/

Let's build something legendary together

  • About
  • Projects
  • Skills

Game-Playing Agent

Apr 2022 - May 2022

This video shows the gameplay between the Cachex Game-Playing Agent playing as blue and the Monte Carlo Tree Search opponent playing as red.

Cachex Game-Playing Agent is a Python program that excels in playing the Cachex game, a variant of the Hex game that allows players to capture opponent pieces.

Cachex Game-Playing Agent's significant strengths lie in its strategic approach to move exploration, which enables it to consider as many possible moves as possible and make optimal decisions within the time limit. This is achieved through its utilization of the Minimax algorithm with Alpha-Beta pruning that enables the agent to search the game tree while discarding unnecessary branches.

Additionally, Cachex Game-Playing Agent's efficiency is further improved through its ability to quickly identify instant winning and losing positions and strategically explore move orders. It prioritizes those moves that are most likely to lead to victory, thereby making Alpha-Beta pruning more effective. As a result, Cachex Game-Playing Agent achieves an average win rate of 85% against a variety of strategies.