Should Computer Models or Human Voters Decide College Football's Playoff Rankings?
This post is about a recurring tension in how college football ranks its teams: human poll voters versus computer-driven formulas, reignited by this week's Top 25 releases.
Heading into Week 4 of the 2026 season, ESPN's model dropped Texas a spot even though the Longhorns had just blown out UTSA, days after the same model had ranked Texas No. 1 the week before its win over Ohio State. Meanwhile, the AP's human poll voters made their own contested calls, dropping Texas A&M several spots after an unranked Kentucky team beat the Aggies at home, and shuffling LSU and Ole Miss after their head-to-head matchup. Oregon's 84-point rout of Portland State also became a flashpoint, with some analysts arguing computer models reward lopsided scores against overmatched opponents in ways that don't reflect playoff readiness.
The debate matters beyond bragging rights. The 12-team College Football Playoff selection committee uses a mix of the eye test, strength of schedule metrics and analytical models like ESPN's Football Power Index to seed the bracket, and disagreements between those inputs have already sparked criticism this season, including public comments from coaches like Nick Saban about how wins and losses get judged.
The case for:
Algorithmic models strip out regional and brand-name bias, weighting strength of schedule and scoring margin consistently instead of rewarding preseason reputation, which is exactly what critics say happened with Texas this week.
Human polls have a track record of clinging to traditionally strong programs even after their on-field résumé is passed by other teams, which computer rankings are built to correct.
The case against:
Formulas can reward blowout scores against weak opponents, like Oregon's 84-point win over Portland State, in ways that don't necessarily reflect a team's readiness for playoff-level competition.
Longtime poll voters can account for context a formula can't, such as a key injury or a backup quarterback, and computer models have no mechanism to weigh that kind of nuance.
What's your take?
Should the College Football Playoff rely more on computer models like ESPN's FPI than human poll voters to rank teams? Yes 
No 
Other 
#CollegeFootball #CFP #AP25 #SportsRankings
Now let's hear from you.
You get one Take and 3 ratings so use them well.
Strong arguments beat loud ones. See Moving The Needle (How to)
Don't forget to rate your own Take.
Please don't feed the trolls.
