Devising a strategy for playing Bangla Hangman (Jhulonto Manob) based on character frequency distribution

In this paper we created an optimal playing strategy for the Bengali (Bangla) Hangman word game which we named Jhulonto Manob. We demonstrated three possible strategies to play Jhulonto Manob. The first strategy tries the most frequent character in the dictionary first and correspondingly moves down the list whereas the second strategy proposed guesses most frequent character in the dictionary of words of length n, where n is the length of the hidden word. But, the optimal strategy states that when a character is guessed incorrectly we should take that information into account and recalculate the frequency distribution on a filtered dictionary where words are of length n and those words does not contain the character(s) which resulted in an incorrect guess. This filtration radically decreases the search space and therefore increases the chance of winning. In the process, interesting statistical data in Bangla words are also revealed.