This repository contains an implementation of the game tic tac toe using minimax with alpha-beta pruning. Apart from the classic 3 x 3 game version, there is implemented a 5 x 5 variant where the player needs 4-in-a-row to win.

The idea behind 5 x 5 version, is to see how minimax scales with depth and how to control it. It was proposed a ranking of the opponent moves, known as "tempo", and discarding the ones with the smallest value. Introducing a way to evaluate the "goodness" of the move and keeping only the best is a good heuristic to start with, but needs improvement.

• conditions.py: contains functions to evaluate the board
• minimax.py: contains an implementation of the minimax algorithm
• moves.py: move availability, other move related functions
• tempo.py: contains a heuristic algorithm that assigns points to moves regarding their goodness. Needs improvement
• tic_tac_toe.py: Initializes the game and display the board in console.

The algorithm was developed as a second part for the Foundations of AI coursework (COMP6231).