Revisiting CFR+ and Alternating Updates

The CFR+ algorithm for solving imperfect information games is a variant of the popular CFR algorithm, with faster empirical performance on a range of problems. It was introduced with a theoretical upper bound on solution error, but subsequent work showed an error in one step of the proof. We provide updated proofs to recover the original bound.

Computer Science Mathematics

View graph Source

Paper

References (8)

01Online Convex Optimization for Sequential Decision Processes and Extensive-Form GamesGabriele Farina, Christian Kroer, T. Sandholm2018 · AAAI Conference on Artificial Intelligence · 65 citations In Library

02Solving Heads-Up Limit Texas Hold'emOskari Tammelin, Neil Burch, Michael Bradley Johanson et al.2015 · International Joint Conference on Artificial Intelligence · 182 citations

03Heads-up limit hold’em poker is solvedMichael Bowling, Neil Burch, Michael Bradley Johanson et al.2015 · Science · 555 citations

04Solving Large Imperfect Information Games Using CFR+Oskari Tammelin2014 · arXiv.org · 177 citations

05Regret Minimization in Games with Incomplete InformationMartin A. Zinkevich, Michael Bradley Johanson, Michael Bowling et al.2007 · Neural Information Processing Systems · 903 citations

06Theory of Games and Economic BehaviorRobert Leonard2006 · 10.0k citations

07A simple adaptive procedure leading to correlated equilibriumS. Hart, A. Mas-Colell2000 · 1.4k citations

08Equilibrium Points in N-Person Games.J. Nash1950 · Proceedings of the National Academy of Sciences of the United States of America · 8.2k citations

Similar papers

Lazy-CFR: fast and near optimal regret minimization for extensive games with imperfect informationYichi Zhou, Tongzheng Ren, J. Li et al.2018 · International Conference on Learning Representations · 15 citations In Library

Lazy-CFR: fast and near-optimal regret minimization for extensive games with imperfect informationYichi Zhou, Tongzheng Ren, Jialian Li et al.2020 · International Conference on Learning Representations In Library

Faster Game Solving via Asymmetry of Step SizesLinjian Meng, Tianpei Yang, Youzhi Zhang et al.2025 · AAAI Conference on Artificial Intelligence · 0 citations In Library

Stable-Predictive Optimistic Counterfactual Regret MinimizationGabriele Farina, Christian Kroer, Noam Brown et al.2019 · International Conference on Machine Learning · 36 citations In Library

Solving Imperfect-Information Games via Discounted Regret MinimizationNoam Brown, T. Sandholm2018 · AAAI Conference on Artificial Intelligence · 197 citations In Library

Sync Pure Counterfactual Regret Minimization in Incomplete Information Extensive Form GamesQi Ju2023 · arXiv (Cornell University) In Library

AutoCFR: Learning to Design Counterfactual Regret Minimization AlgorithmsHang Xu, Kai Li, Haobo Fu et al.2022 · Proceedings of the AAAI Conference on Artificial Intelligence · 13 citations In Library

Accelerating Nash Equilibrium Convergence in Monte Carlo Settings Through Counterfactual Value Based Fictitious PlayQi Ju, Hei, Falin, Ting Feng et al.2023 · arXiv (Cornell University) · 6 citations In Library

Deep Counterfactual Regret MinimizationNoam Brown, Adam Lerer, Sam Gross et al.2018 · International Conference on Machine Learning · 256 citations In Library

Parallel Counterfactual Regret Minimization in Crowdsourcing Imperfect-information Expanded GameJie Zhang, Kefan Li, Baoming Zhang et al.2021 In Library

Source of record

arXiv

44B

Welcome.

This panel sticks with you. Pick anything below and it opens right beside it, so you can dig through 60,000-plus records without ever losing your spot here.

Travel 44B

LibraryPapers, policy, standards, statute LabsEvery organization building AI ModelsIntelligence, price, and speed BenchmarksThe evaluation catalog IncidentsAI failures and harms SearchOne field across everything DashboardThe state of AI accountability in NY 44B RegistryThe Art. 44-B compliance portal