Tic-Tac-Toe v1.0.0

First Release of Tic-Tac-Toe

This project is an intelligent system demonstrating the application of adversarial search strategies in ensuring optimal gameplay. In particular, the agent plays tic-tac-toe (also known as noughts and crosses) against a human player. The system has six levels of rationality; while the level 0 agent selects its moves only randomly, winning against the levels 1 to 5 agents is guaranteed to be an impossibility. Moreover, in the interest of algorithm analysis, the program also tracks the match statistics and the decision time spent per move.

This application follows the classical version of the game, where two players take turns in placing their tokens (X's and O's) on a 3 x 3 board and the winner is the player who first places three of their tokens to form a complete row, column, or diagonal. It is to be noted that tic-tac-toe is a solved game resulting in a forced draw given perfect play from both parties.

A formal discussion of the behavior of the agent is provided in Technical Report.pdf