Data-Driven Classifications of Video Game Vocabulary (2303.07179v1)

Abstract: As a novel and fast-changing field, the video game industry does not have a fixed and well-defined vocabulary. In particular, game genres are of interest: No two experts seem to agree on what they are and how they relate to each other. We use the user-generated tags of the video game digital distribution service Steam to better understand how players think about games. We investigate what they consider to be genres, what comes first to their minds when describing a game, and more generally what words do they use and how those words relate to each other. Our method is data-driven as we consider for each game on Steam how many players assigned each tag to it. We introduce a new metric, the priority of a Steam tag, that we find interesting in itself. This allows us to create taxonomies and meronomies of some of the Steam tags. In particular, in addition to providing a list of game genres, we distinguish what tags are essential or not for describing games according to players. Furthermore, we provide a small group of tags that summarise all information contained in the Steam tags.