Ratings Explained

The number

A team's Rating is the margin it would be expected to win by against an average FBS team on a neutral field. A +20 team should beat a +5 team by about 15 on a neutral field, or by about 18.0 at home, since home field is worth 3.0 points in the model. Zero is average. The best teams in a typical season land near +25 to +30; the worst near −25.

Opponent adjustment

Raw season averages, such as points per game or EPA per play, say nothing about who they came against, which is why a big week one against an FCS opponent can put a team at the top of an unadjusted list.

The Rating fixes that by treating every game as an equation. For each offense in each game, its EPA per play in that game equals its offensive strength, minus the opposing defense's strength, plus a home edge. Every FBS game of the season goes into one system of roughly 1,800 such equations, solved together, so gaining +0.30 EPA per play against the best defense in the country counts for far more than the same number against the worst. A second system does the same thing with points scored and allowed instead of EPA per play.

Adj Off is the points per game a team would score against an average FBS defense on a neutral field, and Adj Def is the points it would allow against an average offense. The two are solved together from points scored and allowed in every game, and Adj Off minus Adj Def is the Rating. Lower is better on defense. The same solve on EPA per play appears in each team's style profile as Adj. EPA per play.

Blending EPA and margin

The points-based offense and defense are blended with the EPA-based ones, scaled into points (about 70 points per one EPA per play, roughly the number of plays a team runs in a game). The blend was chosen by testing how well each mix predicted real games; EPA currently carries 10% of the weight and points the rest. Scores turn out to predict future scores slightly better than per-play efficiency does, but the EPA component steadies the numbers against a single fluky final.

Preseason prior

Before a season starts, every team gets a prior rating built from four things: its final rating from the previous season (kept at about 66% of its value, since teams regress toward average), its 247Sports talent composite, its recruiting class points, and the share of last year's production that returns. The weights were fit on a decade of seasons, so they reflect how much each input has actually predicted the next season's finish.

The prior is worth about 4 games of evidence. After week one it dominates; by midseason it is a small share of the rating, and by the bowls it is mostly gone. Teams new to FBS start below average, and FCS opponents are pooled into one shared unit that is rated like any other team.

How good is it

The model is checked by walking forward through past seasons a week at a time: fit the ratings on every game before that week, predict that week's games, then compare with what happened and with the closing point spread. The closing line is the hardest benchmark in the sport, and the ratings are not expected to beat it, only to stay close.

Beating the spread about half the time is the expected result for an honest model; anything much higher would mean the test was leaking future information. The early weeks miss by more than the late weeks, which is the price of leaning on the prior.

This season so far

The same test runs on 2026 after every data refresh: each week's games are projected from ratings fit only on the games before it, then scored once they are played. Through Wk 9, 485 FBS-vs-FBS games.

Early weeks lean on the preseason prior and miss by more; the season number should settle toward the historical one by midseason. A gap to the closing spread that keeps growing would mean something upstream broke, not that the model changed.

Game projections

On the current season's team pages, unplayed games show a projected spread and win probability. The spread is the rating difference plus the home edge, shown from the team's side with the usual convention: negative means favored. The win probability comes from how often a projected margin holds up in the backtest; actual margins scatter around the projection with a standard deviation of about 16 points, so a 7-point favorite wins roughly two-thirds of the time. The projected record is the current record plus the sum of those probabilities.

Upcoming games

The homepage strip and the Games page rank the coming week's games by a thrill score. It combines three things: the quality of the two teams (the sum of their Ratings, half the weight), how close the projection is (the favorite's win probability's distance from a coin flip, about a third), and tempo (the two teams' combined plays per game, the rest). Each is standardized against every FBS matchup of the season and the total is expressed as a percentile, so a 90 is a better game than nine in ten on the schedule. It says nothing about storylines, rivalries, or what is at stake; it is only what the numbers can see. The projected score is each team's adjusted offense against the other's adjusted defense, with half the home edge to each side, so the two scores always differ by the projected spread.

