To understand how tire and pit strategy affect race outcomes, I built a linear regression model based on the base pace and degradation rate for every driver, then integrated along the average stint length to return an efficiency score. The results showed HARD tires averaged 3.55% higher efficiency than SOFT and 2.03% higher than MEDIUM. Alongside this, I developed an XGBoost regression model trained on tire age, compound, and track conditions to simulate full-race pit strategies, which was validated with race-grouped cross-validation against a 10-race holdout set. The model came within 1.2% (89.8s MAE) of actual outcomes.
Texas A&M spans over 5,000 acres, and as a result, most students have not been fully exposed to the campus. To tackle this issue, we drew inspiration from GeoGuessr, a game that places you in a random location worldwide and challenges you to guess where you are. With over 1,500 players, this game successfully exposed students to new places on campus, motivated them to explore in person, and helped them discover new third spaces.
