How are College Football Teams Ranked?

College football teams in the US are ranked in two ways: by human voters and by computers that use mathematical formulas. Some rankings are polls that use qualified voters who cast ballots indicating their opinions of the best teams in order. Other rankings are compiled by computers based on factors such as the outcomes of games, the scores of the games, the location of the games and perhaps other criteria in the formulas. At the highest level of US college football, a collection of both types of rankings has been used to determine which two teams will play in the annual Bowl Championship Series (BCS) National Championship Game.

Voters in college football polls are chosen by the organization that is conducting the poll. In most cases, the voters are journalists, coaches, school employees or former players. Since the mid-1930s, The Associated Press has polled sportswriters and broadcasters. A poll of coaches has been conducted since 1950 and has been sponsored by various organizations, including United Press International, USA Today and ESPN. The Harris Interactive poll, which uses a combination of media personnel, coaches, school administrators and foreign players, was first conducted in 2005.

There are many mathematical formulas that computers use to rank college football teams. The six most prominent are those developed by Jeff Sagarin, Richard Billingsley, Wesley Colley, Kenneth Massey, Dr. Peter Wolfe and the team of Jeff Anderson and Chris Hester. The factors and methods used by each are different. For example, some of the formulas do not consider the scores of games — only whether a team won or lost. Not all of the formulas that are used by the computers have been publicized.

College football rankings typically are released each week during the season. Many of them also are published before the first game of the season and are known as preseason rankings. Before the late 1960s, it was uncommon for the polls to be updated after the bowl games, which are additional, end-of-season games played by successful teams, but it has become the norm since then. Some polls avoid rankings until after the first several weeks of the season, based on the belief that preconceptions and previous seasons should not affect the current season's rankings.

Rankings that are based on voters' ballots typically include the top 25 teams, with any other teams that received votes being given honorable mention in the order of their voting totals. Before the 1990s, some polls often listed only the top 10 or top 20 teams. Computer rankings typically include every team — more than 100 of them in the highest division of college football.

Starting in 1998, the National Collegiate Athletic Association (NCAA) began using the BCS to determine which football teams would play in the championship game for its highest level of competition, which was named the Football Bowl Subdivision (FBS). These rankings also are used to help determine which teams will play in a few other of the most prestigious bowl games. The BCS system combines the most prominent human polls and computer rankings, along with other factors, into another set of rankings. This system is regularly evaluated by the NCAA, and changes might be made between seasons if the organization's officials determine the adjustments to be necessary.

Although most people are more familiar with the rankings of FBS teams, rankings also exist for football teams at lower levels of the NCAA and for teams that belong to the National Association of Intercollegiate Athletics (NAIA), which is for small colleges. Just like the voter rankings for FBS teams, those for lower-level teams typically are based on ballots submitted by people such as media members, coaches or other school employees. One difference is that the polls are often done on a regional basis because it is more difficult for voters to have a great deal of knowledge about small-college teams that are in other areas. In some cases, these rankings might be compiled on a regional as well as on a national basis.