College football is the most inefficient major betting market in American sports — and that's exactly why we're building a model for it. With 130+ FBS teams, the books can't dedicate the same resources to pricing Missouri vs. Vanderbilt that they give to Chiefs vs. Bills. That pricing gap is where Signal Labs thrives.

Our college football model is built on SP+ (Bill Connelly's efficiency ratings), FPI (ESPN's Football Power Index), and raw EPA/play data. We layer in transfer portal adjustments — the single biggest source of preseason mispricing — along with coaching changes, returning production percentages, and strength of schedule.

Home field advantage is MASSIVE. In the NFL, home field is worth about 1.5 points. In college football, it's 3-4 points on average — and at places like Death Valley, The Swamp, or Autzen Stadium, it can be worth 5-7 points. The market consistently underprices this, especially for mid-major home teams against Power 4 opponents.

The transfer portal changed everything. Before 2021, returning production was the most stable predictor of team quality year-over-year. Now, roster turnover via the portal means teams can gain or lose 15+ players in an offseason. Our model tracks portal impact by position and projects how new additions integrate. This is the single biggest edge in college football analytics right now.

Full coverage across the Power 4 (SEC, Big Ten, Big 12, ACC) and select Group of 5 games where we identify value. Conference championship and College Football Playoff analysis will be published weekly once the season starts. We'll also track early Heisman odds and identify value in futures markets.