Tic Tac Toe, Reinforced Learning and Hash Tables - Don't understand how to create a game

Question

For my data structures course, we were asked to create Tic Tac Toe(3x3)that:

1) Has two player - dumb and smart(computer plays both sides)
2) Dumb player - "X" , always makes first move, and all of his moves are random.
3) Smart player - "O" relies on reinforced learning. Basically, he looks through the hash table(transposition table) created from the different game states learned from prev games and finds good move. Project specifications page has no information on the implementation of the "intelligence"

I don't understand how to handle 3). At what stage of the game, board's hash needs to be inserted into table ? How Smart player decides on his next move ?

Thanks !

Answer 1

You can find something about your problem here. It's some theory about similar project (Reinforcement Learning in Tic-Tac-Toe Game and Its Similar Variations) as you are solving.

Answer 2

I don't know what your course requires, but I can think of a simple approach. You need records of game states, moves, and wins. So the smart player looks up the current game state in the record of past games and makes the move that led to the most wins. (Apologies if this is too obvious.)