AutoChess

Computer chess agent written in Kotlin.

Features
  • Alpha-beta pruning
  • Iterative Deepening Search
  • Transposition Table / Evaluation Cache
  • Material Difference Heuristic
  • Mask-based board representation
Caveats
  • Does not support castling yet.
Usage

Run Entrypoint.kt. The agent will play a game against itself. For each move a search of at least depth 3 is conducted. Depth will be iteratively extended until 15 seconds have elapsed since the turn began.

GitHub

View Github