UC Santa Cruz

Complex games, those with multiple correct strategies and unpredictable outcomes, are seeing increased popularity and integration into high-impact domains such as health, education, and training. Further, even in entertainment contexts, these games have proven benefits for players. The steep learning curves, however, make the games inaccessible to many players, resulting in a lack of diversity in professional circles, cutting people off from the proven benefits of play, and rendering them ineffective in serious domains.

This work examines, in a user-centric manner and through the lens of the Cyclical Phase Model of Self-Regulated Learning, how players learn to play and master complex games and how we can better support these processes through computational support. These thesis, thus, contributes an advanced understanding of how learning and mastery occurs in complex gameplay, and empirical insights into how to support these processes, grounded in theories of learning.

In sum, this work contributes to making complex games more accessible and effective as high-impact tools and identifies generalizable insights for the design of computational support tools for learning relevant to adjacent domains including explainable and user experience of Artificial Intelligence.