As the title may suggest, this is a weird companion piece to my earlier examination of the origins of the skill system in D&D 2000. Instead of looking backwards, I am now looking forward from D&D 2000 to how three iterations on it would attempt to lump and categorize the let's say 50ish individual skills they inherit from it and what I can possibly learn from them.
Ready? Too bad.
Theoryposting, maybe I'll get to the point of concrete details by the end of this post, maybe I won't.
 |
| Artist: Tennis Cramer |
So, my goal of 3eish is to make something that feels like 3e to me, while also keeping it light enough to hold the whole thing in mind and play fast and easy. Can't get rid of levels without it starting to feel like something else entirely so it's worth asking: what's a level. In 3e there are a couple clear answers:
There are some rationalizations that suggest themselves, like making attack bonus or saving throws a part of the skill system, but I've been down most of those design roads at this point and most of them wind up reinventing RuneQuest. So instead I'm going to reinvent True20; a level is a feat and one HD and everything else falls out of that.
Let's nail that down somewhat:
You may notice there's no real purpose left for classes. That's probably for the best, let's lose them. Carve out their interesting bits into feats and leave the rest.
One-feat character creation is probably too bare to be fun. Instead we'll have two-feat character creation (truly, I have a dizzying intellect). There's a special set of feats called backgrounds, and you pick two of them to start. Your starting HP is 4 if both of them have experience with violence, 3 if one does and 2 otherwise.
The 3e PHB had 7 races and 11 classes, so for parity with its 77 combinations a pool of backgrounds would need...a pool of 14 backgrounds (14 choose 2 is 91, fourteen more combinations for a pool with three fewer elements). Now that's of course not the full range of customization of a 1st level 3e character but it's close enough for my purposes.
My 3eish concept has been through a lot of drafts, but I recently arrived at the following rules for action sequences (combat) and I think they're quite close to what I'll use in the final thing. Who knows, maybe I'll even use the same words.
Encounters take place over rounds, representing a few seconds each.
Each character may act once per round. Characters act in order from highest initiative (aka adrenaline) to lowest and then from highest Wisdom to lowest. Changes to initiative affect the next round's sequence, not the current round.
A character may make only one reaction between one action and the next. They may only react a second time by forfeiting their next action.
Actions that take a few seconds require a character to commit to them; a committed character can't react until after their next action.
A character may move up to their Move score as part of their action; if they take no action, they may move twice that amount.
Careful movement---such as balancing or leaping---requires a check; a failed check occupies the character's action. A failure by 5 or more means they fall.
Contested movement---such as climbing, parting tangles or swimming---requires an action, at least one free hand and a check; a failed check takes a few seconds (committing the character). A failure by 5 or more means they fall, become entangled or sink.
The third edition Player's Handbook lists more than forty skills in which its adventurers can have ranks. Although prior editions of the game had skill systems (AD&D's Nonweapon Proficiencies and BECMI's General Skills), the 3rd edition skill system breaks with them in two ways:
I'm not going to spill any (digital) ink addressing those points. If you have strong opinions about them, good for you. Rather, I'm going to address the question: where did 3E's list of skills come from?
The nice thing about fantasy races is that they give character creation choices with a good degree of granularity and comprehensibility and are a neat way to pull folkloric material into your game. The bad thing about them is that they mean that race science (that is, the false belief that races as understood by society correspond to biologically distinct types of human beings) is true in your fantasy world.
Honestly, the bad part is bad enough to completely sink the goodness of the good part.
This is another in my ongoing struggle to formulate different fantasy folk in a way that isn't race science, but emphasizing the ways that real people saw themselves and others, particularly before the race-scientific idea became prominent in the modern era.
The problem with making things nuanced is that also makes them complicated; race science is conceptually pretty straightforward, which makes it very game-able. Trying to embed fantasy folk in a realistically nuanced web of relationships with each other also buries the conceptual clarity of a player looking at the list of options and saying, "I'm going to play an elf, that means I live in the woods and don't particularly like dwarfs." One runs the risk of overloading the player with world-building, particularly if one chooses to foreground linguistic groups which are among the more salient in-universe distinctions between peoples.
I'm far from a solution to this, but my current idea is to use somewhat-transparent names for languages (or groups of languages) wherever possible, and doing as little "hard" worldbuilding as possible to keep things relatively setting-neutral, while still keeping things complex enough to stop collapsing into the fail state of race science. It's a tall order, let's see how I do.
