A Player-Based Approach to Predicting March Madness Tournament Outcomes

In the U.S., the annual NCAA college basketball tournament, known as March Madness, draws in millions of people trying to predict who will win. Just one problem: no one has ever created a perfect bracket. By using a player-based rating system that updates throughout the season, a “predictive model” can be created to accurately predict teams with the best shot of winning the championship, and even show which players had the most impact on a single team in college basketball.