Procedural Level Generation to Improve a Digital Math Game's Development: Does it Impact Player Experience?

Procedural content generation can improve the game development process, however, few studies evaluated how it influences players, especially on digital math games. This work tackles this problem by investigating how procedural level generation influences players of an introduced digital math game. Additionally, we validated the game and analyzed both the relationship between fun, willingness to play the game again (i.e., returnance), and curiosity, and the impact of demographic and in-game data on player experience and performance. A two-sample experiment was designed where participants played a game version with (dynamic) or without PCG (static) in which in-game (n = 724) and questionnaire (n = 506) data were gathered and empirically analyzed. The results demonstrate the experiences of players from the dynamic version were similar to those of the static in all but one question, while being more difficult and providing equivalent engagement. The findings also show: the game is fun and arises players’ curiosity and returnance, players’ curiosity has a strong correlation to fun and returnance, and demographics and in-game performance impact players’ experiences. Our results are valuable to developers and designers, showing the impact of procedural level generation on players, and how and which factors might play a role in their experiences.