When making a character, you must choose a people for them to belong to (or at least stem from). A people is defined by a language, a way of life and an alignment, all of which affect how they see themselves and the world.
To make this your own, you only have to identify the peoples in your setting, their languages, ways of life and alignments. You may choose to give every people a distinct ethnonym; I'm not doing this so as not to overload a player with unfamiliar names. Instead, I'm going to organize peoples by language family, grouping setting languages together into four groups (middling, rhuno-buggish, sylvene and foreigner), the first three of which correspond to actual language families. Each family contains a number of peoples with different ways of life and dramatically different outlooks on the world and their place in it.
Before launching into a long post about economics of the ancient world, let's put the exciting part at the top, the random treasure hoard generator. These are going to be significantly smaller than typical fantasy adventure hoards, but I've tried to callibrate them to generate the kidns of hoards people actually find from the bronze and iron ages:
Now, for the less flashy stuff. Before we can really talk about economics, let's establish some units of measurement. These are going to be largely based on historical units of mesopotamia and the eastern mediterannean in antiquity. My sources are kind of all over the place on these so I'm not going to be citing much, but the system I'm giving here hews pretty close to the coins and measures used in Talmudic literature.
A talent (kikar, if you prefer) weighs sixty pounds (maneh), each of which is fifty weights (shekel) of about eight and a half grams; this makes a pound of about 425 grams rather than the 453.59 of the avoirdupois pound, but that's close enough to use the familiar word in my book. A wet measure or dry measure are both a unit of capacity approximately equal to 144 medium-sized eggs (about nine quarts).
The basic unit of money is a weight of silver. Although actual prices vary by place and season, most people will say that a silverweight should buy:
In the essay "Figures, Doors and Passages," Robin Evans contrasts the Classical and Renaissance approach to dividing rooms into spaces with the Modern (post-19th century) ideas about the same. As the main example of the former, he discusses the Villa Madama:
The first difference he points out from modern plans is that:
...the rooms have more than one door - some have two doors, many have three, others four - a feature which, since the early years of the nineteenth century, has been regarded as a fault in domestic buildings of whatever kind or size. Why? the answer was given at great length by Robert Kerr. In a characteristic warning he reminded readers of The Gentleman's House (1864) of the wretched inconvenience of thoroughfare rooms, which made domesticity and retirement unobtainable. The favored alternative was the terminal room, with only one strategically placed door into the rest of the house.
Yet exactly the opposite advice had been furnished by the Italian theorists who, following ancient precedent, thought that more doors in a room were preferable to fewer. Alberti, for instance, after drawing attention to the great variety and number of doors in Roman buildings, said, 'It is is also convenient to place the doors in such a Manner that they may lead to as many parts of the edifice as possible.' This was specifically recommended for public buildings, but applied also to domestic arrangements. It generally meant that there was a door wherever there was an adjoining room, making the house a matrix of discrete but thoroughly interconnected chambers. Raphael's plan [of the Villa Madama] exemplifies this, though it was in fact no more than ordinary practice at the time.
In modern design, enclosed spaces are distinguished between access spaces to be moved through but not occupied (elevators, passages, stairs) and those that are meant to be occupied but not moved through (rooms). However in earlier eras, even buildings large enough to be subdivided into many rooms (it is worth remembering that most buildings would contain few rooms; even a king might have no more than a hall, a private chamber for his family and a kitchen in his castle) would generally favor what Evans calls the "matrix of interconnected rooms."
This contrasts markedly with the kind of dungeon plan perhaps best exemplified by Castle Geryhawk, in which mostly-empty passages serve isolated terminal rooms to be stocked with various monsters or treasures:
The Greyhawk map is quite interesting because it demonstrates an extremely modern layout sensibility in an ostensibly medieval complex. The language Evans applies to William Morris's plan for the Red House at Bexley Heath applies perhaps more directly this dungeon than the house itself:
Yet his commitment to past practice only went so far. The morality of craft and beauty might transform the procedures of building and the appearance of the finished work, but medievalism did not percolat into the plan, which was categorically Victorian and utterly unlike anything built in the fourteenth or fifteenth centuries. Indeed the Red House illustrates the principles laid down by the bourgeois Robert Kerr better than Kerr's own plans: rooms never interconnect, never have more than one door and circulation is unified and distinct.
