise
2. Good balanced minimum

Opening lead: ♣ K.

The defense starts ♣ K, club to East’s ♣ A, and a third club, which

West ruffs. He now exits with a trump. How should you handle the
diamond suit? The normal play is to lead a diamond to the q J, which
wins whenever West has the q Q. The other possibility is the “backward
finesse” of swinging the q J off the dummy, planning on then finessing
the q 9 if East covers the q J. This requires both East to hold the q Q
and West to hold the q 10 to be successful, hence it will work only half
as often as the straightforward finesse.
Now for the matchpoint cost analysis. Against other tables where
the ♣ K was led the backward finesse gains 1⁄2 matchpoint 25% of the
time and loses 1⁄2 matchpoint 50% of the time, for an average loss of
1⁄ matchpoint 25% of the time. When East holds both the q Q and q 10,
2

25% of the time, the play breaks even. What about tables that receive
another lead? These tables don’t lose the club ruff, so the backward
finesse costs nothing to these tables when it is wrong since they will
make 5 (via the straightforward finesse that they certainly will take),
and this is out of your reach. However, the backward finesse does gain
1⁄ matchpoint the 25% of the time it works, for they now lose a diamond
2

trick and you don’t, so you go from a loss to a tie. Consequently, if you

372
 DECLARER PLAY

think that the majority of Souths will not get the killing club lead, a very
reasonable assumption, then the backward finesse is the best play. The
idea is that if an unusual lead has definitely cost you a trick it may be
correct to take an anti-percentage play to try to get that trick back, for
taking the same play as the rest of the field will leave you 1 trick behind
them whether the play works or not.

♠ rq ♣ ♠ rq ♣

Safety plays have their place in matchpoints, but it is at the bottom of

the totem pole. Some safety plays can only gain, and these should be
made routinely. For example:

North: ♠ Axxx
South: ♠ KQ9xx

If there are no entry problems, the ♠ A should be played first at any form
of scoring to guard against ♠ J10xx in the East hand. However, most
safety plays involve losing a trick that might not have to be lost in order
to ensure the contract against a very bad split. This is fine for IMPs,
where making the contract has the highest priority, but at matchpoints
these plays will result in far more bottoms than tops in normal contracts.
For example:

373
 MATCHPOINTS

♠ 983
r 93
N-S vul.
q 32
♣ AKQ654

♠ AK4
r AK4
q QJ854
♣ 73

West North East South

— — — 1NT
Pass 3NT Pass Pass
Pass

Opening lead: r Q.

At IMPs it would be a routine safety play to duck the first round of

clubs. This assures the contract against a 4–1 club split. Playing
matchpoints, this safety play would be the height of folly. The contract
is trivial to reach, and the opening lead appears normal. Ducking a club
will cost a trick when the clubs split 3–2, and gain at least 1 trick when
they split 4–1. Since a 3–2 split is far more likely than a 4–1 split,
ducking will be wrong more often than it will be right, so you should
ignore the safety play and try to take as many tricks as possible in this
normal contract.
In abnormal contracts, where just making the contract is important,
safety plays can be correct. If you are doubled, overtricks are
virtually inconsequential because making any doubled contract usually
guarantees a top or near top. The same is true about a hard-to-bid slam.
For example:

374
 DECLARER PLAY

♠ AK6
r K432
E-W vul.
q A964
♣ 75

♠ 943
r AQJ
q KJ74
♣ AQ6

West North East South

— — — 1q
Pass 1r Pass 1NT 1
Pass 2♣ 2 Pass 3r
Pass 3♠ Pass 4♣
Pass 4q Pass 4r
Pass 4♠ Pass 5♣
Pass 6q Pass Pass
Pass
1. 15–17 HCP
2. New Minor Forcing

Opening lead: ♣ J.

You went through a rather involved sequence to get to a marginal slam,

which looks a lot better after the ♣ J opening lead. You can be pretty
sure that most of the field won’t be anywhere near this contract after a
strong notrump opening, and the overbidders probably won’t find their
diamond fit and will get to 6NT, which you can’t outscore if it makes.
Therefore, you should take the standard safety play in the trump suit of
q K and then low to the q 9 (unless West shows out). The potential lost
overtrick is virtually meaningless, while the whole board could ride on
making this contract, and this safety play guarantees the contract against
all but the worst of splits.
Another type of hand that calls for a safety play to ensure the
contract for an entirely different reason is one where you have arrived
at an abnormal low-scoring contract. Since you can’t compete with the
normal higher-scoring contracts if they make, you have to assume that

375
 MATCHPOINTS

they will go down, so you must attempt to ensure your contract to avoid
tying other tables at down 1. For example:

♠ A2
r KQ9
N-S vul.
q A843
♣ A943

♠ 987
r AJ7
q K7
♣ K 10 7 6 2

West North East South

— 1q 1♠ 2♣
Pass 3♣ Pass 3r
Pass 3♠ Pass 4♣
Pass 5♣ Pass Pass
Pass

Opening lead: ♠ 3.

Don’t ask me what partner has against notrump, but here you are in
5 ♣. 6 ♣ makes on a club pickup, and 3NT will make many overtricks
if the club suit comes in. Both of these contracts will be down 1 if the
clubs don’t behave since the spades appear to be 5–3 on the bidding and
opening lead. Most of the field will play 3NT, and those that don’t will
probably get to 6 ♣. You can’t compete with either of these contracts
if they make, so you must assume that they don’t. Therefore, it is
important to ensure making 5 ♣, so take the safety play of a low club
to the ♣ 10 at trick 2 to guard against a 4–0 club split. A lost overtrick
figures to cost almost nothing, but going down instead of making can
cost 1⁄2 matchpoint compared to other tables that are down 1 in the more
normal contracts.

