This independent study seeks to construct an in-game win probability model for NCAA Division I football. Using historical data, the model seeks to output the likelihood of winning based on events that have occurred within the observed game. To this end, we find the value of particular situations that teams may experience during the game, and how teams should theoretically respond to these scenarios. During competitive games, the value of 1st down with respect to field position is more than that of any other down played. The model also compares the worth of 4th down conversion attempts to the worth of field goal attempts and punting to determine the optimal decision on 4th down with respect to field position.


Pasteur, Drew



Publication Date


Degree Granted

Bachelor of Arts

Document Type

Senior Independent Study Thesis



© Copyright 2019 Jordan Dennis