Probability · Conditional and Bayes · Medium
A coin of unknown bias
A coin lands heads with probability p. You do not know p: it was drawn uniformly at random between 0 and 1 when the coin was made. You flip the coin three times and see two heads and one tail. What is the probability that the next flip lands heads?
a fraction, an expression or a decimal to two significant figures · 5/8 · ln 2 · pi · inf
Answer and solution
Answer
3/5 ≈ 0.6
Solution
Weight each possible bias by how well it explains the flips, then take the average bias.
The weight on p is proportional to p²(1 − p). The expected p is ∫p³(1 − p) dp / ∫p²(1 − p) dp = (1/4 − 1/5) / (1/3 − 1/4) = (1/20) / (1/12) = 3/5. In general, h heads in n flips gives (h + 1) / (n + 2).
Conditional and Bayes comes up in 8 firms' roundsJane Street · Hudson River Trading · Citadel Securities · +5
These 16 rounds ask this kind of problem live, with follow-ups. Every round on every track is in Premium.