♠ rq ♣ ♠ rq ♣

Sacrifices can lead to interesting play decisions. Obviously your target

is to go down less than the value of the contract against which you are
376
 DECLARER PLAY

sacrificing. You can assume that the field won’t all take the sacrifice,
and it is important to get that full matchpoint swing against tables at
which the save isn’t taken.
♠ 53
r A J 10 6
E-W vul.
q K 10 9 5
♣ 982

♠ 2
r 52
q QJ87432
♣ A63

West North East South

— — 1♠ 3q
4♠ 5q Double Pass
Pass Pass

Opening lead: ♠ K.

West leads the ♠ K, and continues spades, which you ruff. You try a
heart to the r J and East’s r K, and East returns a club. Even though the
second heart finesse is a 2-to-1 favorite, it would be silly to try it. You
figure to get a good board for −500, since not everybody will take the
save and E-W may be able to make 5 ♠ anyway. It would not be a good
idea to risk a full matchpoint against tables where E-W are playing
and making a game in spades in order to perhaps gain 1⁄2 matchpoint
against tables where the contract is the same as yours. With neither
vulnerable, however, the second heart finesse would be clear-cut. Now
your target is 9 tricks, not 8, so you must try it. Even though it requires
two defensive errors on West’s part (not shifting to clubs and not playing
the r Q on the first heart lead), you must assume that these errors have
been made. With neither vulnerable there won’t be much matchpoint
difference between −500 and −800, but there will be a huge difference
between −300 and −500.
When you take a sacrifice, you will always get a bad result if your
save is a phantom. Consequently, it is necessary to assume that the
opponents’ contract is making when you are trying to go for less than

377
 MATCHPOINTS

the value of their game. It makes no sense to play for a lie of the cards
on which the enemy contract could not have made. For example:

♠ KJ52
r 92
E-W vul.
q 6542
♣ 932

♠ A7643
r 74
q A93
♣ A65

West North East South

— — 1r 1♠
4r 4♠ Double Pass
Pass Pass

Opening lead: r Q.

Neither you nor your partner took a totally unreasonable action, yet
the final contract doesn’t look too good. The defense takes the first
two rounds of hearts and shifts to a club. You have 6 losers outside of
trumps, so you must avoid losing a trump trick to get out for less than
the value of their potential game. Both follow when you lead the ace
of trumps, and West follows small on the second round. Do you finesse
or play for the drop? On the bidding the drop seems better, since East
doubled. However, if the ♠ Q drops you still get a very bad board. The
opponents will have 4 losers in 4 r if the spades are 2–2, so you must
hope that they are 3–1. If the finesse works you will be −500 with 4 r
making for a good sacrifice, while if the finesse loses your −800 will
only be a slightly rounder zero than −500 since the save was a phantom.
Therefore, the finesse is clearly correct.

♠ rq ♣ ♠ rq ♣

The most difficult contracts to play at matchpoints are those that may
be in competition with alternative contracts. If there is another likely
contract that may or may not beat yours, it is very important to assess
378
 DECLARER PLAY

the likely results of the other contract depending on the lie of the
defenders’ cards and to plan your play accordingly.

♠ A J 10 4 3
r A65
None vul.
q 52
♣ A62

♠ K7
r KQ932
q A43
♣ K 10 5

West North East South

— — — 1r
Pass 1♠ Pass 1NT
Pass 2♣ 1 Pass 2r
Pass 4r Pass Pass
Pass
1. Artificial and forcing

Opening lead: q Q.

West leads the q Q, which you duck. He continues with a diamond

to the q K and q A, and you ruff the third round of diamonds low in
dummy, East following. You draw trumps pitching a club from dummy;
East follows three times while West pitches a diamond on the third
trump. Now, how do you tackle the spades?
At this point, you might be wishing you were in 6 r. That contract is
not relevant to you. Either it makes or it doesn’t, but those in slam will
decide their fate at their table and nothing you can do will affect your
result against them, so you might as well assume anybody in slam is
playing in another event. How about 3NT? This may well be a popular
contract. Many Souths will open 1NT on your cards, after which the
5–3 heart fit probably won’t be found. How will 3NT fare after the
natural diamond lead? Eventually, declarer will have to take a spade
finesse. If it loses, the notrumpers won’t take more than 10 tricks,
while if it wins they will take at least 11. This is the clue for how
to play the spades in 4 r. If the spades are 3–3 with the ♠ Q offside

379
 MATCHPOINTS

South won’t mind too much losing a spade trick, for he will still score
11 tricks to beat the notrumpers. The danger situation is when West has
♠ Qxxx (if he has ♠ Qxx all plays work). Now the notrumpers will take
exactly 11 tricks, so it is essential to make 12 tricks in hearts. Another
way to look at this situation is to realize that South is a trick ahead of
the notrumpers due to the diamond ruff in dummy. Consequently, if
South plays the spades the same way they will be played in 3NT he
will remain a trick ahead regardless of whether the spade finesse wins
or loses. Therefore, South should play ♠ K and a spade to the ♠ J.
The play of this hand in 3NT is also interesting. How many times
should South hold up in diamonds? There is a strong argument for not
holding up at all, and hoping both major suits behave for 13 tricks. As
long as South doesn’t know how the diamonds are splitting he might as
well go whole hog and assume the best. If it works he is at least sure of
beating 4 r contracts regardless of how many tricks they make.

