2/4/25
One of our favorite analyses is the “Team Recaps”. This is the type of data that the algorithm uses to refine and calibrate ratings for each team. It’s always striving to create a “perfect” scenario where it matches the rating for opponents to maximize the win percentage of the algorithm. As you view the analysis below, if every team loses to higher rated team and wins against lower rated teams, the algorithm reaches “ground truth” or perfection. Of course, this will never happen in a population the size of every game this season for D3 men. But the algorithm strives to get the closest to it.
Here is a good example: Cal Lutheran defeated UC Santa Cruz this season. UC Santa Cruz defeated Mount Union. And Mount Union defeated Cal Lutheran. It’s virtually impossible to put ratings together that produces “ground truth” as it does not exist.
We hope you enjoy this presentation.