Predictive Modeling and Analysis of Sports Using Linear Algebra-based Models

This project will be run by Amanda Harsy.

​​

Ranking sports teams and forecasting post-season outcomes based on regular-season performance is a complex task. Among various mathematically driven ranking methods, linear algebra-based approaches—such as Colley, Massey, Keener, PageRank, and Markov Chain models—stand out for their simplicity and accessibility to undergraduate students with a background in linear algebra. While these models effectively rank teams, they often struggle with accurately predicting future game results. A potential enhancement to these models involves incorporating weighted factors or additional features into these models, along with cross-validation techniques to assess their predictive accuracy. In this research, we will develop and evaluate the predictive capabilities of linear algebra-based models using data from a sport of the student’s choosing.

​

For a pdf with more information, click here.

​​