♠ 742
r KJ62
E-W vul.
q K 10 6 3
♣ 32

♠ A 10 8
r A7
q QJ852
♣ Q 10 8

West North East South

— — — 1q
Pass 1r Pass 1NT
Pass 2q Pass Pass
Pass

Opening lead: ♠ Q.

What a nice comfortable contract. It seems as though only a madman

would take the heart finesse for an overtrick and risk going down. Yet,
careful analysis shows that at matchpoints playing for the overtrick is
the only sensible line of play.

380
 DECLARER PLAY

What are some other possible results? Your opponents were silent,
but not all E-W pairs figure to keep quiet holding half the deck. How
they will fare depends a lot on where the ♣ J and r Q are, but it is not
difficult to envision some N-S pairs collecting +100 against 2 ♠ or 3 ♣.
If this is the case, there will be a big difference between +90 and +110,
for you will swing a full matchpoint against every N-S pair who is +100.
N-S pairs who are pushed to 3 q will be forced to play to make it, so
by taking the heart finesse you tie them if it wins and beat them if it
loses, as you will be −50 to their −100. N-S pairs playing notrump
will either make one or go down a bundle, depending on where the ♣ J
is. If they go down you will beat them regardless, while if they make
one you are as likely to gain as to lose to them by taking the heart
finesse. The point is that there is much more activity between +90 and
+110 on this hand than there is between −50 and +90. No conceivable
score falls completely inside the latter range, so the finesse can cost at
most 1⁄2 matchpoint to any table if it is wrong, while it can gain a full
matchpoint against +100 if it is right. If E-W were not vulnerable, it
would be correct to play safe for the contract. Now +50 falls between
−50 and +90, so the finesse will gain less if it is right and lose far more
if it is wrong.

381
 MATCHPOINTS

♠ A86
r KQ865
N-S vul.
q J63
♣ Q9

♠ KQJ9
r A74
q A5
♣ K842

West North East South

— — — 1NT
Pass 2♣ ∗ Pass 2♠
Pass 3NT Pass Pass
Pass

Opening lead: q 2.

One can sympathize with North’s decision not to search for a 5–3 heart
fit since he holds extra strength and values in every suit, but it doesn’t
appear to be right on this hand. You try the q J at trick 1, but East covers
with the q K and you win the q A. On the r A West plays small and East
plays the r 9, and West plays another small heart when you continue the
suit. Now what?
Playing for the 3–2 split is the percentage play in the heart suit.
The finesse of the r 8 wins only when East started with the stiff r 9,
while going up wins when East started with r J9, r 109, or r J109.
Admittedly the last two holdings are less likely because East might have
played a different card than the r 9 (Restricted Choice again), but the
drop is still the percentage play in the suit. However, at matchpoints the
finesse is the percentage play on the hand!
When partner bid 3NT, he was betting that you could take the same
number of tricks in notrump as in a possible 5–3 heart fit. You must try
to live up to his expectation despite the unfavorable lead. At hearts N-S
can easily take 11 tricks if the hearts split, for they have time to set up
the club trick while you don’t. Consequently you should go ahead and
take the double heart finesse, even though you will be down in a cold
contract (if the diamonds are 4–4) if it fails. If the hearts actually are

382
 DECLARER PLAY

4–1 you will now come to 10 tricks, which is all those playing in 4 r
will take since they will play the hearts normally. Besides, what about
those in 3NT who don’t get a diamond lead? If the hearts split 3–2
they have time to set up a club trick and make 11 tricks, while on a 4–1
heart split they may be held to 9 tricks if they play the hearts normally.
Therefore the double heart finesse gains a full matchpoint against both
4 r and 3NT without a diamond lead if it is right, and loses nothing to
these contracts if it is wrong. This analysis shows that the finesse has
so much more to gain than to lose that it is clearly the best play even
though it is very anti-percentage for the heart suit in isolation.
Incidentally, against weak opposition it might be a good idea to try
to steal a club at trick 2 on this hand. If you get away with this you
will beat the 4 r players if the hearts split. Against good opponents this
should not work. If West has, say, q Q9xx he will be able to deduce
from your play of the q J at trick 1 that you don’t have the q 10, for if
you did you would have played low to ensure two diamond tricks. With
a good player in the West seat, the straightforward approach is best.

383
 DEFENSE

Defense is difficult enough at any form of scoring, but at matchpoints

there are even more problems than usual. The difficulty is that you don’t
always know what your goal is. At IMPs you know your goal is to set
the contract, so you just assume that declarer has whatever hand he
needs for the contract to be defeated, almost regardless of how unlikely
that hand actually is, and defend accordingly. If you are wrong the cost
will only be an unimportant overtrick or two. At matchpoints you can’t
afford to take this reckless sort of approach. Those overtricks can be
very important. On the other hand, it is a catastrophe to let declarer
make a contract you could have beaten because you were trying to hold
in the overtricks. It is often quite a dilemma.

♠ A Q 10 9
r Q 10 9
N-S vul.
q 87653
♣ 5
♠ 8643
r J7
q KQ
♣ A 10 8 4 2

West North East South

— — — 1NT
Pass 2♣ ∗ Pass 2q ∗
Pass 2NT Pass 3NT
Pass Pass Pass

Opening lead: ♣ 4.

