An Ex­ploita­tive Monte-Car­lo Poker Agent, 2009
Im­manuel Schweiz­er, Kamill Pan­itzek, Sang-Hye­un Park and Jo­hannes Fürnkranz
Knowledge Engineering Group, Technische Universität Darmstadt
Pro­ceed­ings of the 32nd An­nu­al Ger­man Con­fer­ence on Ar­ti­fi­cial In­tel­li­gence (KI-09), pages 65--72, Springer-Ver­lag, 2009

Abstract: The poker agent AKI-REALBOT described in this paper was designed to participate in the 6-player Limit competition which was part of the Computer Poker Challenge at the AAAI 2008 conference.It ended up in second place, its performance being mostly due to its ability to exploit weaker bots and still play fairly well against stronger players. This paper describes the architecture of the program and the Monte-Carlo decision tree-based decision engine that was used to make the bot’s decision. It will focus the attention on the modifications which made the bot successful in exploiting weaker bots.

Download: http://www.ke.tu-darmstadt.de/publicati ... 009-02.pdf