Here are the reasons Hurricane Analytics is projecting the field for the 2013 NCAA Basketball tournament:
1. It’s a highly predictable event. Hurricane Analytics believes that many events can be anticipated if the right data questions are asked to build a good, predictive model. The college basketball selection process is well known, but not an exact science. However, we have constructed a logistic regression model to simulate the selection process and we expect to forecast the at-large members of the tournament with at least 90% accuracy.
2. Mike Watkins is a college basketball nerd. I always have been. My passion for the game is only exceeded by my drive to understand what makes hurricanes do what they do (and what happens to people and businesses when they affect land).
3. It’s verifiable. In other words, the numbers will speak for themselves. Because the total number of at-large entries into the tournament are known, accuracy can be determined with precision. My forecast will also be comparable to another entity, ESPN, to judge relative skill – can a model I built in a day compete with an entity who tries to predict the tournament all year long?