Eric S. Smith '30 University Fellows Endowment Fund

Sports analytics are a constantly growing field, but little work has been done to incorporate analytics to video games. This summer through my project, I wanted to see if similar tools that are used in professional sports could be applied to sports video games, specifically to predict Rocket League game outcome. I utilized a variety of statistical learning methods to investigate which statistics were most positively or negatively associated with winning, and built a random forest model that classified games as wins/losses correctly with almost 90% accuracy.

Currently most general consumer prosthetics are basic devices with limited functionality and while sophisticated prosthetics with advanced motion and sensation do exist, they are at a price point out of reach for the general consumer. While there are groups working on increasing the motion and sensation capabilities of prosthetics, often there is not a focus on developing them at low price points and thus they are too expensive for the general consumer. That’s where my research comes in.