Game Analysis

Ohio State (-5.5/-6)  30   Texas  22

The math model, weighing all games equally, favors Ohio State by 5.5 points in this game, but the Buckeyes have been relatively better against better teams while Texas has been relatively worse against good teams because the Longhorns’ offense has been significantly worse against better defensive teams than they have overall.

Ohio State’s offense has averaged 7.2 yards per play with QB Will Howard in the game while facing teams that would allow 5.4 yppl to an average team and the Buckeyes have been even better in two playoff games – averaging 8.2 yppl with Howard in the game against Tennessee and Oregon defenses that would allow 4.7 yppl to an average attack. Ohio State did have that one bad offensive game against Michigan (4.3 yppl) when their play-calling indicated that they were playing not to lose the game rather than playing to win it. The coaching staff has obviously learned from their mistake in that game, and overall the Buckeyes’ offense has been relatively better against better defensive teams even with that Michigan debacle included. In fact, the regression equation projects that the Buckeyes would be 0.4 yppl better, relatively, against an elite Texas defense than they would overall. I decided to adjust 0.2 yppl for that.

The Texas defense is the best in the nation in compensated yards per play allowed, edging out Ohio State by 0.05 yppl. The Longhorns have given up 4.3 yppl to teams that would combine to average 6.1 yppl against an average defense and that unit was a bit better, relatively, against better offensive teams and I made a slight adjustment for that.

The Texas offense has been the problem, as the Longhorns racked up huge numbers against weaker defensive teams but have struggled, relatively, against better stop units. Texas has run 792 plays from scrimmage (not including “TEAM” plays) with quarterback Quinn Ewers in the game and they’ve averaged 6.1 yards per play while facing teams that would allow 5.1 yppl to an average offense. The offense was relatively better against weaker teams, as the slope of the regression equation to project the Texas offense (in terms of yppl) based on the defensive level of the opposing defenses (based on compensated yppl allowed), weighted by how many plays Ewers was in the game for against each team, is +1.59. That means that for every yard better or worse the defense was the Longhorns’ offense was 0.59 yppl better or worse, relatively. A team that plays on the same relative level against both good and bad defensive units would have a slope of 1.00. The equation would project a performance that is 0.4 yppl worse than normal for Texas facing an elite Ohio State defense, but I’ve decided on an adjustment of 0.2 yppl, which equates to 1.3 points.

Ohio State’s defense has allowed just 4.2 yppl to teams that would combine to average 5.9 yppl against an average defense and that unit was also a bit better, relatively, against better offensive teams even with the 7.6 yppl they allowed to Oregon in Euguene in week 7. The Buckeyes also yielded just 4.0 yppl to the Ducks last week while giving up only 3.7 yppl and 3.1 yppl, respectively, to very good offensive teams Tennessee and Indiana. Those 3 recent performances more than made up for the outlier in the first meeting against Oregon. As with the Texas defense, I made a slight adjustment for Ohio State being a bit better against better offensive units.

After the adjustments for how each unit performed, relatively, against better teams I get Ohio State by 8.4 points and 51.8 total points (adjusted for perfect dome conditions – they keep the roof closed at AT&T Stadium).