Maybe it’s time to get rid of the AP Poll once and for all

The AP Poll is broken and has been for years.

When an AP voter ranked Florida higher after losing at home to South Florida while leaving USF completely unranked, then dismissed the controversy by saying “it’s really fun for discussion,” it revealed what we all already knew about a fundamentally flawed system.

“This is indicative of a broken, stupid, antiquated voting system,” said Barstool Sports’ Brandon Walker.

The AP Poll worked perfectly in 1936 because the world was completely different then. Television didn’t exist, radio broadcasts were rare, and beat writers were basically the only way to get comprehensive college football coverage. Those reporters went to every game, knew all the coaches and players personally, and had information nobody else could get.

Beat writers today have completely different jobs that make national coverage impossible.

A USC beat reporter isn’t watching Penn State or tracking Arkansas because they’re busy covering one program full-time. Their Saturday starts hours before kickoff at the Coliseum, checking injury reports, watching pregame warmups for tells about who’s playing, taking notes on every meaningful snap during three hours of football, interviewing players and coaches afterward, filing multiple stories by deadline, then doing it all over again the next week.

That’s what it takes to cover one team properly, but somehow we expect these same people to have informed opinions about 25 teams across multiple conferences while doing their actual jobs.

This is symptomatic of a bigger problem with how sports media works today, and the AP Poll represents everything that’s wrong with pretending that access and titles matter more than actual knowledge. Sixty-two media members vote in the AP Poll, and most cover one team, maybe one conference, but they’re asked to rank 25 teams across the entire country every week, based on what exactly? Highlights on SportsCenter? Box scores? Gut feelings about brand names?

It’s an impossible job that sets everyone up to fail because nobody can watch enough football to make informed decisions about teams from every conference. ESPN shows hundreds of games now, and a dedicated fan with ESPN+ and a YouTube TV subscription can watch more football than any working journalist possibly could. The Florida-USF thing proves the point because any fan who watched that game knew Florida got beaten by a better team. USF outplayed them for four quarters at home, but expecting one voter to have watched every meaningful game that weekend is unrealistic.

The system sets people up to fail, so voters end up ranking teams they haven’t seen because Alabama or Ohio State or whoever “should” be good. Then, when those teams lose, they barely drop because the whole poll has momentum based on assumptions rather than actual performance.

That’s why computer rankings might just work better because they process every play through data, and ESPN’s FPI uses metrics no human could calculate, while the College Football Playoff committee — for all its faults — uses 13 people who can dedicate their time to watching film and evaluating teams systematically.

Professional sports use win-loss records with clear tiebreakers, and the Big 12 already killed its preseason poll this year because Commissioner Brett Yormark said it hurt teams for no reason. Big Ten coaches have also called for eliminating preseason polls, with ESPN’s Adam Rittenberg noting that several coaches and officials agreed “the preseason polls should be obliterated” because “they carry too much meaning and shape debate, overvaluing/undervaluing teams before we know anything.”

Kansas State coach Chris Klieman has been pushing to eliminate preseason polls completely. And they’re not wrong because preseason polls create narratives that stick all season, even when they’re obviously flawed. Teams get ranked high in August and stay ranked until they lose multiple games, while teams that start unranked can’t climb fast enough, even when they’re clearly better.

Perhaps we should stop asking beat writers to vote on teams they don’t cover. Instead, either use computers that actually process every game, or use a committee of people who know football and watch film for a living. The CFP committee isn’t perfect, but at least they’re trying to evaluate teams based on actual performance instead of guessing, and they use strength of schedule, head-to-head results, and context. Revolutionary stuff. Or just admit polls don’t matter until Week 6 when you have enough data to make real comparisons. Danny Kanell has been saying for years to let teams play some games before deciding who’s good.

But here’s the thing: the AP Poll will keep existing because it always has, and college football loves tradition more than logic.

The poll gets clicks and generates debate, which is apparently more important than accuracy, because media members like having a vote because it makes them feel important. Coaches complain about polls but use rankings to recruit and build narratives, while fans argue about rankings because it’s fun, so everyone benefits from the broken system except the teams getting screwed by voters who don’t watch their games.

The Florida voter got harassed into making her social media private, but she’ll keep voting, and other voters will keep ranking teams they’ve never seen while we keep pretending this makes sense. The AP Poll made perfect sense when beat writers were the only source of information about college football, but now everyone has more information than those original voters ever had. We’re still using their system because nobody wants to admit it’s time for something better, and college football deserves ranking systems that match 2025, not 1936.