The Math
Power Digits is a completely unbiased, mathematical way of ranking college football teams. Its goal isn't to predict who will win future games — it's to determine who has the best résumé for the College Football Playoff.
There is no right way to rank college football teams. This is simply one way to do it — one that will sometimes agree with the pollsters and the CFP Committee, and often won't.
Every FBS team receives a Power Digits Rating between 0 and 100, which determines its rank. 100 is the best possible rating; 0 is the worst.
The five components
| Component | Weight | What it measures |
|---|---|---|
| Team W-L Rating | 55% | The team's own weighted wins and losses |
| Strength of Schedule | 25% | The weighted records of every opponent played |
| Top 3 Wins | 10% | The ratings of the three best teams beaten |
| Strength of Wins | 5% | The combined rating of the teams beaten |
| Strength of Losses | 5% | The combined rating of the teams lost to |
Power Digits = 100 × (0.55 × W-L Rating + 0.25 × Schedule + 0.10 × Top 3 Wins + 0.05 × Wins + 0.05 × Losses)
Team W-L Rating (55%)
The base of the algorithm is a weighted win percentage. Instead of every win or loss counting as exactly 1, each one counts for a little more or a little less depending on the strength of the opponent's conference and where the game was played.
Conference strength comes from how each conference has done in non-conference FBS games over the last three seasons plus the current one (for 2026, that's 2023–2026), counting every team in the conference it plays in this season. Records against teams at the same level (Power 4 vs. Power 4, Group of 5 vs. Group of 5) count for 40%, and records across levels count for 60%. Regular-season games count fully, bowl games count a quarter as much, and College Football Playoff games count ten times as much. A conference's base value is its strength plus 0.5.
That base value is what a neutral-site win over a team from that conference is worth. A road win adds 0.03 and a home win subtracts 0.03 — it's harder to win on the road in college football!
A loss is worth 1 divided by the base value, again adjusted by 0.03 for location: a home loss costs a little more and a road loss a little less.
Games against FCS teams count as half a win (0.5) or one and a half losses (1.5).
Strength of Schedule (25%)
The same weighted win/loss formula, applied to the results of every opponent — in their games against everyone except the team being rated. Teams benefit when their opponents go on to win games.
Opponents only count once the game has been played. A late-season rivalry game against a juggernaut isn't part of the résumé yet.
Top 3 Wins (10%)
The average W-L Rating of the three best FBS teams beaten. This rewards teams that have beaten several strong opponents, rather than teams carried by playing (and losing to) tough schedules. The average is always taken over three, so a team with fewer than three FBS wins is held back.
Strength of Wins & Strength of Losses (5% each)
Strength of Wins is the combined W-L Rating of the teams beaten; a winless team receives a 0. Strength of Losses is the combined W-L Rating of the teams lost to; an undefeated team receives a perfect score.
These ratings aren't meant to say "Team X is better than Team Y." They say "Team X deserves a playoff spot before Team Y does." In short, the best outcome for a team is to win games (of course) — and then hope its opponents go on to win games too.
In the early weeks the algorithm has very few results to compare teams on. By the end of the season there are 12 results for each team and roughly 130 opponent results to measure how strong a team really is.
FAQ
Why do the rankings look so "wrong" for the first few weeks?
The rankings only learn from game results, so they improve with every game played. In week 1, the algorithm may only know that Maryland beat Eastern Michigan. By weeks 2, 3 and 4, as both teams play other opponents, it learns how good Eastern Michigan really is — and as Eastern Michigan wins games, Maryland's rating gets a bump. By week 12, Maryland has ten opponents, each with their own ten results.
If we know Alabama is better than Akron, why do they start with no advantage?
Fans may think they know Alabama will be better than Akron, but it has to be proven on the field. The rankings use only the current season's results, so there are no preconceived biases.
College football is volatile. Michigan State went 12-2 with a playoff appearance in 2015, then 3-9 in 2016 — and that was before the transfer portal and today's coaching carousel. Power Digits gives no credit or penalty for previous seasons; every year starts with a blank slate. Do we really know Alabama is better than Akron, or are we just assuming?
Does margin of victory matter?
No. A win is a win, whether it's by 1 point or 50. Only the result, the location and the strength of the opponent count. Points are not 1-to-1 in games, every gameday comes with it's own challenges and situations, your job as a team is to win that day.
How can conference strength work when conferences change so much?
Conference strength uses the non-conference results of the teams currently in each conference, going back three seasons. Texas's past results count for the SEC, Cincinnati's count for the Big 12, and so on.
Do teams that go undefeated against easy opponents get ranked higher?
No. Strength of Schedule carries a lot of weight, and the strength of the teams beaten even more. An undefeated team does benefit from its record, but 45% of its rating is held back by weak opponents.
Do Group of 5 teams benefit from Power Digits?
No — they actually have it harder, because their conference games hurt their weighted records. A win over another Group of 5 team is worth less than one win, and a loss is worth more than one loss. An undefeated Group of 5 team can climb near the top, but after a loss they rarely crack the top 10. With a usually low strength of schedule, they need an almost flawless season.
How do these compare to the AP Poll and the CFP rankings?
Power Digits is an independent ranking and should be seen as something different from the AP Poll and the College Football Playoff committee's rankings. There is no 100% accurate way to rank teams unless they all played each other. Power Digits is a bias-free, alternate way to rank them.
Is scheduling the worst Power 4 teams the best way to beat the algorithm?
To a point, yes. A road win at a struggling Power 4 team credits your W-L Rating the same as a road win at a powerhouse from the same conference. But the difference shows up in Strength of Schedule, Strength of Wins and Top 3 Wins, where a strong opponent's success boosts your rating and a weak opponent's struggles drag it down.
Why do undefeated teams get such a boost?
They're undefeated. If nobody they played could beat them, they should be rewarded. Until a team loses, we can only guess how good it truly is.
How are FBS independents classified?
Independents are grouped with the conference they're most closely tied to, or their previous conference: Notre Dame with the ACC and UConn with the American. Independents never receive a conference's automatic bid in the playoff brackets.
Two teams who played each other are ranked next to each other. Why is the loser higher?
Across all the parts of the rating, those two teams happened to land side by side. Power Digits ranks résumés, not who-beat-who. Swapping teams for head-to-head results can create a never-ending loop, so it's never implemented.
How is this different from the old BCS computer rankings?
Like the computer rankings used in the BCS era, Power Digits is purely mathematical and ignores margin of victory and preseason expectations. Unlike the BCS, there's no human poll blended in — and the equations is replicatable & intentionally simple.
How do games against FCS teams count?
A win over an FCS team counts as half a win (0.5), and a loss to one counts as one and a half losses (1.5). For Strength of Schedule, an FCS opponent counts as if it were the lowest-rated FBS team.
How often are the rankings updated?
Every morning during the season, automatically, using the latest game results from CollegeFootballData.com.
What do the green and red numbers on the rankings page mean?
In each stat column, green marks the top 20 teams in that stat and red marks the bottom 20. Hover over (or tap) a number to see the games behind it.
What are the Playoff Formats?
They show what the College Football Playoff would look like if it started today, using Power Digits to pick and seed the field — for the current 12-team format and for alternatives with 4, 8, 20, 24 and 32 teams. Each bracket explains its own selection rules.