Resume: Wins Above Playoff

The Rating says how good a team is. The Resume page scores what a team has actually done, with Wins Above Playoff (WAP), the football version of college basketball's Wins Above Bubble. Picture a team right on the playoff bubble: the average of the teams ranked 10 to 14 by Rating (+16.4 in 2026). For every game a team has played, the same win probability used for projections says how often that bubble team would have won against that opponent at that site. A win is worth one minus that chance and a loss costs that chance, so beating a good team on the road adds a lot, beating a weak one at home adds almost nothing, and a loss costs more the more beatable the opponent. WAP is the sum: the team's wins minus the wins the bubble team would expect from the same schedule. Near zero is a bubble resume; margin of victory does not count.

Opponents are judged by their current Rating, not what they were on game day, so a win grows or shrinks as the opponent's season plays out. Every completed game counts, including FCS opponents, bowls, and playoff games. Bubble W-L is the record the bubble team would expect against the games played so far, which doubles as a strength-of-schedule measure (rank 1 is the hardest). Proj adds what the team is expected to gain or lose over its remaining games, using its own win probabilities.

Player of the Year

The Player of the Year page ranks every player on one number, in points: production above a replacement-level player at the position, adjusted for the opponents faced, then scaled by a position factor and by how much the team has won. The table shows the parts.

Production is the total EPA on the player's plays: throws and runs for a quarterback, runs and targets for a back, targets for a receiver. The passer and the receiver are both credited for the same completion, which is how the play-by-play data works and why each position is judged against its own replacement level. The play-by-play data has no EPA for defenders, so a defender's production is a points value put on each havoc play: a sack 1.8, a tackle for loss 1.0, an interception 4.5, a pass broken up 0.7, a hurry 0.5, a fumble recovery 3.0, a touchdown 5.0, and a little for every other tackle. A two-way player is one row in the data, so both halves add up.

Opp is the opponent adjustment. The player logs only say who played, not against whom a yard was gained, so the adjustment is the team's: the gap per play between an average defense and the defenses the team has faced (by Adj Def EPA, FCS opponents pooled), times the player's plays. Defenders get the mirror image, scaled by the offenses faced. Replacement is the 75th percentile of the qualified players at the position, so the rating counts only what a player adds over a good starter; this season that is 59 points for a quarterback, 16 for a back and 21 for a receiver.

The position factors and the team factor are fit to ten years of national player of the year voting (2016 to 2025), so the page ranks players the way voters tend to, not the way a pure value model would. Quarterbacks are the yardstick at ×1.00; backs get ×1.50, receivers ×1.25, defensive linemen ×3.00, linebackers ×2.00 and defensive backs ×0.25, which is the voters' taste for pass rushers over cover men and for quarterbacks over everyone. The team factor is 1 + 0.50 × (2 × win% − 1), with win% pulled toward .500 by two pseudo-games so an early-season start is not a full-season bonus. Over those ten seasons the fitted model puts the eventual winner in its top three in nine of them; the exception is 2024, when a two-way player won. The factors are refit once a year after the vote.

What it does not know: sacks taken by a quarterback, forced fumbles, snaps played, and anything about a defender beyond the box score. Early in a season the replacement level and the team factor move with every week, so a September list is a watch, not a verdict.

Coaches Hot Seat

The Hot seat page gives every head coach the chance he is not the school's head coach next season. It is not a judgment about whether he should be; it is how often coaches in the same spot have not been back, learned from every FBS coach-season since 1988 (4,052 of them, 14% of which ended with the coach gone). A coach who leaves for another FBS job is set aside in the training, so the number covers firings, resignations and retirements together; it cannot tell those apart.

The inputs are few and all on the page: tenure (seasons at the school, counted up to ten; after that it stops mattering), this season's record, scored as its projected final record while the season is on, last season's record, what the program expects (its winning percentage over the ten seasons before the coach was hired, pulled a little toward .500), the change in the team's margin-based strength since last season, and how tenure and this season's record interact, since a bad year costs a seventh-year coach more than a second-year one.

