NCAA Basketball: Seeking out title contenders by the numbers

A deep dive into historical data can provide some insight into what teams might have what it takes to win the NCAA Basketball Tournament.

Every year, there are a handful of articles that try to use math to solve the age old-riddle: is there a formula that can accurately calculate who is going to win the NCAA Tournament? The short answer to that question is a definitive “no”, but the longer answer is a bit more interesting.

Data patterns cannot be used to accurately predict every tournament champion. For this, I’m glad, because it’d be terribly boring if that was the case. In reality, however, there are historical data patterns that allow us to identify what teams have the best chance at making a run when March rolls around.

When it comes to learning in life, experience is often the best teacher – when it comes to analyzing statistics, the same holds true. The more data you have to analyze, the higher the potential accuracy of your projections and modeling. Thus, the more games that are completed this season, the more accurate the analysis will become when evaluating a team’s championship qualities.

After combing over piles and piles of data, I have compiled my magnum opus. Through both linear and non-linear regression analysis, I have identified the variables that have the strongest correlation factors to making deep March runs. The following is an unveiling of the statistics I have found to be the most indicative of championship pedigree, and the teams that fit the mold so far this season.