Bringing Sanity to March Madness


  • Austin Tice
  • Alexi Marti
  • Kendall Land
  • Max Layer


Every year come mid-March, the NCAA College Basketball tournaments start and mania promptly ensues. Games that should be blowouts become nail-biters, upsets happen, and a few underdog teams become what are known as “cinderellas”. All of which shows how it has earned the name March Madness®. Our team plans to address the problem of being able to predict the outcome of a game in the tournament. Our goal is to try and quantify and/or explain a team’s ability to “stay in a game”, their competitiveness, and their “cinderella-ness” based on how they performed in the regular season. Then, using the attributes we determine for each team, can we predict our overall goal of who will win each game in the tournament. We will be using the dataset provided by Kaggle, which includes data on how both Men’s and Women’s basketball teams performed and placed in tournaments in previous years. Our methodology will be strongly rooted in exploratory data analysis with a strong emphasis on clean, properly processed data to try and achieve the most accurate predictions from our machine learning models.






Computer Science