SpinGPT: A Large-Language-Model Approach to Playing Poker Correctly
This work addresses the problem of scaling AI to popular multi-player poker formats like tournaments, where traditional methods like CFR are limited, offering a potential new approach for game developers and AI researchers.
The authors tackled the challenge of applying AI to multi-player poker, specifically the Spin & Go format, by developing SpinGPT, a large language model trained with supervised fine-tuning and reinforcement learning, which achieved 78% tolerant accuracy against a solver and a win rate of 13.4 +/- 12.9 BB/100 against Slumbot in heads-up play.
The Counterfactual Regret Minimization (CFR) algorithm and its variants have enabled the development of pokerbots capable of beating the best human players in heads-up (1v1) cash games and competing with them in six-player formats. However, CFR's computational complexity rises exponentially with the number of players. Furthermore, in games with three or more players, following Nash equilibrium no longer guarantees a non-losing outcome. These limitations, along with others, significantly restrict the applicability of CFR to the most popular formats: tournaments. Motivated by the recent success of Large Language Models (LLM) in chess and Diplomacy, we present SpinGPT, the first LLM tailored to Spin & Go, a popular three-player online poker format. SpinGPT is trained in two stages: (1) Supervised Fine-Tuning on 320k high-stakes expert decisions; (2) Reinforcement Learning on 270k solver-generated hands. Our results show that SpinGPT matches the solver's actions in 78% of decisions (tolerant accuracy). With a simple deep-stack heuristic, it achieves 13.4 +/- 12.9 BB/100 versus Slumbot in heads-up over 30,000 hands (95% CI). These results suggest that LLMs could be a new way to deal with multi-player imperfect-information games like poker.