That may have been said to Morris's shame, but I don't mean it to shame the Castle or its designer. If anything, its plan is less typical of modern dungeons than the kind of dungeon plan perhaps best exemplified by (my beloved) Nethack, wherein corridors lead from specific rooms to specific other rooms, rather than forming a network of circulation for the whole level:
However, I think embracing the matrix of interconnected rooms is a useful tool to keep in mind when designing dungeons or other structures because it naturally introduces a high degree of interconnectivity and makes chance encounters with the inhabitants of the dungeon likely throughout, not simply in their designated areas or when they take to wandering the halls. Here are a last few thoughts in the form of suggestions when drawing a dungeon plan:
This is a thinky post and modest proposal about modularity in TTRPG design.
Since modularity can mean a lot of things to a lot of people, I'll be specific. I am not talking (at least not directly) about the role that discrete packages play in character customization; I am talking about the more general organizational principle of giving a unique name to a particular rule and referring to it at need elsewhere in the text.
When a rulebook describes a number of statuses that can affect a character's combat options: blinded, dazed, deafened, prone, surprised, stunned and so on, and then elsewhere gives players and enemies the option to blind, daze, deafen, knock prone, surprise and stun one another, that's an example of modularity. Modularity is a strategy for sacrificing some amount of clarity (the consequences of stunning a foe may not be explained in the same place in the text as the source of the condition) for to make the rules more consistent (all abilities that stun foes achieve the same effect) and concise (it is not necessary to quote a modular rule everywhere it is referenced). The cost in clarity can be offset by choosing descriptive names for the modular rules so that readers' expectations about what applies to stunned characters map well onto what the actual rules provide for those situations.
A place where modularity is particularly useful is in magical effects. By nature, magical affects are at least somewhat contrary to the readers' expectations and, if precise boundaries of the effect are needed, a lot of text may be the result. Delta has a whole series of posts analyzing the rules texts of spells from Chainmail through 3E; it's evident at a glance that spells have accumulated more and more text as time goes on and the attempt is made to nail down exactly what such and such a piece of magic can do.
Whether you like detailed spells descriptions or vague ones, is not practical to quote the full text of a spell description every time it is referenced. As such, the rules are full of effects defined in reference to specific spells. An unscientific survey of a pretty orthodox old school magic item list had 10-40% (depending on category) of magic item listings specifically referencing a spell effect. Excluding scrolls (which are mostly spell scrolls), potions are next most likely to reference spell effects, followed by wands, miscellaneous items and finally rings. Monster abilities in early editions and their imitators seem much less likely to reference spell effects; however 3/3.5E can't get enough of monster abilities directly referencing spell effects, between traditional spellcasting, spell-like abilities and the very confusing situation of "supernatural" (and sometimes "extraordinary") abilities that duplicate spell effects without certain features of spells such as the possibility to be interrupted or dispelled.
Honestly, I think 3/3.5E is pushing in the right design direction there, even if they did it in the messiest possible way. Nevertheless, I'd wager few people have the system mastery to accurately tell you off hand the differences between a spell, spell-like ability, supernatural ability and an extraordinary ability. And even fewer would be able to accurately guess the effects of an unfamiliar spell from its name. In 4E, the designers pulled hard in the opposite direction by removing the modularity from spells (and prayers and maneuvers and psionic powers) entirely: all abilities got siloed to the class or monster type that granted them where they are described in full. In particular, this means that while the information is conveniently organized for the DM running the encounter, the economics of space and how many unique abilities a DM can actually keep in mind while running the encounter suddenly place pretty hard limits on how intricate monsters can be.
Rather than continue talking about the historical vicissitudes of spell modularity, I will instead make a modest proposal:
Keep magical causes and magical effects separate:
Different ways of working the same wonder can be made to feel very different if, for example, different costs are demanded from a holy person and a necromancer to bring about the same healing of a broken body. The details of the wonder equip a ref to rule on its applications and edge cases, while the details of the causes are able to be more diegetic and flavorful without sacrificing modularity for that.
A book of wonders could be used by a referee both as an at-the-table reference for whatever magic is in play but also as a source for describing new wizarding traditions (by selecting a set of wonders and describing that tradition's means of causing each), artifacts or creatures. If the rules texts were sufficiently concise, a deck of cards bearing wonder descriptions could be physically handed out to players when they become able to use them.
