Video Game Development is characterized by lengthy design lifecycles that include the construction of interweaving systems that ultimately result in emergent experiences for players. These systems need to be correctly tuned on the initial approach, lest they result in player experiences that are frustrating and unfair. To remedy this, a common practice and term of art used in game development and design is that of “Balance,” which generally refers to the adjustment of different game elements to present an experience to a player that is fair, challenging, engaging, and full of a variety of gameplay choices. That being said, balance is often used intuitively and is reduced to singular definitions and quantitative measurements that miss out on the complexity of the term and practice overall. This can result in highly lengthy iteration cycles to get the overall feel of a game correct, sometimes massively outstripping the core development time of the game systems themselves. To complicate this, the goals, targets, and methods used to achieve balance can be haphazard and miss the mark of what an audience might want to see. Any automation or assistance to the balancing process could meaningfully reduce game development costs while increasing player enjoyment. In this proposal, I perform theory work to taxonomize balance definitions, identifying the many gameplay elements that can be targeted and the procedural methods that can be used to adjust them. I then present two implementations targeting several definitions of balancing in a procedural context, showing how different techniques can be used for various goals for game balance. I finally integrate quantitative and qualitative evaluations to understand if procedural methods are adequate to achieve the objectives of balance outlined in my taxonomy work.

Event Host: Samuel Shields, Ph.D. Candidate, Computational Media

Advisor: Edward Melcer