This one-off experiment is composed of a Hangman game interface and a Hangman player. The goal of the experiment is to evaluate the efficacy of different strategies available to the Hangman player.

The Hangman game takes an input dictionary and randomly selects a target word. Meanwhile, the player is provided with the same input dictionary and produces letter guesses given the current game state. The player's score is determined by the number of incorrect guesses before the player guesses all the letters in the target word.

Extensions

One might imagine modifying the game to select words from the dictionary with a non-uniform distribution. The player's strategy could then be adjusted to account for this, given the distribution. Another variation might have the distribution be unknown to the player apriori, to be discovered through repeated iterations of the game.