Solution — step by step (with careful arithmetic)
Step 1 — likelihoods of drawing two reds from each box
• From Box A (3 red, 2 blue, total 5):
3 2
𝑃(two reds ∣ 𝐴) = × .
5 4
Compute digit by digit:
3 2 3×2 6 3
× = = = .
5 4 5 × 4 20 10
• From Box B (2 red, 4 blue, total 6):
2 1
𝑃(two reds ∣ 𝐵) = × .
6 5
Compute:
2 1 2×1 2 1
× = = = .
6 5 6 × 5 30 15
Step 2 — total probability of observing two reds
1 1
Priors: 𝑃(𝐴) = 2 , 𝑃(𝐵) = 2.
So
1 3 1 1
𝑃(two reds) = ⋅ + ⋅ .
2 10 2 15
Compute each term with a common denominator (use 60):
1 3 3 9
• ⋅ = 20 = 60.
2 10
1 1 1 2
• ⋅ = 30 = 60.
2 15
Add:
9 2 11
𝑃(two reds) = + = .
60 60 60
Step 3 — use Bayes to get 𝑷(𝑨 ∣ two reds)
1 3
𝑃(𝐴) 𝑃(two reds ∣ 𝐴) 2 ⋅ 10
𝑃(𝐴 ∣ two reds) = = .
𝑃(two reds) 11
60
Compute numerator:
1 3 3
⋅ = .
2 10 20
11
Divide by :
60
3/20 3 60 3 × 60
= × = .
11/60 20 11 20 × 11
Simplify 60/20 = 3:
3×3 9
= = .
11 11
9
So (a) 𝑃(Box A ∣ two reds) = 11 .
9 2
(Then 𝑃(Box B ∣ two reds) = 1 − 11 = 11.)
Step 4 — probability the third ball (from the same box) is blue, given two reds
We condition on which box it is (use total probability with the posterior weights).
• If it was Box A: after removing 2 reds from the original (3R,2B), remaining are 1 red and 2
blue (total 3).
2
So 𝑃(blue ∣ 𝐴,two reds) = 3.
• If it was Box B: after removing 2 reds from (2R,4B), remaining are 0 red and 4 blue (total 4).
So 𝑃(blue ∣ 𝐵,two reds) = 1.
Now combine using posteriors:
2
𝑃(blue ∣ two reds) = 𝑃(𝐴 ∣ two reds) ⋅ + 𝑃(𝐵 ∣ two reds) ⋅ 1.
3
Plug numbers:
9 2 2
= ⋅ + ⋅ 1.
11 3 11
Compute first term:
9 2 9×2 18 6
⋅ = = = (since 18 ÷ 3 = 6 and 33 ÷ 3 = 11).
11 3 11 × 3 33 11
2
Now add :
11
6 2 8
+ = .
11 11 11