This thesis aims to create a model to predict the expected value of a given shot in women's NCAA Division III basketball. For each shot taken, we collect information on several variables relating to the shot. Some of these variables include whether the offensive team was the home team or the away team, the type of shot (i.e. off rebound, post move, off pass, etc.), distance from the basket, and offensive and defensive shooting percentage. The resulting data set contains approximately 2200 data points from 20 games in the 2016-17 season. We then use two-fold cross validation and forward stepwise regression to create several models based on court location that predict expected shot value. Division III women's basketball coaches will find this research particularly useful, as it will help with implementing offensive and defensive research. We conclude by examining common scenarios that may occur during a game and calculating the expected shot value in those scenarios.


Pasteur, Drew




Applied Statistics

Publication Date


Degree Granted

Bachelor of Arts

Document Type

Senior Independent Study Thesis



© Copyright 2018 Paige McKean