Characters have to come from peoples, and those peoples have ways of life. Where is home for your people, and how do they see themselves?
Alignment describes how a people sees themselves and how they divide up the rest of the world. Some see themselves as worldly and do not share the alignment of their folk. Even those who share an alignment do not necessarily like each other or agree about where one another fit into the order of things.
A civilized fellow knows that without the nameless gods of heaven to teach mortals the arts of civilization, we would be naked heathens all. Elders who still name the gods of the earth may know no better, but the barbarian who call on demons from behind the sky and beneath the sea. A heaven-fearing fellow knows a 'worldly' one for what they really are: a scoffer not to be trusted.
Civil folk all see the peoples of cities and houses as civil (though they may war with one another just the same), woods-dwellers as elders and nomads under the open sky as barbarians. However, they differ on a few points:
Freedom is especially treasured by those who live under the open sky, viewing the labors of settled peoples as a kind of thralldom. They honor their beloved dead and all the spirits of the air and earth that touch their lives, chanting their names and deeds for later generations. The 'worldly' fellow who scorns these gods is a lonesome one, a tiny fellow adrift without ancestor to guide their feet or shelter them from the cruel winds.
Those of cities, the deep earth and houses are seen as thralls, while those who range in boats and carts, from the hillsides, or in the deep woods are seen as half-thralls. A free neck does not bend to the yoke of a thrall, and a warrior certainly does not fear one.
Goodwrights have learned the secrets of craft from the earth itself. They celebrate the good earth with the work of skilled hands. Though a goodwright will take spoil from war, they detest nothing as much as a thief. And what is a 'worldly' creature who pays no mind to their mother earth but a thief of her gifts?
All outsiders are seen with a measure of suspicion, but wrights of the deep earth and of the hillsides revile one another as thieves nearly without exception. Hillside wrights will treat with most other outsiders, but those of the deep earth consider the folk of boats and carts or the woods as thieves as bad as hated hill-wrights.
The day is for the sun, the night for the stars, but the moon wanders through day and night in its turnings. Moonfolk chase the balance of the moon, honoring the earth in all its seasons. Moonfolk cultivate the woods around them with a gardener's care and live from the game and fruits thereof.
The enemies of the moonfolk are the nightfolk who harry them from the deep earth and open sky, while city and house dwellers are dayfolk and useful allies against the nightfolk when they can be roused to it. Those of boats and carts or hillsides or 'worldly' folks who think themselves wiser than everyone are smallfolk and not worth serious consideration.
The worldly are not many, and they are called lonesome, scoffer, smallfolk, thief or worse where they go. Nevertheless, a worldly fellow hates to suffer fools and sees no reason to give up the better parts of life to distant gods. They can't abide the small-mindedness of boat and cart travelers, the cruelty and hypocrisy of the cities, the incurious provincialism of the deep earth, the baseless boasting of hillside folk, the credulous superstition of house holders and open sky nomads both or the sanctimonious mysticism of the woods.
This post is an attempt to sketch out a different way that the base book classes of AD&D 2e, an edition of the game I never directly experienced, could have been developed in another edition.
Obviously, now we have the benefit of three additional editions of official D&D and numerous offshoots and hacks widely available, so this isn't to say, "this is what 3e/4e/5e should have been" as much as, "this is another direction that could be explored," now that we've taken some probing steps in various directions.
So: classes in 2e AD&D were derived from the four types:
| Warrior | Wizard | Priest | Rogue |
| Fighter | Mage | Cleric | Thief |
| Ranger | Illusionist | Druid | Bard |
| Paladin |
Let me single out a feature that i rather like: only the four core classes are always available to players. From the DM's chair, I find this very appealing; it's a much lighter demand of a setting that it have fighters, mages, clerics (though clerics present some problems) and thieves than that it also have barbarians, bards, druids, monks, paladins, rangers and sorcerers (to say nothing of warlocks) as well. Bards, druids and monk are particularly egregious in terms of the amount of baggage that their inclusion brings. Although 4e and 5e took the step of distinguishing common and uncommon "races," they do not do the same for classes (which, for my money, do a lot more heavy lifting in terms of providing implied setting).