Your club lead goes to partner’s ♣ Q and declarer’s ♣ K. He leads the

♠ J to dummy’s ♠ Q, partner playing the ♠ 2. Now declarer plays a small
diamond from dummy, partner plays the q J, declarer plays small, and
you are in. Now what?
 MATCHPOINTS

It is pretty easy to reconstruct most of declarer’s hand. We know

from trick 1 that he has the ♣ J, and partner’s play of the ♠ 2 (showing
an odd number) marks declarer with the ♠ KJ doubleton. Declarer must
have started with q A109xx for the diamond plays to make sense; he
obviously went to dummy to try to duck a diamond trick safely to your
hand. Therefore, declarer either has ♠ KJ r Kxx q A109xx ♣ KJx or
♠ KJ r Axx q A109xx ♣ KJx, but which one? If East has the r A,
a heart switch will garner in a 2-trick set, while after a passive return
declarer has 9 tricks. On the other hand, if declarer has the r A, a heart
shift will allow him to take the rest of the tricks (4 diamonds, 3 hearts,
4 spades, and 1 club), while a passive return holds him to 10 tricks. It
seems sensible to reason as follows: If declarer has the hand with the
r A he has a good 16 HCP and a fairly clear 3NT bid, but if he has
the hand with the r K he has only 15 HCP and might not have bid 3NT
despite his good 5-card suit. Therefore, it is better to play safe and avoid
giving up overtricks.
Let’s try a matchpoint cost analysis on this one. If we go for the
set and are wrong it costs 1⁄2 matchpoint against other tables in 3NT, as
they will presumably be faced with the same problem, but nothing to
tables in any other contract since we were beaten by those tables in the
bidding. Since N-S have only 24 HCP and a 5–5 diamond fit with North
having some distribution, there will be many pairs not in 3NT. Let’s say
that half the field reaches 3NT. Then we lose 1⁄2 matchpoint to half the
tables, for an average cost of 1⁄4 matchpoint per table.
Suppose we defend passively and are wrong. This means that South
has the weaker hand, which makes 3NT a very unlikely contract at most
other tables. If we fail to set 3NT when we could have we will lose
1⁄ matchpoint to other tables in 3NT and a full matchpoint to tables
2

in any other reasonable contract. Consequently, our average cost per

table will be almost one full matchpoint. In other words, failing to
set a super-thin game that could have been beaten and will not be bid
at other tables is nearly a full top-to-bottom swing. Therefore we are
giving about 4-to-1 matchpoint odds if we don’t try for the set, so we
would have to be very sure that South wouldn’t have bid 3NT on the
hand with the r K before we take this position.
The moral of this hand is: When there is a reasonable lie of the
cards that allows the contract to be set and the contract is not a normal

386
 DEFENSE

one if that lie of the cards exists, then you should play for the set. If
you fail to defeat such a contract it will be very costly, while if you give
away an overtrick or two it might not cost very much, particularly if
the contract is not normal. This approach must not be overdone. There
must be a reasonable chance for the set, not an extremely unlikely one.
On the previous hand if North had owned the ♠ J, then South would
have to have a 14-count for the hand to be defeated. This would not be
reasonable at all if the opponents were playing strong notrumps, so it
would be better to go passive and hold in the overtricks rather than play
for the impossible.

♠ rq ♣ ♠ rq ♣

When you know you are defending against a normal contract, it is

a different story altogether. Now giving up an overtrick is just as
expensive as letting declarer make the contract, so you should simply
make the play that maximizes your probable tricks.

♠ 43
r AQ85
N-S vul.
q AQJ2
♣ 10 9 5
♠ A852
r 72
q 7543
♣ AK2

West North East South

— — — 1r
Pass 2NT 1 Pass 4r
Pass Pass Pass
1. Forcing raise

Opening lead: ♠ Q.

It is clearly right to win and shift to clubs, but which club? If partner
has the ♣ Q it doesn’t matter, but if declarer has it the choice will make
a difference. At IMPs there is no problem. The defense must take three
club tricks if the hand is to be defeated, so East should shift to a low club

387
 MATCHPOINTS

and hope that declarer has ♣ Qxx, with which he will almost certainly
misguess. At matchpoints it is another story. The contract is normal,
and the opening lead is probably fairly standard. Leading a low club
gains only when declarer has specifically ♣ Qxx or ♣ Qxxx. It loses
whenever declarer has the ♣ J with the ♣ Q or when he holds a singleton
or doubleton ♣ Q (assuming he remembers to put up the queen), for
declarer will then score at least one extra trick by discarding clubs on
dummy’s diamonds. There are more holdings on which the low club
play loses than holdings on which is wins, so East should just cash his
♣ AK, even though this is not his best play to set the contract. On this
hand the overtrick is at least as important as the setting trick, perhaps
even more so. The full hand is:
♠ 43
r AQ85
N-S vul.
q AQJ2
♣ 10 9 5
♠ QJ76 ♠ A852
N
r 10 9 r 72
W E
q 10 9 8 q 7543
S
♣ 8743 ♣ AK2

♠ K 10 9
r KJ643
q K6
♣ QJ6

Your partner did well to lead a spade. At many other tables a red
suit was led, and declarer easily took 11 tricks by pitching 2 clubs on
dummy’s diamonds. If you try for the set declarer will make 12 tricks
instead of 10, so you will lose a full matchpoint to tables that didn’t get
a spade lead. Even if going for the set had worked it would have gained
1⁄ matchpoint only against other tables where a spade was led, since
2

simply cashing out beats tables receiving a red-suit lead.

♠ rq ♣ ♠ rq ♣

As with declarer play, the most difficult defensive problems at match-

points occur when there are possible alternative contracts that are in

388
 DEFENSE

competition with the one you are defending. These can be very hard to
analyze, but the effort must be made. If you can determine the likely
fate of the alternative contract, this will help you decide on your trick-
taking goal.

