Ian's annual dunk on ESPN's FPI: 2025 edition

I don't know why ESPN has FPI when they already have SP+ but, let's dive in anyways.

I’ve always strongly disliked FPI.

I never understood why ESPN developed it while simultaneously poaching Bill Connelly and his SP+ model. Naturally they’ve tended to promote FPI over it, which I find to be an inferior product. Also, when it was launched I ridiculed many of its projections as it had playoff outcomes that were not possible because of future games. At one point in 2014 they were insisting that Oklahoma, TCU, and Baylor were three of the best four teams, or some such nonsense. I followed the Big 12 and knew this was not the case. Also, they were all going to play each other and knock each other out of elimination so putting them all in the playoffs didn’t make much sense.

When I made this criticisms on X.com (formerly known as Twitter), I found ESPN writers in my mentions defending the product with some of them disparaging me for daring to think I could see something a model could not. This did not prove to be a lesson in humility for me.

My dislike of models and their proponents has only grown since then (cough COVID cough) but I don’t mind when we reference them as something to consider rather than something to treat with the certainty in which it presents itself with the ridiculous percentages. They can be a useful tool like anything else.

Let’s see what FPI has decided may be in store for the 2025 season and try to anticipate what it’s missing.