Ohio State scored 42 points on Tennessee in 32 degrees on a Saturday night at Ohio Stadium on December 21, and the first 12-team College Football Playoff first round was, by the end of the weekend, exactly the lopsided event that the analytical community had been predicting since the bracket got released. Texas beat Clemson 38-24. Ohio State beat Tennessee 42-17. Penn State beat SMU 38-10. Notre Dame beat Indiana 27-17. Four games. Four favorites. Combined margin of victory: 65 points. The format that had been sold as a way to extend competitive college football into late December produced its first weekend of football that the higher-ranked team won every single time, by an average of 16 points.
The framing of “first 12-team weekend exposed the format” started on Sunday morning and dominated the next three days of college football media. The framing was understandable but mostly wrong. The lopsided results were not an indictment of the 12-team format. They were a reasonable outcome of seeding 12 teams in a sport where the talent gap between the 5-12 band and the 8-12 band is structurally larger than the gap inside the top four. The format produced exactly the games it was designed to produce. The question worth asking is what those games actually told us about the eight teams still alive, and what the round itself reveals about the format’s structural problem.
What follows is what the first round actually showed about each surviving team, where the public efficiency models actually disagreed with the on-field results, and what the quarterfinal matchups suggest about the rest of the bracket.
What the four first-round results told us
Texas-Clemson was the closest game on the scoreboard and the cleanest mismatch underneath. Texas’s offense produced 0.41 EPA per play; Clemson’s defense allowed a season-high yardage total. The 38-24 final included two late Clemson touchdowns in garbage time that compressed the margin. The model-implied win probability had Texas above 90% by the third quarter. The Longhorns are a real top-five team and the score confirms it without flattering them.
Ohio State-Tennessee was the largest margin of the round and the clearest expression of how the Buckeyes’ front-loaded offensive identity travels. The Buckeyes scored on six of their first seven drives and produced 0.51 EPA per play across the first half. Tennessee’s defense had been top-fifteen during the regular season; against Ohio State’s specific offensive structure, the unit gave up explosive plays at a rate that no Volunteers defense had produced all year. The 42-17 result is the cleanest example yet of the offensive ceiling that Ryan Day’s roster has been building toward.
Penn State-SMU was the most-decisive defensive performance of the round. The Nittany Lions held SMU’s top-fifteen offense to 10 points, six first downs, and 198 total yards. The Penn State defense, which had been bottom-half in pass-pressure during the regular season, generated pressure on 41% of dropbacks against SMU. The structural reason was a defensive coordinator adjustment that pressed SMU’s receiver routes more aggressively than the regular-season tape suggested.
Notre Dame-Indiana was the only game where the underlying numbers had been ahead of the committee. The committee ranked Notre Dame seventh; SP+ ranked them fourth. The 27-17 result aligned with the model rather than the committee. Indiana’s defense had been a top-ten unit but Notre Dame’s offense exposed structural weaknesses against the run that the regular-season schedule had hidden.
The model-vs-result gaps, summarized
| Game | Final score | Model-implied margin | Gap (model vs result) |
|---|---|---|---|
| Texas 38-24 Clemson | +14 | +11 | +3 Texas (close to model) |
| Ohio State 42-17 Tennessee | +25 | +15 | +10 OSU (above model) |
| Penn State 38-10 SMU | +28 | +10 | +18 PSU (well above model) |
| Notre Dame 27-17 Indiana | +10 | +7 | +3 ND (close to model) |
The pattern is interesting. Two games landed close to the model (Texas, Notre Dame). One landed well above (Penn State). One landed somewhat above (Ohio State). The Penn State result specifically is the one worth pausing on — the +18-point gap above the model implied margin is the kind of single-game outlier that usually corrects in the quarterfinals. Penn State plays Boise State next. The model has them favored by 9. If they win by 25, the SMU performance was structural. If they win by 7, the SMU game was variance.
Where this gets weird
The clean “first round was lopsided, format is fine” reading misses three nuances the broader CFP coverage has not surfaced.
The first is that the first-round format itself produces structural mismatches. Seeds 5-8 host seeds 9-12, with the home team having earned its higher ranking through a stronger regular-season resume. The home-field advantage compounds the seeding advantage. Across the entire history of college football regular-season home games between teams in the top-15, the home team has won approximately 68% of the time. The CFP first round narrowed that further because the talent gap between the 5-8 band and the 9-12 band is larger than the average inter-conference matchup. The 100% home-team win rate is high but not shocking.
The second is that the first-round games may have been competitively boring but they were commercially excellent. The four games drew an average of 11.8 million viewers — well above the regular-season college football average and competitive with NFL ratings for the same weekend. The argument that the format needs to produce closer games to be commercially viable was already partially refuted by the audience response. Viewers showed up for college football in late December even when the games were not close.
The third is that the four winners now face four teams that did not play in the first round. Oregon, Georgia, Boise State, and Arizona State got byes. The bye-week teams have historical advantages in tournament basketball but the football precedent is thinner. Whether Oregon’s 14-0 regular season translates to a quarterfinal advantage against Ohio State is the open empirical question. The format’s first real competitive test is the quarterfinal round, not the first round.
What the quarterfinals will actually test
- Oregon vs Ohio State at the Rose Bowl. The model has it as a one-score game. Oregon’s defense was the best in college football all season. Ohio State’s offense was the most efficient. Strength vs strength matchup.
- Texas vs Arizona State at the Peach Bowl. The model has Texas as a 10-point favorite. Arizona State’s run-heavy offense is structurally hard to play against in a single game. The Sun Devils have a higher ceiling than most pieces are giving them credit for.
- Notre Dame vs Georgia at the Sugar Bowl. The model has Georgia by 6 despite the SEC champion’s late-season offensive struggles. Notre Dame’s defense will be tested by a Georgia offense that has more explosive-play potential than the regular season suggested.
- Penn State vs Boise State at the Fiesta Bowl. The model has Penn State by 9. If Penn State’s first-round defensive performance was structural, they cover comfortably. If it was variance, this game gets closer than expected.
The callback
That cold Saturday night at Ohio Stadium when the Buckeyes scored 42 on Tennessee in single-digit wind chill was the loudest expression of what the 12-team format actually produces in the first round. The home team wins. The favorite covers. The audience shows up regardless of whether the game is close. None of this is a format failure. It is the format working as designed against opponents who had earned the lower seed through measurably weaker regular seasons. The quarterfinals are the real test of whether the bracket produces competitive football across the next two rounds. The four games next weekend will tell us more about the 12-team format than the first round did. The home team will not always win. The favorite will not always cover. By January 1, the brackets that survive will be the ones the format was designed to produce. The SEC vs Big Ten piece covers the broader version of how conference structure interacts with playoff seeding. The first round happened. The bracket continues.
Game data via NCAA.com; SP+ ratings via Bill Connelly at ESPN; viewership data via CFP official.
