Procedural Progression Model for Smash Time

This paper addresses the problem of improving the player experience in single player endless games and encouraging the player to be engaged with procedurally generated content for longer periods of time. We present a model in which the procedural content generation process takes into consideration the dynamics of two dimensions of the player experience: the performance of the player when overcoming the challenges created by the game, and the variety of challenges presented to the player over time. We discuss the implementation of the model in the endless mode of the mobile game Smash Time, and describe how its evaluation supports that the model was able to increase both the number and duration of play sessions as well as having a better game experience reported by the participants, when compared to the original game. These results suggest that this approach could improve replayability and, as a consequence, the lifetime of a digital game.