A demonstration of the Polaris poker system

ABSTRACT

Poker games provide a useful testbed for modern Artificial Intelligence techniques. Unlike many classical game domains such as chess and checkers, poker includes elements of imperfect information, stochastic events, and one or more adversarial agents to interact with. Furthermore, in poker it is possible to win or lose by varying degrees. Therefore, it can be advantageous to adapt ones' strategy to exploit a weak opponent. A poker agent must address these challenges, acting in uncertain environments and exploiting other agents, in order to be highly successful. Arguably, poker games more closely resemble many real world problems than games with perfect information. In this brief paper, we outline Polaris, a Texas Hold'em poker program. Polaris recently defeated top human professionals at the Man vs. Machine Poker Championship and it is currently the reigning AAAI Computer Poker Competition winner in the limit equilibrium and no-limit events.

REFERENCES

• 1: Michael Bowling , Michael Johanson , Neil Burch , Duane Szafron, Strategy evaluation in extensive games with importance sampling, Proceedings of the 25th international conference on Machine learning, p.72-79, July 05-09, 2008, Helsinki, Finland [doi>10.1145/1390156.1390166]
• 2: M. Johanson. Robust strategies and counter-strategies: Building a champion level computer poker player. Master's thesis, University of Alberta, 2007.
• 3: M. Johanson, M. Zinkevich, and M. Bowling. Computing robust counter-strategies. In Advances in Neural Information Processing Systems 20 (NIPS), 2008. To appear (8 pages). A longer version is available as a University of Alberta Technical Report, TR07-15.
• 4: M. Zinkevich, M. Johanson, M. Bowling, and C. Piccione. Regret minimization in games with incomplete information. In Advances in Neural Information Processing Systems 20 (NIPS), 2008.
• 5: M. Zinkevich and M. Littman. The AAAI computer poker competition. Journal of the International Computer Games Association, 29, 2006. News item.

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
        I.2.11 Distributed Artificial Intelligence

Additional Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE
      I.2.3 Deduction and Theorem Proving
          Subjects: Uncertainty, "fuzzy," and probabilistic reasoning
      I.2.4 Knowledge Representation Formalisms and Methods
  I.6 SIMULATION AND MODELING
      I.6.4 Model Validation and Analysis
      I.6.5 Model Development
          Subjects: Modeling methodologies

General Terms:
Algorithms, Design, Performance, Theory