Authoring Adaptive Game World Generation

Abstract

Current research on adaptive games has mainly focused on adjusting difficulty in a variety of ways, for example, by providing some control over adaptive game world generation. These methods, however, are mostly ad-hoc and require quite some technical skills. To the best of our knowledge, so far, there has been no adaptive method that is truly generic and explicitly designed to actively include game designers in the content creation loop. In this paper, we introduce a generic method that enables designers to author adaptivity of game world generation, in a very expressive and specific fashion. Our approach uses adaptation rules, which build atop gameplay semantics in order to steer the online generation of the game content. Designers create these rules by associating skill profiles, describing skill proficiency, with content descriptions, detailing the desired properties of the specific game world content. This game content is then generated online using a matching and retrieval approach. We performed user studies with both designers and players and concluded that adaptation rules provide game designers with a rich expressive range to effectively convey specific adaptive gameplay experiences to players.