HOW TO READ THIS SITE

Most rankings ask “who would win on a neutral field?” This site asks a different question: who has earned it? That's the question the playoff committee is supposed to answer, and it's a résumé question — what did you actually do, against whom, and how convincingly.

The catch with most résumé rankings is that you can't see inside them. Ours shows its work: click any team and every game gets a line item, and those line items add up exactly to the team's score. We call that the receipts. If you ever wonder “why is this team ranked 8th?”, the answer is always the same: a list of games you can point at.

One number, two ingredients. Rating measures how good a team has looked — every game, weighted by how thoroughly they controlled it, adjusted for who they played and where. WAB measures what the record is worth. Add the two together (each scaled so they pull equal weight) and you get the Résumé Score that orders the board. Looking great while beating nobody won't carry you; grinding out wins while looking shaky won't either. You need both.

A team's rating is set so it's consistent with every game they've played — beat good teams convincingly and it rises, squeak past bad ones and it sags. Margin matters, but not the way you'd think: we score control, not style points. Leading wire-to-wire counts for more than a garbage-time cover. And because we re-price the whole season every week, a September win over a team that turns out to be great keeps getting more valuable in November. The number is centered so 1500 means an average FBS team. You'll meet this again on the receipts: the Performance (Perf) column is the Rating, one game at a time.

Imagine a team that's good enough to be in the playoff conversation but not safely in — a bubble team. WAB asks: playing your exact schedule, how many games would that team be expected to win, and how many did you win? Beat someone a bubble team almost always beats, and you get almost no credit. Beat someone a bubble team usually loses to, and you bank most of a win. It's the cleanest single number for “what has this record actually earned?” — a 9-3 record against a brutal schedule can out-WAB an 11-1 record against nobody. This one also lives on the receipts: the Strength of Win (Str) column is WAB, one game at a time.

Strength of schedule — just the average rating of everyone you've played. The score already accounts for schedule through WAB, so SOS is context: it's here to help you read the board, not to double-count difficulty.

Here's where it all connects: the receipts are the same two ingredients, taken apart game by game. Performance (Perf) is the Rating's share of one game: how well you played, given the opponent and the venue. Strength of Win (Str) is WAB's share: how much this particular result was worth against the bubble. And Total Résumé Value (Tot) combines them exactly the way the Résumé Score combines Rating and WAB. Same recipe at every scale — one game or a whole season.

All three are on a 0–100 scale, and it's a fixed one: 50 is an average team-game, roughly 70 is the top fifth, and 90+ is a performance almost nobody has. The scale is pinned to every FBS game played since 2022, not to this week or this season — so a 72 in September means exactly what a 72 in December means, and you can compare a 2026 game to a 2023 one without adjusting anything in your head. Str runs wider than Perf on purpose: winning and losing is a yes-or-no thing, so its scores bunch toward the ends, while Perf spreads out in the middle the way quality of play actually does. Tot lands close to the average of the two — near enough to eyeball, not exact, because each is scaled to its own spread.

That's what makes the ± column possible: each game's exact contribution to the team's score, and the honesty check — the rows sum to the score at the top, to the penny, every time. The receipts assemble the ranking, one game at a time. (Games against non-FBS opponents show greyed out: they count for the win-loss record, but a résumé is built on FBS competition.)

The Factors page asks who a team is, rather than who's earned it. Rush offense, pass offense, rush defense, pass defense, each rated the same opponent-adjusted way. These are descriptive: they tell you a team's identity, and they deliberately do not feed the rankings. A team can be a top-10 résumé with a mediocre pass defense — identity and résumé are just different questions, each with its own honest answer. (Pass ratings for run-first service-academy-style offenses show in grey — there just isn't much passing to rate.)

After two weeks of football, nobody's résumé means much — every ranking system that pretends otherwise ends up ranking whoever scheduled the softest openers. So early boards lean on a preseason-informed model rating and hand over to the résumé on a fixed schedule as real evidence piles up: mostly model in weeks 1-2, mostly résumé by week 6, résumé alone from week 7 on. The meter above the board always shows the current mix, and a team's receipt carries a “model prior” line for exactly the share the model contributes — even September gets receipts. Teams whose play is wildly out of line with expectations (in either direction) get eased onto their résumé a bit sooner, so the handoff doesn't jolt.

If the season ended today applies the actual format to the current board: five highest-ranked conference champions get automatic bids (before title games exist, the highest-ranked team in each conference holds its spot — those are the gold seeds), seven at-larges fill the rest, straight seeding. (2022-24 didn't actually run under this format — it was 4 teams in 2022-23, and 2024's byes went strictly to conference champs. We rebuild those seasons under today's rules so every year on this site compares apples to apples.) Projected end of season runs thousands of simulated seasons — every remaining game, every title race — and reports honest probabilities. An 85% favorite misses one time in seven; that's the sport.

The bracket and the probabilities answer different questions. The bracket is the field this board says a team earned. The probabilities are our forecast of what the committee will do. A team can be seeded 6th here and still be a long shot to actually get in.

The playoff field is picked by a committee of people, so predicting it means predicting them. We went and measured how they actually decide: every pair of teams in four years of final rankings, checked against what our board said. Two things came out clearly.

They weight the record far above the rating. Roughly four parts WAB to one part Elo, where we weight them evenly. And they pay a premium for a handful of marquee wins — beyond the credit those wins already earn. A team with two enormous scalps and three bad losses will be ranked above a team that was quietly better in all thirteen games.

Our view is that this is a real bias, and it's a defensible one to disagree with. Big wins already count enormously here — a win over a great team is worth many times a win over a bad one, in both Elo and WAB. The committee counts them again on top of that. The trouble is that beating one great team tells you less about how good you are than thirteen games do, and the extra weight doesn't predict how a team plays in January.

Plenty of fans will side with the committee here, and we understand why — a signature win is the most memorable thing a season produces, and it feels like it should count double. This site just doesn't score memory.

Credit where it's due, though: rewarding marquee wins gives programs a real reason to schedule hard non-conference games, and the sport is better for it. A system that only counted the full body of work would quietly reward ducking. So the committee's thumb on the scale buys something — we just think it costs more than it buys, and we show you both numbers so you can decide.

Games sorted by stakes — how much playoff and title-race probability actually swings on the outcome. Tap a game and you get the rooting guide: who should want what, including the teams that aren't playing. Some of the answers will surprise you; the fun ones are why the feature exists.

The slider above the board replays the season week by week: each week is an independent solve on only the games played to that point, receipts included, not today's rankings recolored. It's the best way to watch a résumé get built or come apart.