Modifying MCTS for human-like general video game playing

Ahmed Khalifa, Aaron Isaksen, Julian Togelius, Andy Nealen

    Research output: Contribution to journalConference articlepeer-review


    We address the problem of making general video game playing agents play in a human-like manner. To this end, we introduce several modifications of the UCT formula used in Monte Carlo Tree Search that biases action selection towards repeating the current action, making pauses, and limiting rapid switching between actions. Playtraces of human players are used to model their propensity for repeated actions; this model is used for biasing the UCT formula. Experiments show that our modified MCTS agent, called BoT, plays quantitatively similar to human players as measured by the distribution of repeated actions. A survey of human observers reveals that the agent exhibits human-like playing style in some games but not others.

    Original languageEnglish (US)
    Pages (from-to)2514-2520
    Number of pages7
    JournalIJCAI International Joint Conference on Artificial Intelligence
    StatePublished - 2016
    Event25th International Joint Conference on Artificial Intelligence, IJCAI 2016 - New York, United States
    Duration: Jul 9 2016Jul 15 2016

    ASJC Scopus subject areas

    • Artificial Intelligence


    Dive into the research topics of 'Modifying MCTS for human-like general video game playing'. Together they form a unique fingerprint.

    Cite this