Research map / Learning theory and algorithms / Combinatorial and search optimization
Nested Monte Carlo Search for Two-Player Games
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add discounting (playout-length) to NMCS (ablation) | NMCS level 2 · Breakthrough | win rate vs undiscounted NMCS (%): 99.4 | Helped |
| add discounting (ablation) | NMCS level 0 · misere Breakthrough | win rate vs undiscounted (%): 42.4 (level 0) -> 80.8 (level 1) -> 90.0 (level 2) | Mixed |
| add discounting to NMC(3) (ablation) | NMC(3) · Breakthrough trap position | correct move frequency (%): 11.9 -> 64.1 | Helped |
| add Cut on Win pruning without discounting (ablation) | NMC(3) · Breakthrough trap position | states visited (k): 4, 459 -> 1, 084 (COW<=1) | Helped |
| add Cut on Win pruning without discounting (ablation) | NMC(3) · Breakthrough trap position | correct move frequency (%): 11.9 -> 12.3 | No clear effect |
| add Prune on Depth with discounting (ablation) | NMCD(3) · Breakthrough trap position | states visited (k): 2, 775 -> 627 (POD<=3) | Helped |
| add Prune on Depth with discounting (ablation) | NMCD(3) · Breakthrough trap position | correct move frequency (%): 64.1 -> 62.4 | No clear effect |
| nested playouts (NMCS n=1, COW) in UCT instead of random playouts | UCT · misere Breakthrough | win rate vs MCTS (%): 99.8 | Helped |
| nested playouts (NMCS n=1, COW) in UCT | UCT · Breakthrough | win rate vs MCTS (%) at 320ms: 39.6 | Hurt |
| nested playouts (NMCS n=1, COW) in UCT | UCT · Domineering | win rate vs MCTS (%) at 320ms: 50.0 | No clear effect |
| use nesting level 2 instead of level 1 in playouts | NMCS in UCT · Breakthrough | win rate vs MCTS (%) | Hurt |
| minimax rollouts (MCTS-MR) instead of NMCS | MCTS · misere Breakthrough | win rate vs MCTS (%): 49.4 | No clear effect |
Closest papers by meaning
- Split Moves for Monte-Carlo Tree Search (AAAI 2022 · similarity 0.88)
- Mastering Board Games by External and Internal Planning with Language Models (ICML 2025 · similarity 0.86)
- An Efficient Algorithm for Thresholding Monte Carlo Tree Search (NeurIPS 2026 · similarity 0.86)
- Implicit Search via Discrete Diffusion: A Study on Chess (ICLR 2025 · similarity 0.85)
- Monte Carlo Tree Search in the Presence of Transition Uncertainty (AAAI 2024 · similarity 0.84)
- Mastering Zero-Shot Interactions in Cooperative and Competitive Simultaneous Games (ICML 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.