Belmont University Research Symposium (BURS)

Anarchy? Nope. Just Probability.

Publication Date



Sciences and Mathematics, College of


Math and Computer Science, Department of

BURS Faculty Advisor

Dr. Christina Davis

Presentation Type

Poster Presentation


The yearly NCAA march madness tournament is one of the most difficult prediction problems, as a perfect bracket is effectively impossible. However, predicting the outcomes of each individual game is a much more surmountable challenge. In my project I use free in-season and tournament data from Kaggle’s march machine learning mania challenge and from public advanced stats websites in an attempt to assign a win probability to each possible matchup in the 2023 March Madness tournament, as well as probabilities of each team to advance for each round through a hierarchical model. I have built multiple predictive models and will compare them and examine what is driving their differences.

This document is currently not available here.