Another interesting thing about this grid is the idea that we could make the rows meaningful as well as the columns. for example, the Ranger and Druid do seem to have something in common, as do the Bard and Illusionist, suggesting:
| Warrior | Wizard | Priest | Rogue | |
| Fighter | Mage | Cleric | Thief | |
| Ranger | Druid | Wilderness | ||
| Illusionist | Bard | Trickery | ||
| Paladin |
...hey, this looks a lot like:
| Defender | Controller | Leader | Striker | |
| Fighter | Warlord | Rogue, Ranger |
Martial | |
| Wizard | Warlock | Arcane | ||
| Paladin | Cleric | Divine |
I'll freely admit that this is a part of 4e's design that i genuinely do like. I think you could potentially communicate a lot of setting using a matrix like this, especially if your sources of power corresponded to diagetic factions. even the gaps in the table (though later books, of course, fill them) serve a purpose. It says something about what these sources of power are about that there would be no martial controller, arcane defender or divine striker, for example.
However, this tableau is of basically no use for restricting available classes, which was one of the things that the 2e one does so well. Furthermore, little enough is directly shared between classes of the same source of power that they feel, to me, a lot more like palette swaps on the same game math than genuinely distinct types of character. This is exacerbated by the insistence on filling in gaps in the source/class matrix
Returning to our spread of classes in 2e AD&D, let's use a different axis to split things. Instead of source of power, let's use (3 position) alignment:
| Warrior | Wizard | Priest | Rogue | |
| Ranger, Paladin | Cleric | Lawful | ||
| Fighter | Mage | Druid | Bard, Thief | Neutral |
| *Berserk | *Assassin | Chaotic |
Italicized classes are ones that we'd expect DM veto power on. I'm adding in a couple of thematic Chaotic classes, nothing to see here. there's a couple of things that i don't love about this. First, as in 4e there's no clear way to use this table to restrict available classes; the neutral classes don't correspond to the core classes (druids in particular probably should not be the core priest class). Second, i don't like the idea of the core classes being restricted to neutral alignment. i'd prefer something like this:
My proposal is to distinguish the alignment of druids and bards from 'neutrality' (probably replacing that with a better term). I'm merging paladins into rangers (as we don't necessarily need two distinct Lawful classes about defending the innocent), and making clerics no longer a core class but instead a Lawful-only specialty.
| Warrior | Wizard | Priest | Rogue | |
| Fighter | Mage | Thief | any | |
| Ranger | Cleric | Lawful | ||
| Druid | Bard | Old Believer | ||
| Berserk | Assassin | Chaotic |
This setup involves four alignments:
This is better, in my book, but it can be further refined. Notice that there is no longer a core Priest class, and take the obvious step of collapsing that column with Wizard:
| Warrior | Wizard | Rogue | |
| Fighter | Mage | Burglar | any |
| Druid | Bard | Old Believer | |
| Ranger | Cleric | Lawful | |
| Berserk | Assassin | Chaotic |
This is approaching what I'd really like to see. We understand fighter, mage and thief to always be on offer. The specialty classes are all bound up with a diagetic alignment (in this case a mostly religious meaning). Any type is available to any alignment, but that doesn't mean that every alignment has a specialty in every type. Indeed, in the first draft of this post I had a Necromancer Chaotic wizard specialty rather than Berserk. While that could work, it seemed better to me to place all wizards under the suspicion of being Chaos magicians by associating wizardry and Chaos.
Available talents ("nonweapon proficiencies" need rebranded) would be chosen by combining the Warrior, Wizard or Rogue list with the appropriate alignment list. Other distinctions (say, what a cleric can do that a lawful mage cannot) would be as noted in the specialty class.
If one wished to customize the array of classes offered to a given
setting, identifying alignments and their associated specialties seems
like a natural way to do this; for example, someone could remove the Old
Faith and add in something home-brewed.
In my dream vision of this, the core classes each get a spread to themselves. Each aligned pair of specialty classes could then be presented in a single spread, one on each facing page. Nine classes in twelve pages.
| Roll | 1 | 2,3 | 4,5 | 6 |
| Skills | 4 | 3 | 2 | 1 |
| Ranks | Tiers |
|---|---|
| 0 | 1 |
| 1-2 | 2 |
| 3-6 | 3 |
| 7+ | 4 |
