What is it about?
We develop a mathematical programming model to select, rank, and seed 68 teams for the National Collegiate Athletic Association (NCAA) Men’s Basketball Tournament. The selections and seeding are the responsibility of a 10-person selection committee, which chooses from 351 Division I college basketball teams and considers many team-performance attributes. Our approach provides an unbiased initial selection and seeding that has been shown to closely align with the final decision of the selection committee, in recent years. The approach requires no training data from prior seasons or previous committee recommendations and is based on such team- performance attributes as strength of schedule, quality wins, and winning percentage. The method can be applied to other situations where decision makers need to select and rank alternatives based on multiple performance attributes.
Featured Image
Why is it important?
Broadcast and multimedia rights to the tournament’s 67 games were recently extended eight years for a total fee of $8.8 billion. These payments are justified by the widespread interest in the tournament, which is colloquially known as the Big Dance or March Madness. Moreover, athletic programs are compensated based on the number of tournament games in which their teams appear, which makes entering the tournament, and doing so with a favorable seed, an important revenue stream.
Perspectives
My co-author and I are both college basketball fans that have enjoyed March Madness for decades. This work gave us the opportunity to explore a topic of considerable personal interest and learn much more about it in the process. We had initially submitted a paper for ranking and seeding the teams once the Selection Committee had identified at-large bids. But the editor-in-chief Dr. Gorman pushed us to include the selection process as well. The result is a method that could provide a soup-to-nuts unbiased starting point that could assist the NCAA Selection Committee and similarly-charged committees. The idea here is that the committee would likely make changes from our recommended starting point, but they would have to articulate their reasons for doing so. This process would help them gain insight to their own decision-making criteria and avoid potentially biasing their decisions based on factors that they are explicitly to avoid, such as conference affiliation, past tournament performance, or program history.
Bruce Reinig
San Diego State University
Read the Original
This page is a summary of: Using Mathematical Programming to Select and Seed Teams for the NCAA Tournament, INFORMS Journal on Applied Analytics, February 2018, INFORMS,
DOI: 10.1287/inte.2017.0939.
You can read the full text:
Contributors
The following have contributed to this page
