dynamic-programming

Optimal algorithm for Letter Boxed word game: bitmask DP, BFS on (mask, last), admissible heuristics like set-cover bounds, pruning techniques for minimal words. Fast for L=12 puzzles with chaining and coverage.

Compute the exact expected total payoff at stopping time using finite-state Markov chain and dynamic programming. Includes minimal state space, recurrence equations, backward substitution, and Python code for precise fraction.