How College Football Rankings Work: Polls, Committees and Computers
There is no single college football ranking. Learn how the AP poll, the coaches poll, the playoff committee and computer ratings each work, and why they disagree.
Two fans argue in a parking lot about who is the best team in the country. One quotes a poll. The other quotes a different poll. A third brings up a computer rating that disagrees with both. They are all citing real rankings, and none of them is wrong to do so. College football does not have one ranking. It has several, built by different people, for different reasons, with different rules.
That fact is worth understanding because, in college football, the ranking is not just commentary. For decades, rankings have helped decide who plays for a national championship. Today they still shape the playoff, bowl matchups, television attention and a team's sense of itself. Knowing who makes a ranking, and how, tells you how much weight to give it.
This guide covers the main types: polls voted on by writers and coaches, the committee that picks the College Football Playoff field, and the computer ratings that try to remove human opinion. It ends with practical advice for reading them.
Why college football needs rankings at all
The NFL does not need a ranking to pick its playoff field. Its teams all play a schedule built from a common formula, and the standings decide the playoffs. College football cannot do that. There are well over a hundred teams at the top level, grouped into conferences of different sizes, and no team plays more than a small share of the others. A team in one league may never meet a team from another. Schedules differ in strength: one team plays a hard conference schedule while another faces a weaker set.
With no common schedule, results cannot be compared directly. Somebody has to judge, and the judgment is what a ranking is. In the past, those rankings decided who was crowned national champion. Now they decide which teams reach a playoff. For the structure of the playoff itself, see our guide to the College Football Playoff format.
The AP poll
The Associated Press poll, often called the AP Top 25, is the oldest and best-known ranking. It began in 1936. A panel of sportswriters and broadcasters votes each week during the season, and the results are published for the public.
How it works
- Each voter submits a ballot ranking his or her 25 choices.
- A team receives 25 points for a first-place vote, 24 for second, and so on down to 1 point for a 25th-place vote.
- The points are added across all ballots. The team with the most total points is No. 1.
- The poll is released weekly during the season, with a preseason version before play begins.
The ballots are made public, so you can see how each voter ranked each team. The voters are not employed by schools. They are media members, and their number and selection are decided by the AP.
Strengths and weaknesses
The AP poll is a snapshot of the opinion of people who watch a great deal of football. It is simple to understand and widely reported. Its weaknesses are also well known.
- Preseason anchoring. The first poll of the year is based on what voters expect, because there are no results yet. Once placed, a team tends to move only when it wins or loses. A team that begins ranked No. 3 and wins ugly often stays above a team that began unranked and wins impressively.
- Limited viewing. No voter can watch every game, and the games watched are usually the biggest ones.
- Voter lean. Voters can favor teams from the conferences they cover most, or teams with famous names. Individual ballots are sometimes criticized.
The AP poll is also not used to select the playoff field. It is a poll of opinion, not an official selection process.
The Coaches Poll
The Coaches Poll is voted on by a panel of head coaches at the top level of the sport. It dates back to 1950 and has been sponsored by different news organizations over the years. Check the poll's own page for who administers it today.
It works much like the AP poll: each voter ranks 25 teams, and the points are added up. Two things distinguish it.
- The voters have a stake. Coaches are participants, and they compete against the teams they are ranking. Critics have long pointed out that a head coach has little time to watch other teams' games closely during the season, and that some ballots are prepared by staff members.
- The result tends to track the AP poll. The two lists are usually close, with differences of a spot or two in most weeks.
For many years, the AP poll and the Coaches Poll each named their own national champion. In some seasons the two disagreed. In 1997, for example, the AP poll ended with Michigan at No. 1 and the coaches' poll chose Nebraska. Those split verdicts were part of the reason the sport moved toward a more formal system.
The BCS era: polls plus computers
From the 1998 season through the 2013 season, the sport used the Bowl Championship Series (BCS). The BCS tried to decide a title-game matchup by combining human polls with computer ratings. The two top-ranked teams in a combined formula played in the national championship game.
The BCS is gone, but it is worth remembering because it made computer ratings famous, and because the arguments it started shaped what followed. Fans disagreed about how much weight the polls and computers should receive, and about whether the formula was too secretive. After the 2013 season the BCS was replaced by a selection committee. For the history, see from polls to playoff: the history of the college football championship.
The College Football Playoff committee
Since the 2014 season, the teams that compete for the national championship are chosen by a selection committee of the College Football Playoff. This is the ranking that matters most today.
Who is on the committee
The committee is a group of people with football experience: athletic directors, former coaches, former players, university administrators and, historically, a former journalist. The initial committee had 13 members, and the membership rotates over time. Committee members who have a conflict of interest, such as an employment tie to a school, do not take part in discussions or votes involving that team.
What the committee does
Beginning in early November, the committee releases a ranking each week until the final one that sets the playoff field. Members meet in person, review data and film, discuss and vote. The process uses repeated rounds. Small groups of teams are chosen, ranked, set aside, and the next group is considered. Each round narrows the field in order to produce the list.
