About THE MODEL.

Look — we love college football, but we also love numbers. That’s why we built a custom predictive model to cut through the media hype and tell you who’s actually good (and who’s just got a flashy social team and deep NIL pockets).

This model blends several key data sources to generate projected power ratings and spreads for every FBS team. Here’s what we factor in:

• Previous Season Power Ratings (SP+ based) – Because past dominance usually means something.

• Returning Production – Who’s coming back on offense, defense, and special teams? Continuity counts.

• Recruiting Rankings – Talent wins games. A stacked class — especially if true freshmen start right away — can shift the needle fast.

• Strength of Schedule – A 10-2 record against cupcakes ain’t the same as 10-2 in the SEC.

• Transfer Portal Impact – Game-changing additions (or brutal losses) are baked in.

We trained the model using ridge regression, which basically helps avoid overreacting to weird outliers (like that one time Hawaii was ranked #2). The goal? To get the most realistic, forward-looking rankings possible — and then turn those into projected spreads for every matchup.

Is it perfect? Hell no. No model is. But it sure as hell beats listening to another preseason take from a guy in khakis who thinks Iowa’s offense is “due.”

Note: Kennesaw State, Delaware, and Missouri State are excluded from our rankings due to a lack of sufficient historical data. Blame their transition status — not us. I'm sure that breaks your heart.

Want to see the numbers yourself? Hit “The Model” tab up top — it’ll open in a new window, so don’t freak out.