♠ 10 4 3
r A Q 10 7 2
Both vul.
q A Q 10
♣ 63
♠ AJ95
r 98
q 9432
♣ A52

West North East South

— — — 1q
Pass 1r Pass 2r
Pass 2♠ Pass 3q
Pass 3r Pass 3NT
Pass Pass Pass

Opening lead: ♣ Q.

North has taken quite a position by choosing 3NT rather than 4 r. It is

apparent that many other pairs will play 4 r, and you are in competition
with those defending the heart game. On the bidding South seems to
have 5 diamonds and 3 hearts, and the red suits will certainly run once
declarer gets in. So how should you defend after winning the ♣ A at
trick 1?
The way to beat 3NT, of course, is to shift to the ♠ J, playing
declarer for ♠ Qxx r KJx q KJxxx ♣ Kx. This is by no means an
unreasonable construction, and the ♠ J shift would certainly be correct
at IMPs. The problem with the play is that it gives declarer his twelfth
trick whenever he holds the ♠ K, assuming he has five diamonds. A
club return doesn’t set up anything for declarer, but will be wrong when
partner holds the ♠ K. How about ♠ A and a spade? This gives up on
some legitimate chances of beating the hand, but at least allows us to
score 3 tricks when West holds the ♠ K and 2 tricks when he doesn’t.

389
 MATCHPOINTS

Let’s look at some hands declarer might hold and see what is likely to
happen in 4 r. Keep in mind that East is not likely to lay down an ace
against 4 r, since he doesn’t know about the diamond threat.

a. ♠ Qxx r KJx q KJ10xx ♣ Kx. 4 r will make 5 after a red-suit

lead (declarer pitches 2 spades on the diamonds and eventually
leads to the ♣ K). The ♠ J is the killing shift, but ♠ A and a spade,
holding declarer to 4, will probably get a good board. A club
continuation, allowing declarer to make 5, would be very costly.

b. ♠ Kxx r KJx q KJ10xx ♣ Kx. Again, 4 r will make 5. Either

the ♠ A or a club continuation holds declarer to 5, while the ♠ J
shift allows 6 to make. Whatever we do will probably result in a
bad score, but the ♠ J shift is worst of all.

c. ♠ KQx r Jxx q KJ10xx ♣ Kx. Now 4 r will make 6 on a red-

suit lead (declarer can pitch his losing clubs and establish two
spade tricks). This time the ♠ J shift is really bad, as it costs a full
matchpoint to most tables in 4 r.

This analysis shows that ♠ A and a spade, while never correct in an

absolute sense, is the only play that is not very costly when it is
wrong, so it is the best defense. It may seem odd to take this sort of
a compromise, but it is the best play.
On the same hand, suppose East had diamonds stopped, giving him
♠ AJ95 r 98 q J982 ♣ A52. Now the club continuation is best, for
this will hold declarer to 9 tricks (he won’t duck since he has no reason
to expect the diamonds not to split). If East chooses instead to attack
spades it is now correct to lead the ♠ J, going for the set, since declarer
can’t run the rest of the tricks even if the play doesn’t work. ♠ A and a
spade now becomes the worst of the three possible plays.

390
 DEFENSE

♠ 62
r Q 10 9 6 5 3
None vul.
q4
♣ Q853
♠ AJ5
r A84
q A Q 10 8 5
♣ 74

West North East South

— — — 1♠
1NT 2r Double 2♠
Pass Pass Double Pass
Pass Pass

Opening lead: ♣ 7.

Dummy wins the opening lead with the ♣ Q, partner playing the ♣ J. A
diamond is led from dummy, partner plays the 2 q , declarer the J q , and
you win the Q q . What now?
The bidding and the carding on the first 2 tricks indicate that
declarer’s distribution is 6=0=4=3. It is clear to shift to a trump, for any
trump tricks that may be lost by the trump shift will come back in the
diamond suit. But which trump should West shift to? If East has ♠ Qx,
then ♠ A and a trump is necessary to defeat the contract; the defense
would score two spade tricks and four diamond tricks. A low trump
shift would allow the contract to be made, for declarer would score four
trump winners in his hand, a ruff in dummy, and three club tricks. On
the other hand, if East has ♠ Kx of trumps a low trump shift will defeat
the contract 2 tricks, while ♠ A and a trump will allow declarer to escape
for down 1. Declarer apparently has either ♠ K109xxx r — q KJxx
♣ AKx or ♠ Q109xxx r — q KJxx ♣ AKx.
Let’s see what is likely to happen at other tables. Other N-S pairs
may not be so frisky or other E-W pairs may pass up the penalty and
play a notrump game or partial. It is clear that the cards lie very
favorably for E-W, and they can take at least 2 spades, 3 hearts, and
4 diamonds at notrump. You can’t compete with the tables at which
3NT is bid, since they will outscore you whatever you do. To beat tables

391
 MATCHPOINTS

that stop at a notrump part-score it is essential to collect +300, for +100

will lose to them anyway. Consequently you should play a low spade,
even though this risks allowing declarer to make 2 ♠ doubled when you
had a sure set. The difference in matchpoints between +100 and +300
on this hand figures to be far greater than the difference between +100
and −470.
Vulnerability is the key to hands such as this one. If it had been E-W
vul., a low spade would be correct for the same reasons, but with both
vul., ♠ A and a spade would be clear-cut. Again you can’t compete
against the 3NT bidders, and your +200 will outscore the partscore
bidders, so you play for the set. At this vulnerability there is little
difference between +200 and +500. If it were N-S vul., you could have
a serious problem. Failing to collect +500 when available would cost a
full matchpoint to tables at which 3NT is bid, for their +400 now beats
your +200 instead of losing to your +500. On the other hand, leading
a low spade and being wrong costs a full matchpoint to tables that stop
in a partscore, for you now get −670 instead of +200 while they are
scoring +150. This is a close decision, but it is probably correct to
play for the sure set. Even if East has the ♠ Kx it will only give him
8 points, so many E-W pairs won’t bid the game anyway. Therefore,
being wrong will not lose to every table. On the other hand, if East has
♠ Qx he has only 7 points, so game doesn’t figure to be bid very often.
If this is the case, failing to beat the contract will be almost a complete
top-to-bottom swing.