The committee does not use a formula. Its published protocol tells members to consider factors such as:
- Strength of schedule: who a team played and how those opponents fared;
- Conference championships: a team that wins its league gets credit for it;
- Head-to-head results between teams that are being compared;
- Results against common opponents; and
- the overall quality of wins and losses.
The protocol has been refined over time, so check the current published version for the list. The committee has also said it looks at performance on the field, not at a team's reputation or the preseason ranking.
What its ranking determines
The committee's final ranking, released after conference championship weekend, is the main input in deciding who enters the playoff and how teams are placed, subject to whatever rules the playoff has for conference champions. The format and the number of teams have changed since the committee started, which is why you should look at the current rules on the playoff page. The committee's rankings during the season are provisional. They are the committee's view that week and can change, which is why the weekly releases create so much debate.
Computer rankings
Computer rankings take human opinion out of the process, at least in theory. A computer rating is a mathematical model that takes the results of games and produces a number for every team.
How they work
Most computer systems use some combination of:
- the winning or losing result of each game;
- the strength of the opponent, so beating a strong team counts for more;
- location, since a road win is harder than a home win; and
- sometimes the score, or the efficiency of a team's play.
The models differ in what they include. Some are designed to predict future results. Others are designed to reward what a team has already accomplished. The distinction matters, as we explain below. Well-known examples published over the years include the Sagarin ratings, the Massey ratings, SP+ and ESPN's Football Power Index.
Strengths and weaknesses
Computers are consistent. They apply the same math to every team, and they never favor a famous program. But they are only as good as their inputs and assumptions. Early in the season, when there is little data, they rely heavily on preseason estimates. They cannot see an injury unless it is built in, and the way they treat margin of victory is a design decision made by the person who built the model. Most important, their output reflects the choices that go into the model, which are human choices.
Rating versus ranking
Here is the most useful idea in this article. Two different questions get answered by rankings:
- Who is the best team? This is a question about strength. If two teams played tomorrow, who would win? A computer rating that predicts games answers it.
- Who has the best résumé? This is about what a team has accomplished. Who has earned a place by beating good opponents and avoiding losses?
These can produce different answers. A team with a powerful roster that lost twice may be strong but have a weaker résumé than a team with a perfect record against a modest schedule. The AP poll and the committee both mix the two questions, which is one reason they argue.
A hypothetical comparison
Imagine two teams, which we will call Team North and Team South.
| Team North | Team South | |
|---|---|---|
| Record | 12-0 | 10-2 |
| Schedule | Mostly weaker opponents | Several top-ranked opponents |
| Losses | None | Two, both close, both to highly ranked teams |
| Margin of victory | Often narrow | Often large |
Which is better? A résumé-based ranking might favor North, because they did not lose. A strength-based rating might prefer South, who beat a harder set of opponents and by larger margins. Neither answer is wrong. They answer different questions, and the argument is about which question should matter. Committee members, voters and computer designers each make a choice.
How rankings affect the season
- Playoff selection. The committee ranking chooses the field.
- Bowl matchups. Games outside the playoff often draw on rankings and conference tie-ins; see bowl games explained.
- Recruiting and television. A high ranking draws attention from recruits and broadcasters.
- Scheduling. Opponents who finish ranked help your strength of schedule. Teams think about this when they build future schedules.
- Perception. A ranking can become a self-fulfilling label, since voters often treat past rankings as evidence.
Rankings beyond the top division
FBS gets the most attention, but other levels have their own rankings. The Football Championship Subdivision, Division II, Division III and the NAIA each publish polls and, in some cases, official ranking committees that build playoff brackets. Their rules differ from those of FBS. If you follow a smaller program, check the ranking source for your division. For the divisions, see FBS, FCS, Division II and III explained.
How to read a ranking sensibly
- Check who made it. Opinion poll, committee or computer?
- Check what it measures. Strength, résumé or a blend?
- Watch the trend, not just the number. A team rising from No. 12 to No. 8 is more informative than a static No. 8.
- Ignore preseason rankings after a few weeks. They fade as results arrive.
- Look at strength of schedule. A ranking with no context is easy to over-read.
- Treat the rankings as opinion. None of them is a rule. Outcomes on the field decide who advances.
You can follow the current lists on our rankings page and check results on the scores page. If you want to follow a team that is climbing or sliding, our college section shows the latest results for each program.
The short version
- College football has no single ranking because teams do not play a common schedule.
- The AP poll (writers, begun 1936) and the Coaches Poll (head coaches, begun 1950) are opinion polls, each ranking 25 teams by points.
- The BCS combined polls with computer ratings from 1998 to 2013.
- Since 2014, a selection committee ranks teams, and its final ranking sets the playoff field.
- Computer ratings apply consistent math but depend on the choices built into each model.
- Ratings estimate strength, while rankings often reward résumé, and the two can disagree.
- Read every ranking by asking who made it, what it measures and how many games stand behind it.