Evolving personas for player decision modeling

Christoffer Holmgard, Antonios Liapis, Julian Togelius, Georgios N. Yannakakis

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

    Abstract

    This paper explores how evolved game playing agents can be used to represent a priori defined archetypical ways of playing a test-bed game, as procedural personas. The end goal of such procedural personas is substituting players when authoring game content manually, procedurally, or both (in a mixed-initiative setting). Building on previous work, we compare the performance of newly evolved agents to agents trained via Q-learning as well as a number of baseline agents. Comparisons are performed on the grounds of game playing ability, generalizability, and conformity among agents. Finally, all agents' decision making styles are matched to the decision making styles of human players in order to investigate whether the different methods can yield agents who mimic or differ from human decision making in similar ways. The experiments performed in this paper conclude that agents developed from a priori defined objectives can express human decision making styles and that they are more generalizable and versatile than Q-learning and hand-crafted agents.

    Original languageEnglish (US)
    Title of host publicationIEEE Conference on Computatonal Intelligence and Games, CIG
    PublisherIEEE Computer Society
    ISBN (Print)9781479935468
    DOIs
    StatePublished - Oct 21 2014
    Event2014 IEEE Conference on Computational Intelligence and Games, CIG 2014 - Dortmund, Germany
    Duration: Aug 26 2014Aug 29 2014

    Other

    Other2014 IEEE Conference on Computational Intelligence and Games, CIG 2014
    CountryGermany
    CityDortmund
    Period8/26/148/29/14

    Fingerprint

    Decision making
    Experiments

    ASJC Scopus subject areas

    • Artificial Intelligence
    • Computer Graphics and Computer-Aided Design
    • Computer Vision and Pattern Recognition
    • Human-Computer Interaction
    • Software

    Cite this

    Holmgard, C., Liapis, A., Togelius, J., & Yannakakis, G. N. (2014). Evolving personas for player decision modeling. In IEEE Conference on Computatonal Intelligence and Games, CIG [6932911] IEEE Computer Society. https://doi.org/10.1109/CIG.2014.6932911

    Evolving personas for player decision modeling. / Holmgard, Christoffer; Liapis, Antonios; Togelius, Julian; Yannakakis, Georgios N.

    IEEE Conference on Computatonal Intelligence and Games, CIG. IEEE Computer Society, 2014. 6932911.

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

    Holmgard, C, Liapis, A, Togelius, J & Yannakakis, GN 2014, Evolving personas for player decision modeling. in IEEE Conference on Computatonal Intelligence and Games, CIG., 6932911, IEEE Computer Society, 2014 IEEE Conference on Computational Intelligence and Games, CIG 2014, Dortmund, Germany, 8/26/14. https://doi.org/10.1109/CIG.2014.6932911
    Holmgard C, Liapis A, Togelius J, Yannakakis GN. Evolving personas for player decision modeling. In IEEE Conference on Computatonal Intelligence and Games, CIG. IEEE Computer Society. 2014. 6932911 https://doi.org/10.1109/CIG.2014.6932911
    Holmgard, Christoffer ; Liapis, Antonios ; Togelius, Julian ; Yannakakis, Georgios N. / Evolving personas for player decision modeling. IEEE Conference on Computatonal Intelligence and Games, CIG. IEEE Computer Society, 2014.
    @inproceedings{82e80602f5124478a620b0a994e5f87b,
    title = "Evolving personas for player decision modeling",
    abstract = "This paper explores how evolved game playing agents can be used to represent a priori defined archetypical ways of playing a test-bed game, as procedural personas. The end goal of such procedural personas is substituting players when authoring game content manually, procedurally, or both (in a mixed-initiative setting). Building on previous work, we compare the performance of newly evolved agents to agents trained via Q-learning as well as a number of baseline agents. Comparisons are performed on the grounds of game playing ability, generalizability, and conformity among agents. Finally, all agents' decision making styles are matched to the decision making styles of human players in order to investigate whether the different methods can yield agents who mimic or differ from human decision making in similar ways. The experiments performed in this paper conclude that agents developed from a priori defined objectives can express human decision making styles and that they are more generalizable and versatile than Q-learning and hand-crafted agents.",
    author = "Christoffer Holmgard and Antonios Liapis and Julian Togelius and Yannakakis, {Georgios N.}",
    year = "2014",
    month = "10",
    day = "21",
    doi = "10.1109/CIG.2014.6932911",
    language = "English (US)",
    isbn = "9781479935468",
    booktitle = "IEEE Conference on Computatonal Intelligence and Games, CIG",
    publisher = "IEEE Computer Society",

    }

    TY - GEN

    T1 - Evolving personas for player decision modeling

    AU - Holmgard, Christoffer

    AU - Liapis, Antonios

    AU - Togelius, Julian

    AU - Yannakakis, Georgios N.

    PY - 2014/10/21

    Y1 - 2014/10/21

    N2 - This paper explores how evolved game playing agents can be used to represent a priori defined archetypical ways of playing a test-bed game, as procedural personas. The end goal of such procedural personas is substituting players when authoring game content manually, procedurally, or both (in a mixed-initiative setting). Building on previous work, we compare the performance of newly evolved agents to agents trained via Q-learning as well as a number of baseline agents. Comparisons are performed on the grounds of game playing ability, generalizability, and conformity among agents. Finally, all agents' decision making styles are matched to the decision making styles of human players in order to investigate whether the different methods can yield agents who mimic or differ from human decision making in similar ways. The experiments performed in this paper conclude that agents developed from a priori defined objectives can express human decision making styles and that they are more generalizable and versatile than Q-learning and hand-crafted agents.

    AB - This paper explores how evolved game playing agents can be used to represent a priori defined archetypical ways of playing a test-bed game, as procedural personas. The end goal of such procedural personas is substituting players when authoring game content manually, procedurally, or both (in a mixed-initiative setting). Building on previous work, we compare the performance of newly evolved agents to agents trained via Q-learning as well as a number of baseline agents. Comparisons are performed on the grounds of game playing ability, generalizability, and conformity among agents. Finally, all agents' decision making styles are matched to the decision making styles of human players in order to investigate whether the different methods can yield agents who mimic or differ from human decision making in similar ways. The experiments performed in this paper conclude that agents developed from a priori defined objectives can express human decision making styles and that they are more generalizable and versatile than Q-learning and hand-crafted agents.

    UR - http://www.scopus.com/inward/record.url?scp=84910088081&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84910088081&partnerID=8YFLogxK

    U2 - 10.1109/CIG.2014.6932911

    DO - 10.1109/CIG.2014.6932911

    M3 - Conference contribution

    AN - SCOPUS:84910088081

    SN - 9781479935468

    BT - IEEE Conference on Computatonal Intelligence and Games, CIG

    PB - IEEE Computer Society

    ER -