392
 DEFENSE

♠ K9
r J 10 6 5
None vul.
q A 10 9
♣ J532

♠ A Q J 10 8 5 2
rK
q KJ6
♣ AQ

West North East South

— Pass Pass 3r
4♠ 5r Double Pass
Pass Pass

Opening lead: ♠ A.

You lead ♠ A and a spade, everybody following, and dummy’s r J rides

to your stiff r K. How do you exit?
So far, things have gone well. You have a chance of scoring +500
to beat your potential +420 at 4 ♠. If partner has both the ♣ K and
the q Q it won’t matter what you do, but if declarer has one of these
cards a shift to the wrong minor will blow a trick. If declarer’s hand
is ♠ xx r AQ10xxxx q Qx ♣ xx a rather neat defense is necessary to
collect +500. You must shift to the ♣ Q, and partner must overtake and
put a diamond through. This spectacular defense would occur to some
players holding the West hand and they might try it, but the practical
matchpoint player would reject it and shift to a diamond. The reason
that the diamond shift is correct is that if declarer has the q Q you cannot
make 4 ♠, for there would be no way to avoid losing 1 spade, 1 heart,
and 2 diamonds against competent defense. Therefore, if the diamond
shift costs a trick the 5 r sacrifice is a phantom, and you will still get
an excellent result for +300. However, if declarer holds the ♣ K and
partner the q Q then 4 ♠ will make, so it is important to get +500 if this
is the case. The full hand:

393
 MATCHPOINTS

♠ K9
r J 10 6 5
None vul.
q A 10 9
♣ J532
♠ 64
N
r4
W E
♠ A Q J 10 8 5 2 q Q7432
S
♣ 98764
rK
q KJ6
♣ AQ
♠ 73
r AQ98732
q 85
♣ K 10

4 ♠ is cold, and the diamond shift collects the necessary +500. The
“brilliant” ♣ Q shift would actually allow declarer to get out for down 1,
for he not only loses only one club trick but also gets to pitch his losing
diamond on the dummy’s ♣ J. This would be catastrophic, for you would
score only +100, losing not only to the game bidders but to the partscore
bidders as well. It would be a complete top-to-bottom swing.
Incidentally, what about declarer’s line of play—finessing the heart
instead of playing for the split? It didn’t work on this hand, but it is
the correct play. Declarer has no reason to think that you don’t have a
club loser in 4 ♠, so from his point of view if the hearts are 1–1 his 5 r
save is likely to be a phantom. Consequently, he played the hand on
the assumption that 4 ♠ would make, and the most likely way for this to
happen is for hearts to be 2–0.

394
 DEFENSE

♠ 987
r K5
E-W vul.
q A K Q 10
♣ 7542
♠ KQ5
r Q J 10 8
q J52
♣ AJ9

West North East South

— — — 1♠
Pass 1NT 1 Pass 2♣
Pass 3♠ Pass Pass
Pass
1. Forcing

Opening lead: r Q.

Declarer wins the r K in dummy and passes the ♠ 9 to your ♠ Q. If

declarer happens to have ♣ Qxx, a low club shift will defeat 3 ♠. This
would give declarer ♠ AJ10xx r Ax q xxx ♣ Qxx, a minimal opening
bid. In the more likely case where South holds the ♣ K, a club shift
may give away an overtrick. What is likely to be happening at other
tables? At your table North took a very conservative position. Most
other Norths will drive to game opposite an opening 1 ♠ bid. If game
makes it won’t matter what you do, for you will beat those tables
anyway. However, if 4 ♠ is down 1, it is important to try to tie this
result by defeating 3 ♠. Overtricks won’t matter, but beating 3 ♠ a trick
will turn a loss into a tie against 4 ♠ down 1, so West should go for the
set by underleading his ♣ A. A further consideration is that if South has
the super-light hand not all Souths will open. If South passes N-S are
likely to be in 2 ♠ for +110, so there will be a big difference between
+50 and −140 for E-W.

395
PLAY REVIEW PROBLEMS

Problem 1

E-W vul., West holds ♠ J5 r 106 q 1098 ♣ K108764

West North East South

— — Pass 1NT
Pass 3NT Pass Pass
Pass
Opening lead?
 MATCHPOINTS

Problem 1 Solution

E-W vul., West holds ♠ J5 r 106 q 1098 ♣ K108764

West North East South

— — Pass 1NT
Pass 3NT Pass Pass
Pass
Opening lead?

q 10. This is not an attempt to defeat the contract by hitting partner’s

suit; it is merely the safest lead on a hand on which you have little
chance of a set. The best lead to beat 3NT is a club, but the bidding
indicates likely minor-suit length in dummy since North didn’t use
Stayman, so the club lead is more likely to cost a trick than to gain
one. Partner doesn’t figure to have substantial strength in either major
since he passed as dealer, so the safest lead is best.

