Exploring player behavior with visual analytics

In designing puzzle and educational games, it is critical to be able to understand player behavior, in order to provide feedback when a player needs help, or redesign a game to keep players on-task. However, building a system that can react to all possible player behaviors can be very time intensive, and if a redesign is needed, can be a wasted effort. We propose a novel visual analytic approach to analyzing playtest data to help the game design process, and demonstrate its application to BeadLoom Game. The approach helped the game developers identify uninterested players, and refine the game so players could get a better sense of how close they were to the puzzle’s solution.