We called Texas by two and Texas won by one, and nothing in between was calm. Ohio State led 23–3 with 15 minutes to play after Texas turned it over on its first two snaps of possession and kicked a field goal from the 14. Then three straight touchdown drives of 91, 72 and 73 yards, two fourth downs converted, a 45-yard Buckeye miss with 4:16 left, and Hollywood Smothers from the 1 with 25 seconds to go. The market had Texas by 1.5; our model had Ohio State, the season's first split, and the market won it. Every pregame claim is graded in the recap, misses first.
Before the game we wrote down Texas 26, Ohio State 24, and the game gave us Texas 24, Ohio State 23, so we had the right winner and missed the margin by one point, which sounds calm and was anything but. Texas gave the ball away on its first two possessions, settled for a field goal from the 14-yard line, and with 15 minutes left was losing 23 to 3, which is usually the kind of score that means the game is over. Then Texas drove 91 yards for a touchdown, then 72, then 73, converted two fourth downs along the way, and watched Ohio State miss a 45-yard field goal with 4:16 to go. Hollywood Smothers went in from the 1-yard line with 25 seconds left, and that was the winner. One more thing. The betting line had Texas by a point and a half and our computer had Ohio State, the first time all season the two disagreed, and the computer was wrong and the betting line was right.
The words we use: the call was the score we wrote down before the game and left alone · the market is the betting line, the number the people wagering real money expected the favorite to win by · the model was our computer's own guess, printed so you could compare · tells were three early signs; if they showed up, the game was going the way we said · edges were which team had the better side of each matchup · graded means we went back over each thing we said and marked it right, partly right, or wrong.
The chips below are our pregame calls, left as printed. Seven of the thirteen board pairings graded RIGHT, three PARTLY, three WRONG, and all three WRONG chips were about Texas throwing the ball: Even on Coleman and Wingo, Even on the Texas deep game, and a slight Texas edge on covering Sayin's deep shots. The rebuilt Buckeye secondary held Manning to 5.3 an attempt and nothing longer than 21; Sayin hit Smith for 43 twice. The full regrade is in the recap.
Before the game we picked a winner for each of 13 matchups, and the list below is exactly what we printed. We got 7 of them right, 3 half right, and 3 WRONG, and all three misses were about Texas throwing the ball: we called the Texas receivers against Ohio State's defensive backs a toss-up, called the Texas deep passing game a toss-up, and gave Texas a slight edge covering Ohio State's long throws. Ohio State's rebuilt group of defensive backs held Arch Manning to 5.3 yards a throw and nothing longer than 21, while Julian Sayin hit Jeremiah Smith for 43 yards twice. The full grading is in the recap.
The upset math held: we said there was no upset here, that 1-to-3-point favorites went 111-115 since 2023. A coin flip landed on the home side by one point.
Why nobody should be shocked: we said before the game that there was no upset to be had here, because since 2023 teams favored by 1 to 3 points have won 111 games and lost 115, which is a coin flip. This coin landed on the home team's side by a single point.
Margin: 1 · Line was Texas −1.5 · Model said Ohio State by 2.4
Cooper out as expected; Goosby started; Russaw played for Ohio State. No late scratches to grade.