Dynamic Difficulty Adaptation for Heterogeneously Skilled Player Groups in Multiplayer Collaborative Games

Master thesis German OPEN
Greciano, Miguel Cristian (2016)
  • Subject:
    acm: ComputingMilieux_PERSONALCOMPUTING

This work focuses on the combination of two key concepts: Dynamic Difficulty Adjustment/Adaptation (video games adapting their difficulty according to the in-game performance of players, making themselves easier if the player performs poorly or more difficult if the player performs well) and Collaborative Multiplayer Games (video games where two or more human players work together to achieve a common goal). It considers and analyzes the challenges, potential and possibilities of Dynamic Difficulty Adaptation in Collaborative Multiplayer Games, which has to date been quite unexplored. In particular, it addresses the heterogeneously skilled player groups challenge: players with different skill levels play together in a video game, but how should the game adapt its difficulty if one player performs well when another performs badly? We use previous research on Dynamic Difficulty Adaption in single player games to mold, define and classify general approaches to Dynamic Difficulty Adaptation in collaborative games. We then focus on a subgroup of collaborative games - distinct-role collaborative video games, where we believe there is a viable way to address the heterogeneously skilled player groups challenge. To test our general approach to Dynamic Difficulty Adaptation in collaborative games we present a game that has been exclusively developed for this purpose: Co-op Craft. It is a collaborative game that uses the StarCraft II™ engine and includes three Dynamic Difficulty Adaptation Algorithms. We analyze, justify and classify these algorithms and we also outline other valid alternatives. We had players from different gaming backgrounds test this game both with Dynamic Difficulty Adaptation active and with Dynamic Difficulty Adaptation disabled. We analyzed their performance in both cases and asked their opinion on the matter based on their experience. From the numerical data representing their performance and their impressions we then extract conclusions about our approach and also the potential of including Dynamic Difficulty Adaptation in popular collaborative games of today.
Share - Bookmark

  • Download from
    tuprints via tuprints (Master thesis, 2016)
  • Cite this publication