Framework for adjudicating a chess match with Computer Agents. A pair programming effort with Joyce Lee

• Minimax implemented, capable of evaluating material exclusively
  • passes promotion test

Requirements for running this project:

1. Python 3.12
2. uv package management
3. CMake, for openspeil

brew install python@3.12
curl -LsSf https://astral.sh/uv/install.sh | sh
brew install cmake

Afterwards simply run

uv sync
source .venv/bin/activate # Unix only

Starting a new game is simple. To play against Minimax run:

python3 main.py

To run the test suite, run:

python3 main.py test
python3 main.py test --filename=endgame_test.txt --depth=10

Game is adjudicated in chengine/main.py

Players (agents) are located in chengine/players. To add a new player, make a new folder within players

1.1 - new testing framework and bugfix 1.2 - alpha beta pruning 1.3 - changed top level to use pruning. Added best continuation tracker 1.4 - added late move reduction 1.4.1 - changed late move reduction to prioritise queen promotion 1.5 - added new evaluation function that combines different heuristics. Added pawn occupation heuristic 1.5.1 - added preprocessing step to convert fen into matrix 1.6 - added temporal advantage including a to-move bonus and piece development heuristic 1.7 - added king safety heuristics

### Engine

Minimax

[ ] Map for piece positioning, where each piece is best placed for each color, or for both colors

[ ] devalue king safety. Ensure my engine does not like the wayward queen attack

[ ] implement openings book

[ ] Implement translator for UCI to work with GUI

[ ] learn to fish, what is required, releasing GIL or using parallelisable language