398
 PLAY REVIEW PROBLEMS

Problem 2

N-S vul., West holds ♠ 9653 r 842 q QJ7 ♣ AQ8

West North East South

— — — 1r
Pass 1♠ Pass 3r
Pass 3♠ Pass 4r
Pass Pass Pass
Opening lead?

399
 MATCHPOINTS

Problem 2 Solution

N-S vul., West holds ♠ 9653 r 842 q QJ7 ♣ AQ8

West North East South

— — — 1r
Pass 1♠ Pass 3r
Pass 3♠ Pass 4r
Pass Pass Pass
Opening lead?

♣ A. The bidding and your major-suit holdings indicate that both major
suits may run, so this lead is best. Partner may have the ♣ K, and even
if he doesn’t the cash may save an overtrick.

400
 PLAY REVIEW PROBLEMS

Problem 3

Both vul., West holds ♠ J107 r 6432 q 965 ♣ KJ6

West North East South

— — — 1q
Pass 1r Pass 1NT
Pass 3q Pass 4q
Pass 5q Pass Pass
Pass
Opening lead?

401
 MATCHPOINTS

Problem 3 Solution

Both vul., West holds ♠ J107 r 6432 q 965 ♣ KJ6

West North East South

— — — 1q
Pass 1r Pass 1NT
Pass 3q Pass 4q
Pass 5q Pass Pass
Pass
Opening lead?

♣ [Link] the opponents have clubs stopped they will do well in notrump at
other tables, so you assume that they do not have club strength. Partner
can’t have a good spade suit since he failed to overcall.

402
 PLAY REVIEW PROBLEMS

Problem 4

Both vul., West holds ♠ Q52 r QJ83 q QJ97 ♣ 104

West North East South

— — — 1♣
Pass 1r Pass 1♠
Pass 1NT Pass 2r
Pass 2♠ Pass Pass
Pass
Opening lead?

403
 MATCHPOINTS

Problem 4 Solution

Both vul., West holds ♠ Q52 r QJ83 q QJ97 ♣ 104

West North East South

— — — 1♣
Pass 1r Pass 1♠
Pass 1NT Pass 2r
Pass 2♠ Pass Pass
Pass
Opening lead?

♠ 2. The opponents apparently are playing a 4–3 fit, and their extra trick,
if any, over notrump figures to come from ruffs since you have the red
suits bottled up. Consequently, it is correct to lead a trump, even away
from the queen.

404
 PLAY REVIEW PROBLEMS

Problem 5

♠ K 10 9 6 4
r KQJ9
None vul.
q 854
♣ A

♠ Q53
r A52
q A Q 10
♣ K765

West North East South

— 1♠ Pass 2♣
Pass 2r Pass 3NT
Pass Pass Pass

The q 3 opening lead goes to East’s q K

and South’s q A.
Plan the play.

405
 MATCHPOINTS

Problem 5 Solution

♠ K 10 9 6 4
r KQJ9
None vul.
q 854
♣ A

♠ Q53
r A52
q A Q 10
♣ K765

West North East South

— 1♠ Pass 2♣
Pass 2r Pass 3NT
Pass Pass Pass

The q 3 opening lead goes to East’s q K

and South’s q A.
Plan the play.

Most of the other tables figure to be in 4 ♠. This will make 11 or

12 tricks depending on the spade suit. You can make the same number
of tricks as they do if you play the spades the same way, so lead a heart
to dummy and play a spade to the ♠ Q, planning to finesse for the ♠ J. If
the finesse loses and East returns a diamond you must go up and hope
that East started with ♠ AJx and a doubleton diamond, in which case
you will take the same 11 tricks as the spade players. Suppose East had
played the q J at trick 1. Now you would know that the spade players
could make at most 11 tricks, since the diamond finesse would lose for
them, so if you could guarantee 11 tricks you would beat them even
if you were to lose an extra spade trick. Therefore, you would go to
dummy with a heart and ride the ♠ 10. Even if this lost to West’s ♠ J
you would be assured of 11 tricks and a very good board, while if East
had the ♠ J you have a super-top. It would be important not to lose the
first spade trick to East, for a diamond return would put your 11 tricks
in jeopardy.

406
 PLAY REVIEW PROBLEMS

Problem 6

♠ J7
r 32
N-S vul.
q A K 10 9 8 5 4
♣ 63

♠ A 10 6 4
r A Q 10 5
q 63
♣ AK7

West North East South

— — — 1NT
Pass 3NT Pass Pass
Pass

West leads the ♣ 2 and East plays the ♣ Q.

Plan the play.

407
 MATCHPOINTS

Problem 6 Solution

♠ J7
r 32
N-S vul.
q A K 10 9 8 5 4
♣ 63

♠ A 10 6 4
r A Q 10 5
q 63
♣ AK7

West North East South

— — — 1NT
Pass 3NT Pass Pass
Pass

West leads the ♣ 2 and East plays the ♣ Q.

Plan the play.

This is a normal contract, and the lead is moderately unfavorable. You

can guarantee 10 tricks by ducking a diamond, but this does not figure
to be a very good result. The best line is to go whole hog by winning
the opening lead, playing a diamond to the q A, heart to the r 10, and
hoping for a 2–2 diamond split (but finessing if East drops an honor
on the first round). This will produce 13 tricks if everything works,
and will get the maximum out of the hand provided the diamonds suit
comes in. If the diamonds don’t behave—unlucky. Had you gotten a
more favorable lead such as a heart away from the r KJ, you would be
more inclined to duck a round of diamonds so as to not blow your good
start.

