The Gambler's Problem

Using Value Iteration to find the optimal betting policy.

0.40

Value Estimates (Probability of Winning)

Final Policy (Stake)