Opponent Modelling and Search in Poker
نویسندگان
چکیده
Poker is a challenging domain that contains both elements of chance and imperfect information. Though progress has been made in the domain, there is still one major stumbling block on the way to creating a world-class calibre computer player. This is the task of learning how an opponent plays (i.e., opponent modelling) and subsequently coming up with a counter-strategy that can exploit that information. The work in this thesis explores this task. A program is implemented that models the opponent through game play and then plans an exploitive counter-strategy using expectimax search. This program is evaluated using two different heads-up limit poker variations: a small-scale variation called Leduc Hold’em, and a full-scale one called Texas Hold’em.
منابع مشابه
Integrating Opponent Models with Monte-Carlo Tree Search in Poker
In this paper we apply a Monte-Carlo Tree Search implementation that is boosted with domain knowledge to the game of poker. More specifically, we integrate an opponent model in the Monte-Carlo Tree Search algorithm to produce a strong poker playing program. Opponent models allow the search algorithm to focus on relevant parts of the game-tree. We use an opponent modelling approach that starts f...
متن کاملIdentifying Features for Bluff Detection in No-Limit Texas Hold'em
Abstract Poker is increasingly becoming an area of inter-Poker is increasingly becoming an area of interest in AI research, partly because of the complex qualities it exhibits which are absent from more traditionally studied games, such as chess. One of the most difficult but also most important aspects of poker is the need to infer information about your opponent while also handling his attemp...
متن کاملParticle Filtering for Dynamic Agent Modelling in Simplified Poker
Agent modelling is a challenging problem in many modern artificial intelligence applications. The agent modelling task is especially difficult when handling stochastic choices, deliberately hidden information, dynamic agents, and the need for fast learning. State estimation techniques, such as Kalman filtering and particle filtering, have addressed many of these challenges, but have received li...
متن کاملThe Effectiveness of Opponent Modelling in a Small Imperfect Information Game Contents
Opponent modelling is an important issue in games programming today. Programs which do not perform opponent modelling are unlikely to take full advantage of the mistakes made by an opponent. Additionally, programs which do not adapt over time become less of a challenge to players, causing these players to lose interest. While opponent modelling can be a difficult challenge in perfect informatio...
متن کاملArtificial Intelligence Techniques in Games with Incomplete Information: Opponent Modelling in Texas Hold'em
Games have been widely used as an application for Artificial Intelligence techniques because of their simplicity and well-defined rules but in the other hand, for their huge range of possible and complex strategies to reach the final objective. In the last years, Artificial Intelligence study applied to games has focused more in games with incomplete information and non-deterministic games wher...
متن کامل