The AP Poll Dropped Ohio State Five Spots. The Betting Market Moved Them About One Percentage Point.

Ohio State led 23-3 with fifteen minutes to play in Austin, gained 49 yards in the fourth quarter, and lost 24-23. Over the previous twenty-two seasons, AP No. 1 teams holding a fourth-quarter lead of twenty or more points were 197-0. They are now 197-1.

On Sunday afternoon the Associated Press panel responded by dropping Ohio State from first to sixth, behind Texas, Georgia, Notre Dame, Indiana and Miami. Four of those five teams spent Saturday beating opponents they were supposed to beat. Ohio State's poll total fell to 1,407 points against Texas's 1,678, and the Buckeyes are now the highest-ranked one-loss team in the country, which is another way of saying the panel moved every unbeaten team above them regardless of what any of those teams had done.

FanDuel's national-championship board says something different. Texas, which won, is +600. Ohio State, which lost, is +650, level with Notre Dame and Georgia. Convert those: 14.3% against 13.3%. The market separated the winner and the loser of a one-point game by roughly one percentage point of championship probability. The poll put five teams in between.

What actually happened, because it matters to the argument

Texas did not play well. Arch Manning finished 23-of-37 for 195 yards with a touchdown and an interception — 5.3 yards an attempt. Hollywood Smothers, the leading rusher, carried 21 times for 74. Ryan Wingo caught four passes for 22 yards, one of them the touchdown that started the rally. Julian Sayin threw for 278 and Jeremiah Smith caught eight for 164, and Ohio State was the better team for three quarters by a distance that was not close to debatable.

Then the fourth quarter happened, and the mechanism was not glamour. Texas put together three consecutive touchdown drives of twelve plays or more. Ohio State gained 49 yards in the entire period. Connor Hawkins missed a 45-yard field goal that would have made it a two-score game — ESPN puts the miss at just over four minutes left, Pro Football Network at 2:55, and we could not reconcile the two, so treat the clock as approximate and the miss as certain. Smothers scored from a yard out with 25 seconds remaining.

Ryan Day's own accounting was the accurate one: "When you look at not being able to get off the field on defense, when you're really not able to sustain any drives on offense, and then you miss the field goal — all three phases didn't get it done in the fourth quarter."

One note on sourcing, because we would want it flagged if somebody else did it. ESPN's game story says Manning "completed 11 of his final 15 passes for 196 yards." ESPN's own box score has him at 195 yards for the game. Those cannot both be true. We used the box score.

The poll is not measuring the thing people think it measures

Here is the structural fact that the Sunday rankings discourse routinely forgets: the Associated Press poll plays no role in selecting the College Football Playoff. None. The selection committee produces its own rankings, it does not begin releasing them until November, and it is not bound by anything a panel of writers thought in September. The AP poll is a weekly opinion survey with a long history and real cultural weight, and its mechanical effect on Ohio State's path to a national championship is zero.

The market, by contrast, is pricing exactly one thing: the probability that a given team is holding the trophy in January. It has to price the whole distribution — schedule, roster, seeding, the twelve-team bracket, the fact that a one-point road loss to the team that is now ranked first is close to the most survivable defeat available in college football.

And the precedent is only two seasons old. Ohio State lost 13-10 at home to Michigan on November 30, 2024, entered the playoff as the No. 8 seed, and beat Tennessee, Oregon, Texas and Notre Dame to win the national title. The poll had opinions about that team in December too.

The best argument that the market is the one that is wrong

We should take this seriously, because college football futures are among the softest markets in American sports and it is not close.

The objection runs like this. Futures boards are not probability estimates; they are inventory management. A book sets a price to balance its liability, and blue-blood futures attract enormous, price-insensitive public money from fan bases who bet their team in August and never close the position. Ohio State is one of the three or four largest of those liabilities in the sport. A book holding a mountain of Ohio State tickets written at +450 in the summer has every commercial reason not to shorten that number much after a one-point loss, because the bettors it wants are the ones buying Texas at a worse price today. On that reading, +650 does not mean the market thinks Ohio State is the second-most-likely champion. It means the market is managing a book, and the AP panel — whatever its flaws — is at least trying to answer the question directly.

That objection is correct about the mechanism and we are not going to pretend otherwise. Futures prices carry a large vig, a documented favourite-longshot distortion, and a public-money thumb on exactly the teams involved here. The board is not a clean probability.

But it does not rescue the poll, because the poll has the same problem without the discipline. A panel of voters who watch one game and reorder a hundred and thirty teams is also running on narrative and also has no money at risk. The difference is that somebody who thinks +650 is wrong can bet against it and be paid for being right. Nobody can be paid for being right about the AP poll. That asymmetry does not make markets oracles; it makes them the only one of the two instruments with a feedback loop.

There is also a test available that does not depend on trusting either. Ohio State plays a full Big Ten schedule and still controls its own conference path. If the panel's read is correct — that a fourth-quarter collapse revealed something structural — the Buckeyes should lose more games. If the market's read is correct, they should not, and by November the committee's first ranking should have them within a spot or two of where the futures board has them now. We will know inside eight weeks.

What we are not claiming

That Texas is the best team in the country. The Longhorns were outgained, out-quarterbacked and outplayed for three quarters and won because the opponent stopped functioning. The AP panel gave them 56 first-place votes and the first regular-season No. 1 ranking Texas has held since October 13, 2024, which is a reward for the last fifteen minutes and not the first forty-five.

That the market's view of Ohio State will survive contact. It might not.

That the collapse was meaningless. A team that gains 49 yards in a quarter with a national championship in its hands has something wrong with it, and Day's postgame answer conceded as much in three clauses.

What we are claiming is narrower and harder to argue with: the poll moved a team five places for losing by one point to the team now ranked above it, the market moved the same team by about a percentage point, and only one of those two numbers has to be defended by anyone.