Predictive Modeling of the University of Vermont Women’s Basketball Games

Conference Year

January 2020

Abstract

Every year the University of Vermont Women’s Basketball team plays 16 games for the America East Conference. Data was collected over two seasons to create a linear model designed to predict the score of a game. Zone percentages were also collected in addition to score-book statistics. Variables in the model were first picked based on linearity, then finalized using Akaike’s Information Criterion. The resulting model was then evaluated in R. This model is intended to help the UVM Women’s Basketball Team understand what the most important factors of their games are.

Primary Faculty Mentor Name

Bernard Cole

Status

Undergraduate

Student College

College of Engineering and Mathematical Sciences

Program/Major

Statistics

Primary Research Category

Engineering & Physical Sciences

Abstract only.

Share

COinS
 

Predictive Modeling of the University of Vermont Women’s Basketball Games

Every year the University of Vermont Women’s Basketball team plays 16 games for the America East Conference. Data was collected over two seasons to create a linear model designed to predict the score of a game. Zone percentages were also collected in addition to score-book statistics. Variables in the model were first picked based on linearity, then finalized using Akaike’s Information Criterion. The resulting model was then evaluated in R. This model is intended to help the UVM Women’s Basketball Team understand what the most important factors of their games are.