While SRS can go a long way towards calculating a difficult problem like schedule strength, it has its pitfalls. Logistic-Regression/Markov Chain (LRMC) ratings provide an alternate approach.
The previous post in this series covered how LRMC works and its mathematical implementation. Now it's time to take things a step further and apply this concept to ranking college football teams.
You might wonder, why is it so difficult to calculate some of these metrics? We have games and data, right? While that may be true, there are several factors that serve to complicate things this season.
As always with machine learning models, a common question that comes up is “Why?”. Why did the model choose this team to win vs. the other? What variables are the most and least influential in a given predictive model? Time to take a deep dive.
Predictive models for college football are a great application of machine learning techniques. Today, we'll look at one technique called gradient boosted decision trees using the LightGBM and NGBoost libraries.