Held out from the fit, the seasons from 2016 on say how well it works: a coach who was gone the next year is ranked above one who stayed 78% of the time, and the percentages mean what they say, with the top tenth of coaches by the number gone 53% of the time against 50% predicted. Tiers: Cool under 10%, Warm to 25%, Hot to 50%, Scorching above that. It knows nothing about buyouts, contracts, athletic directors, recruiting or scandal, and a coach hired during the season is listed without a number.

2025 scorecard. The model's top 15 entering the 2025 hiring cycle, fit on seasons through 2024 only, and what happened: 5 of the 18 coaches who were not back are in its top 10.

Coach	School	Hot seat	Outcome
Mike Gundy	Oklahoma State	94%Scorching	Not back
Thomas Hammock	Northern Illinois	69%Scorching	Not back
Troy Calhoun	Air Force	64%Scorching	Stayed
Sam Pittman	Arkansas	49%Hot	Not back
Mike Norvell	Florida State	47%Hot	Stayed
Brent Pry	Virginia Tech	45%Hot	Not back
Mark Stoops	Kentucky	44%Hot	Not back
Chris Creighton	Eastern Michigan	40%Hot	Stayed
Mike Locksley	Maryland	39%Hot	Stayed
Dave Aranda	Baylor	38%Hot	Stayed
Shane Beamer	South Carolina	32%Hot	Stayed
Luke Fickell	Wisconsin	32%Hot	Stayed
Trent Dilfer	UAB	31%Hot	Not back
Jay Norvell	Colorado State	30%Hot	Not back
Billy Napier	Florida	30%Hot	Took another job

What it does not know

Injuries, suspensions, weather, and motivation are not modeled. Garbage time is not removed. Every game of the season counts the same regardless of how long ago it was played. FCS opponents are treated as one team, so a win over a strong FCS program is undervalued and a win over a weak one is overvalued. Treat the ratings as a well-calibrated starting point, not a betting sheet.

Lines and edges

The Lines tab of the Games page puts the market spread and over/under beside the projected spread and total for every game of the week, biggest disagreement first. Each game's market line comes from one source, chosen in a fixed order of preference. The numbers are captured once a day with the site's refresh, at about 5am Eastern, so they are that morning's lines rather than the kickoff lines. The opening line is shown under the current one wherever it has moved, and a played game keeps its closing line, so a finished week reads as a record of what happened against the number.

What is the edge?

The projected spread minus the market spread. If the market has a team favored by 7.5 and the ratings make it 11.2, that is a 3.7-point edge toward the favorite. Totals work the same way: the two projected scores added together against the posted over/under, as Over or Under. Edges of three points or more are in bold.

Which side does the model like?

The team named in the Edge column, which is the side of the market line the projection falls on, and Over or Under for the total. The win probability is the model's chance for the favorite, from the projected margin and the 16-point spread of misses. Under it is the market's chance for the same team, from the moneyline with the vig removed (each side's implied probability divided by their sum), and the gap in percentage points; the market often posts no moneyline on a big favorite, so that line is missing where the spread edge tends to be largest. On a played game the projection shown is the one published the morning of the game, and a check or a cross records whether that side covered.

How often does the model beat the closing line?

About half the time: 47.9% this season (468 games) and 50.5% over 2022–2025, against the 52.4% it takes to break even at standard −110 odds. That is a coin flip, and it should be: the ratings measure how good teams are, and the closing line is the sharpest public number there is. Read the edges as places where the ratings and the market disagree, not as bets.

Why is a game missing a line, and when do lines update?

The market posts a weekend's full slate about a week out, usually by the Tuesday or Wednesday before, so the coming week is normally complete while the week after shows only a few look-ahead lines on the biggest games. The refresh picks up new lines each morning, and every game that has been played shows its closing line.

Why are FCS opponents listed separately?

Every FCS team is rated as one pooled opponent, so the projection knows less about those games and their edges are the least trustworthy on the page. They are left out of the model's record above.

For information only, not betting advice: the site takes no responsibility for any wager placed on it.