Generating beginner heuristics for simple Texas hold'em

Fernando De Mesentier Silva, Frank Lantz, Julian Togelius, Andrew Nealen

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    Beginner heuristics for a game are simple rules that allow for effective playing. A chain of beginner heuristics of length N is the list of N rules that play the game best. Finding beginner heuristics is useful both for teaching a novice to play the game well and for understanding the dynamics of the game. We present and compare methods for finding beginner heuristics in a simple version of Poker: Pre-Flop Heads-Up Limit Texas Hold'em. We find that genetic programming outperforms greedy-exhaustive search and axis-aligned search in terms of finding well-playing heuristic chains of given length. We also find that there is a limited amount of non-transitivity when playing beginner heuristics of different lengths against each other, suggesting that while simpler heuristics are somewhat general, the more complex seem to overfit their training set.

    Original languageEnglish (US)
    Title of host publicationGECCO 2018 - Proceedings of the 2018 Genetic and Evolutionary Computation Conference
    PublisherAssociation for Computing Machinery, Inc
    Pages181-188
    Number of pages8
    ISBN (Electronic)9781450356183
    DOIs
    StatePublished - Jul 2 2018
    Event2018 Genetic and Evolutionary Computation Conference, GECCO 2018 - Kyoto, Japan
    Duration: Jul 15 2018Jul 19 2018

    Other

    Other2018 Genetic and Evolutionary Computation Conference, GECCO 2018
    CountryJapan
    CityKyoto
    Period7/15/187/19/18

    Fingerprint

    Genetic programming
    Teaching

    Keywords

    • Empirical study
    • Games
    • Genetic programming

    ASJC Scopus subject areas

    • Computer Science Applications
    • Software
    • Computational Theory and Mathematics

    Cite this

    De Mesentier Silva, F., Lantz, F., Togelius, J., & Nealen, A. (2018). Generating beginner heuristics for simple Texas hold'em. In GECCO 2018 - Proceedings of the 2018 Genetic and Evolutionary Computation Conference (pp. 181-188). Association for Computing Machinery, Inc. https://doi.org/10.1145/3205455.3205601

    Generating beginner heuristics for simple Texas hold'em. / De Mesentier Silva, Fernando; Lantz, Frank; Togelius, Julian; Nealen, Andrew.

    GECCO 2018 - Proceedings of the 2018 Genetic and Evolutionary Computation Conference. Association for Computing Machinery, Inc, 2018. p. 181-188.

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    De Mesentier Silva, F, Lantz, F, Togelius, J & Nealen, A 2018, Generating beginner heuristics for simple Texas hold'em. in GECCO 2018 - Proceedings of the 2018 Genetic and Evolutionary Computation Conference. Association for Computing Machinery, Inc, pp. 181-188, 2018 Genetic and Evolutionary Computation Conference, GECCO 2018, Kyoto, Japan, 7/15/18. https://doi.org/10.1145/3205455.3205601
    De Mesentier Silva F, Lantz F, Togelius J, Nealen A. Generating beginner heuristics for simple Texas hold'em. In GECCO 2018 - Proceedings of the 2018 Genetic and Evolutionary Computation Conference. Association for Computing Machinery, Inc. 2018. p. 181-188 https://doi.org/10.1145/3205455.3205601
    De Mesentier Silva, Fernando ; Lantz, Frank ; Togelius, Julian ; Nealen, Andrew. / Generating beginner heuristics for simple Texas hold'em. GECCO 2018 - Proceedings of the 2018 Genetic and Evolutionary Computation Conference. Association for Computing Machinery, Inc, 2018. pp. 181-188
    @inproceedings{54b91361a9b04e9c8f763d99da709134,
    title = "Generating beginner heuristics for simple Texas hold'em",
    abstract = "Beginner heuristics for a game are simple rules that allow for effective playing. A chain of beginner heuristics of length N is the list of N rules that play the game best. Finding beginner heuristics is useful both for teaching a novice to play the game well and for understanding the dynamics of the game. We present and compare methods for finding beginner heuristics in a simple version of Poker: Pre-Flop Heads-Up Limit Texas Hold'em. We find that genetic programming outperforms greedy-exhaustive search and axis-aligned search in terms of finding well-playing heuristic chains of given length. We also find that there is a limited amount of non-transitivity when playing beginner heuristics of different lengths against each other, suggesting that while simpler heuristics are somewhat general, the more complex seem to overfit their training set.",
    keywords = "Empirical study, Games, Genetic programming",
    author = "{De Mesentier Silva}, Fernando and Frank Lantz and Julian Togelius and Andrew Nealen",
    year = "2018",
    month = "7",
    day = "2",
    doi = "10.1145/3205455.3205601",
    language = "English (US)",
    pages = "181--188",
    booktitle = "GECCO 2018 - Proceedings of the 2018 Genetic and Evolutionary Computation Conference",
    publisher = "Association for Computing Machinery, Inc",

    }

    TY - GEN

    T1 - Generating beginner heuristics for simple Texas hold'em

    AU - De Mesentier Silva, Fernando

    AU - Lantz, Frank

    AU - Togelius, Julian

    AU - Nealen, Andrew

    PY - 2018/7/2

    Y1 - 2018/7/2

    N2 - Beginner heuristics for a game are simple rules that allow for effective playing. A chain of beginner heuristics of length N is the list of N rules that play the game best. Finding beginner heuristics is useful both for teaching a novice to play the game well and for understanding the dynamics of the game. We present and compare methods for finding beginner heuristics in a simple version of Poker: Pre-Flop Heads-Up Limit Texas Hold'em. We find that genetic programming outperforms greedy-exhaustive search and axis-aligned search in terms of finding well-playing heuristic chains of given length. We also find that there is a limited amount of non-transitivity when playing beginner heuristics of different lengths against each other, suggesting that while simpler heuristics are somewhat general, the more complex seem to overfit their training set.

    AB - Beginner heuristics for a game are simple rules that allow for effective playing. A chain of beginner heuristics of length N is the list of N rules that play the game best. Finding beginner heuristics is useful both for teaching a novice to play the game well and for understanding the dynamics of the game. We present and compare methods for finding beginner heuristics in a simple version of Poker: Pre-Flop Heads-Up Limit Texas Hold'em. We find that genetic programming outperforms greedy-exhaustive search and axis-aligned search in terms of finding well-playing heuristic chains of given length. We also find that there is a limited amount of non-transitivity when playing beginner heuristics of different lengths against each other, suggesting that while simpler heuristics are somewhat general, the more complex seem to overfit their training set.

    KW - Empirical study

    KW - Games

    KW - Genetic programming

    UR - http://www.scopus.com/inward/record.url?scp=85050587453&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=85050587453&partnerID=8YFLogxK

    U2 - 10.1145/3205455.3205601

    DO - 10.1145/3205455.3205601

    M3 - Conference contribution

    AN - SCOPUS:85050587453

    SP - 181

    EP - 188

    BT - GECCO 2018 - Proceedings of the 2018 Genetic and Evolutionary Computation Conference

    PB - Association for Computing Machinery, Inc

    ER -