Deep Search: 150 MCTS simulations per move explores thousands of positions Pattern Learning: Policy network learns winning move sequences Forced Captures: Properly implements mandatory jump rules King Awareness: Values promotions and king positioning Multi-jump Sequences: Handles complex capture chains Center Control: Mimics grandmaster position.
deep-reinforcement-learning game-theory monte-carlo-tree-search policy-network combinatorial-game-theory checkers-rl mcts-simulations forced-captures king-awareness multi-jump-sequences
-
Updated
Dec 14, 2025 - Python