Team sports for Game AI benchmarking revisited

Sport games are among the oldest and best established genres of computer games. Sport-inspired environments,such as RoboCup, have been used for AI benchmarking for years. We argue that, in spite of the rise ofincreasingly more sophisticated game genres, team sport games will remain an important testbed for AIbenchmarking due to two primary factors. First, there are several genre-specific challenges for AI systems thatare neither present nor emphasized in other types of games, such as team AI and frequent re-planning. Second,there are unmistakable non skill-related goals of AI systems, contributing to player enjoyment, that are mosteasily observed and addressed within a context of a team sport, such as showing creative and emotionaltraits. We analyze these factors in detail and outline promising directions for future research for game AIbenchmarking, within a team sport context.

@Article { MPB21,
  author       = "Mozgovoy, Maxim and Preuss, Mike and Bidarra, Rafael",
  title        = "Team sports for Game AI benchmarking revisited",
  journal      = "International Journal of Computer Games Technology",
  number       = "5521877",
  volume       = "2021",
  month        = "may",
  year         = "2021",
  doi          = "10.1155/2021/5521877",
  note         = "https://doi.org/10.1155/2021/5521877",
  keywords     = "Sport games, Team sports, AI benchmarking, Game AI",
  url          = "http://graphics.tudelft.nl/Publications-new/2021/MPB21"
}