408
 PLAY REVIEW PROBLEMS

Problem 7

♠ A J 10
r AKJ54
Both vul.
q 732
♣ Q3

♠ 87
r 86
q A K 10
♣ K J 10 9 8 6

West North East South

— — — 1♣
Pass 1r Pass 2♣
Pass 2♠ Pass 2NT
Pass 3♣ Pass 3q
Pass 4NT ∗ Pass 5r ∗
Pass 6♣ Pass Pass
Pass

Opening lead: ♠ 2.
Plan the play.

409
 MATCHPOINTS

Problem 7 Solution

♠ A J 10
r AKJ54
Both vul.
q 732
♣ Q3

♠ 87
r 86
q A K 10
♣ K J 10 9 8 6

West North East South

— — — 1♣
Pass 1r Pass 2♣
Pass 2♠ Pass 2NT
Pass 3♣ Pass 3q
Pass 4NT ∗ Pass 5r ∗
Pass 6♣ Pass Pass
Pass

Opening lead: ♠ 2.
Plan the play.

Partner obviously got carried away. The rest of the room figures to be
in 3NT, which is cold, so you must go all out to make 6 ♣ regardless
of possible extra undertricks. It is inconceivable that West would
underlead a KQ against a slam, so you must go up ♠ A, play a diamond
to the q 10, heart to the r J and cash the r AK before playing trumps. If
everything works it is your lucky day; if not, you can’t get worse than a
bottom.

410
 PLAY REVIEW PROBLEMS

Problem 8

♠ 75
r J 10 8 7 6 4
E-W vul.
q Q5
♣ AKJ
♠ A8642
rQ
q KJ
♣ 10 9 7 5 3

West North East South

— — — 1r
Pass 4r Pass Pass
Pass

Partner leads the ♠ Q; you win the ♠ A as

declarer follows small.
How do you continue?

411
 MATCHPOINTS

Problem 8 Solution

♠ 75
r J 10 8 7 6 4
E-W vul.
q Q5
♣ AKJ
♠ A8642
rQ
q KJ
♣ 10 9 7 5 3

West North East South

— — — 1r
Pass 4r Pass Pass
Pass

Partner leads the ♠ Q; you win the ♠ A as

declarer follows small.
How do you continue?

r 4 can be beaten by leading the q K if West has the q A and South has a
minimum opening such as ♠ Kxx r AKxxx q xxx ♣ Qx. It is far more
likely that South has the q A, in which case the diamond shift will cost a
trick. Since everybody figures to get to 4 r, it is better to exit passively
and accept −450.

412
 PLAY REVIEW PROBLEMS

Problem 9

♠ K872
r K5
E-W vul.
q QJ82
♣ 943
♠ AQ
r QJ763
q 74
♣ AJ65

West North East South

— — — 1♠
Pass 3♠ 1 Pass 4♠
Pass Pass Pass
1. Limit raise

You lead the r Q; declarer wins in his

hand with the r A and leads the ♠ J to
your ♠ A. Plan the defense.

413
 MATCHPOINTS

Problem 9 Solution

♠ K872
r K5
E-W vul.
q QJ82
♣ 943
♠ AQ
r QJ763
q 74
♣ AJ65

West North East South

— — — 1♠
Pass 3♠ 1 Pass 4♠
Pass Pass Pass
1. Limit raise

You lead the r Q; declarer wins in his

hand with the r A and leads the ♠ J to
your ♠ A. Plan the defense.

If East has the ♣ K, 4 ♠ may be beaten by a low club shift. This would
leave South with a minimum opener such as ♠ J109xx r Ax q AKx
♣ Qxx, which is not out of the question. Is it worth risking a possible
overtrick to play for the set? I think it is. Not many Norths will consider
their hand worth more than a 2 ♠ bid, which means that game will not
be bid at many tables. Consequently, it is worth going all out for the
set, since it is a reasonable possibility and you will be headed for a bad
result if they make the contract, with or without overtricks. A further
consideration is that a club shift probably won’t cost a trick if South has
both the ♣ K and ♣ Q.

414
 PLAY REVIEW PROBLEMS

Problem 10

♠ AQJ
r KQJ9543
E-W vul.
qK
♣ K7
♠ 7654
rA
q QJ95
♣ A863

West North East South

— 1r Pass 1NT
Pass 3r Pass 3NT
Pass Pass Pass

You lead the q Q; dummy’s q K wins and

partner plays the q 2. Declarer leads a
low heart to the r 8, partner playing the
r 2. What now?

415
 MATCHPOINTS

Problem 10 Solution

♠ AQJ
r KQJ9543
E-W vul.
qK
♣ K7
♠ 7654
rA
q QJ95
♣ A863

West North East South

— 1r Pass 1NT
Pass 3r Pass 3NT
Pass Pass Pass

You lead the q Q; dummy’s q K wins and

partner plays the q 2. Declarer leads a
low heart to the r 8, partner playing the
r 2. What now?

Partner’s q 2 warns you away from a diamond continuation, but should

you shift to a club or a spade? If partner has the ♠ K it makes little
difference what you do: declarer has 10 tricks if you shift to a spade,
and even if you play a club into declarer’s strong club holding you
can win the club return and shift to a spade, tangling declarer’s entries
sufficiently so that he can’t take the rest of the tricks. What if declarer
owns the ♠ K? Now a low club return will hold declarer to 4 if he
has the ♣ J and misguesses, but allows him to make 6 if anything else
happens. The pairs in the more normal 4 r contract will always make 5
if South has the ♠ K. Consequently, there is little difference between
−460 and −490, but a great difference between −430 and −460, so the
risky underlead is correct even though it will be wrong more often than
not.

416