2018/07/01 by Baoxiang Wang, Wang, Baoxiang, Tongfang Sun +3
Computer Science · Psychology · Social Sciences · #Artificial Intelligence (cs.AI) #Artificial Intelligence in Games #Digital Games and Media #Educational Games and Gamification #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Reinforcement Learning in Robotics
paper · pdf · doi:10.48550/arxiv.1807.00366
openalex publication_date 2018/07/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
In recent years, reinforcement learning (RL) methods have been applied to\nmodel gameplay with great success, achieving super-human performance in various\nenvironments, such as Atari, Go, and Poker. However, those studies mostly focus\non winning the game and have largely ignored the rich and complex human\nmotivations, which are essential for understanding different players' diverse\nbehaviors. In this paper, we present a novel method called Multi-Motivation\nBehavior Modeling (MMBM) that takes the multifaceted human motivations into\nconsideration and models the underlying value structure of the players using\ninverse RL. Our approach does not require the access to the dynamic of the\nsystem, making it feasible to model complex interactive environments such as\nmassively multiplayer online games. MMBM is tested on the World of Warcraft\nAvatar History dataset, which recorded over 70,000 users' gameplay spanning\nthree years period. Our model reveals the significant difference of value\nstructures among different player groups. Using the results of motivation\nmodeling, we also predict and explain their diverse gameplay behaviors and\nprovide a quantitative assessment of how the redesign of the game environment\nimpacts players' behaviors.\n