Research map / Bandits and online learning / Online convex optimization and games
Faster Game Solving via Predictive Blackwell Approachability: Connecting Regret Matching and Mirror Descent
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 |
|---|---|---|---|
| Use predictive regret matching+ in CFR with alternating updates and quadratic averaging | CFR for extensive-form games · Battleship | Nash gap: 3-6 orders of magnitude | Helped |
| Use predictive regret matching+ in CFR with alternating updates and quadratic averaging | CFR for extensive-form games · Pursuit-evasion | Nash gap: 3-6 orders of magnitude | Helped |
| Use predictive regret matching+ in CFR with alternating updates and quadratic averaging | CFR for extensive-form games · Battleship and Pursuit-evasion | Nash gap: around 10 orders of magnitude | Helped |
| Use predictive regret matching+ in CFR with alternating updates and quadratic averaging | CFR for extensive-form games · Goofspiel | Nash gap: 0.5-1 order of magnitude | Helped |
| Use predictive regret matching+ in CFR with alternating updates and quadratic averaging | CFR for extensive-form games · River endgame | Nash gap | Mixed |
| Use predictive regret matching+ in CFR with alternating updates and quadratic averaging | CFR for extensive-form games · Kuhn poker | Nash gap | Helped |
| Use predictive regret matching+ in CFR with alternating updates and quadratic averaging | CFR for extensive-form games · Leduc poker | Nash gap | Mixed |
| Use predictive regret matching+ in CFR with alternating updates and quadratic averaging | CFR for extensive-form games · Liar's Dice | Nash gap | No clear effect |
| Add quadratic averaging to CFR+ (ablation) | CFR+ · Most tested games other than poker and Liar's Dice | Nash gap | Hurt |
| Add quadratic averaging to CFR+ (ablation) | CFR+ · Poker games and Liar's Dice | Nash gap | Helped |
| Use predictive regret matching+ in CFR instead of non-predictive CFR variants | CFR for extensive-form games · Non-poker games | Nash gap: by several orders of magnitude | Helped |
| Use predictive regret matching+ in CFR instead of non-predictive CFR variants | CFR for extensive-form games · Poker games | Nash gap | Mixed |
Closest papers by meaning
- Extensive-Form Game Solving via Blackwell Approachability on Treeplexes (NeurIPS 2024 · similarity 0.94)
- Convergence of Regret Matching in Potential Games and Constrained Optimization (ICLR 2026 · similarity 0.91)
- Optimistic Regret Minimization for Extensive-Form Games via Dilated Distance-Generating Functions (NeurIPS 2019 · similarity 0.90)
- Optimism Without Regularization: Constant Regret in Zero-Sum Games (NeurIPS 2025 · similarity 0.89)
- Learning Not to Regret (AAAI 2024 · similarity 0.88)
- Thresholds for sensitive optimality and Blackwell optimality in stochastic games (NeurIPS 2025 · similarity 0.88)
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.