Problems With The AP Top 25 Poll Just Keep Getting Worse

The Associated Press Top 25 poll continues to be a mess. And given its importance in the college football world, it needs a dramatic overhaul, and quickly.

We're only a few weeks into the 2025 regular season, and already there are plenty of examples of poll voters heavily impacting the current rankings, future rankings, and how strength of schedule and resume arguments play out over the rest of the season. And it's only getting worse as the season rolls on.

It started with the preseason poll, which, as always, was a mess, with several teams highly overrated, and some very much underrated. Mostly based on brand names, an over-reliance on results from 2024, and lack of awareness. It's accelerated through the first part of the year. Just last week, one voter in the AP Poll submitted an incoherent set of rankings, then shrugged it off as fun to see people debate over it.

RELATED: AP Poll Is A Joke After Florida, USF Debacle, And It Needs To Be Fixed

But week three provided even more examples of how poor rankings can heavily impact the College Football Playoff, even if the poll isn't directly related.

LSU Tigers head coach Brian Kelly on the sidelines during a recent game. Ken Ruinard / USA Today Network South Carolina / USA TODAY NETWORK via Imagn Images

Preseason Rankings, Lack Of Updates Hurting AP Poll

One of the best examples of what a poor job the AP Poll has done thus far is the LSU Tigers. 

LSU entered the preseason as the No. 9 team in the country, which was a little optimistic, relative to more analytically minded models, but good enough. But then they moved up to No. 3, thanks to a week one win over then-No. 4 Clemson. 

Here's the problem – it's obvious that Clemson was extremely overrated at No. 4. The Tigers are 1-2, with an unimpressive win over the Troy Trojans, the LSU loss, and a week three defeat to Georgia Tech on their resume. What that should tell voters is that LSU's win over Clemson, instead of being treated like one of the best wins of the year, should instead be viewed as a win over say, the 30th best team in the country. 

But because LSU moved up to No. 3, voters won't move them down as their season becomes less impressive. And it has been much less impressive. The Tigers struggled to dominate an overmatched Louisiana Tech team, scoring just 23 points. Then their offense was once again middling against the Florida Gators. They picked up just 10 first downs and scored just one offensive touchdown. 

In a reasonable poll, LSU would be dropped behind more impressive teams like, say, the Oregon Ducks. Instead, they've remained at three and even received first place votes. Because of a win over Clemson that matters much less than it appeared to.

Another way of demonstrating the issues with the AP Poll is by comparing it to a much more accurate method of assessing team strengths: SP+. 

SP+, housed at ESPN, is a predictive tool, similar to how Las Vegas and betting sites rank teams, that takes into account efficiency on offense, defense, and special teams. It also adjusts for opponent. So, for example, it's how good were you at stopping successful plays on defense, and how good were you at creating successful plays on offense. It takes out the "brand name" and recency bias and lack of adjustment that humans struggle with. 

And there are huge disparities already between the two. Here's a list of the biggest variances:

USC: #9 (SP+) vs #25 (AP) → SP+ is 16 spots higher 

LSU: #17 (SP+) vs #3 (AP) → SP+ is 14 lower 

Missouri: #11 vs #23 → SP+ 12 higher. 

Iowa State: #23 vs #12 → SP+ 11 lower. 

Alabama: #4 vs #14 → SP+ 10 higher 

Texas A&M: #20 vs #10 → SP+ 10 lower. 

Tennessee: #6 vs #15 → SP+ 9 higher. 

Ole Miss: #5 vs #13 → SP+ 8 higher. 

Illinois: #16 vs #9 → SP+ 7 lower.

Miami: #10 vs #4 → SP+ 6 lower 

This is just one system – others will have different results. But the point spreads it generates are remarkably close to the best ranking systems on earth; those from sportsbooks. And it'll often outperform Vegas, by a marginal percentage. USC has been almost completely ignored by AP Poll voters, despite a dominant start to 2025. Missouri too, is being overlooked. Several SEC teams are actually underrated by the polls, like Tennessee, Alabama and Ole Miss. 

Does this mean that the polls must follow SP+ exactly? Of course not. But even its "resume" rankings, a list much closer to what the AP Poll is trying to accomplish, makes much more sense than what humans are doing. Strength of schedule is often based off the AP Poll. It's important to get it right, or at least somewhat close to right.