This crate provides the base components for implementing a player in a two-person game.
The game-player crate provides the core traits and search components needed to build a player for a two-person game.
It provides:
- A min-max game tree search algorithm using alpha-beta pruning and transposition tables for optimal performance.
- A Monte Carlo Tree Search algorithm with UCT-based selection, pluggable value estimation
(
ValueEstimator— random playouts, static evaluation, or any other strategy), configurable initial-value weighting, and eager or lazy child expansion.
- User's Guide — overview of the library, criteria for choosing a search algorithm, an explanation of the minimax search, and a walkthrough of a complete tic-tac-toe player.
- API reference —
cargo doc --open. - Runnable example — examples/tic_tac_toe.rs, via
cargo run --example tic_tac_toe.