My dissertation ties together my interests in level design, design patterns, game data analysis, and player modeling.

Level designers create gameplay through geometry, AI scripting, and item placement. There is little formal understanding of this process, but rather a large body of design lore and rules of thumb. As a result, there is no accepted common language for describing the building blocks of level design and the gameplay they create. This dissertation presents a set of level design patterns for first-person shooter (FPS) games, providing cause-effect relationships between level design patterns and gameplay. These relationships are explored through analysis of data gathered in an extensive user study.

Our goal is to better understand how designers use level design to create gameplay, and give them tools that aid their craft. For this research we choose to examine single player FPS levels, as it is a popular and well-established genre. Based on discussion with designers and extensive analysis, we formed a theory of how design patterns are used to create gameplay in FPS levels. We identified recurring elements in levels and the designer's intent in using them to affect the pacing, tension, and challenge of a level. We also explored ways these patterns could be varied, and how that would alter the basic effects of the patterns. The resulting pattern collection can be seen here.

Explicitly using the identified patterns, we built levels for a mod of Valve's Source Engine (a.k.a. the Half-Life engine). This mod serves as the testbed to collect data on how player behavior is affected by these patterns. Analysis of this data gave insights into these connections between level design and gameplay.

This work is the first scientific study of level design, laying the foundation for further work in this area. Data driven approaches to understand gameplay have been attempted in the past, but this work takes it to a new level by showing specific cause-effect relationships between the design of the level and player behavior.

The result of this dissertation is a resource for designers to help them understand how they are creating gameplay through their art. The pattern collection allows them to explore design space more fully and create richer and more varied experiences.

Get the final version of my dissertation here!