For the fifth time in five years of the Playoff, the No. 1 seed didn’t win the whole thing. Two of those times, the No. 1 got knocked out even before the title game.
- 2014 Alabama lost to No. 4 Ohio State in a semifinal (as No. 6 TCU destroyed an Ole Miss that’d also beaten Bama)
- 2015 Clemson lost to Alabama in the Championship
- 2016 Alabama lost to Clemson in the Championship
- 2017 Clemson got blown out by No. 4 Alabama in a semifinal
Along with telling us Alabama and Clemson are by far the two best programs, maybe No. 1 seeds’ losses show four isn’t an automatic magic number.
There are other ways to argue that point. The current Playoff excludes mid-majors. Giving more teams a Playoff shot would make more games feel important. There’s money to be made. But No. 1 seeds losing in the Playoff every year makes a case about competition.
The numbers fluctuate a bit each year, but according to S&P+, there are about seven Playoff-caliber teams and three title-caliber teams per season. In 2018, you could argue there were only two or three Playoff-caliber teams.
But you could have also argued, based on S&P+ and just watching the teams play, that it would’ve been nice to have Georgia and Ohio State in the Playoff this year (at least before watching Georgia’s Sugar Bowl). Maybe you wanted to see undefeated UCF in the field too.
Top seeds’ repeated Playoff defeats shouldn’t even be that surprising, given how little information we have about every team in the country
College football has the shortest season, 12 games, of any of the major American sports. Teams only play between 11 and 13 of the 130 FBS teams before the Playoff selection committee picks the field. The sport where teams only play 10 percent of the other teams at their level is the one with the smallest playoff of any major American league.
The majority of that 10 percent is conference opponents, which makes it hard to be certain about schedule strength, though analytics have made a lot of progress in assessing that. Teams that play each other in the Playoff might have no common opponents, or one or two.
Even that’s not much of a reliable indicator — Maryland beat Texas, for instance — but in a low-information sport, the Playoff committee uses it to help pick teams. The committee also cares to some degree about who wins conference championships, despite teams taking vastly different paths to win their leagues.
If there were ever an example of how little we really know about these teams of 18- to 22-year-olds after 13 games against different competition, it was Bama losing by 28 in the biggest game of the year. Hours before that, Bama was poised to join the Best Team Ever competition. Maybe having more games to sort it out on the field would be good, perhaps if there’s a way to trim games we don’t need and manage the student-athlete load.
It’s not even like these losses by top seeds have been that shocking.
2014 Bama was a 7.5-point favorite against Ohio State in Vegas. But S&P+ had been higher than Vegas on the Buckeyes pretty much all year, so it’s not like that was a stunner.
2015 Clemson was a 6-point Vegas underdog to Bama.
2016 Bama was only a 6.5-point favorite against Clemson.
2017 Clemson was a 3.5-point underdog to Bama, which S&P+ had as better anyway.
2018 Bama was a 5-point favorite in the Championship, despite S&P+ saying the teams were separated by just 1 point.
This is a more convincing argument than pointing to bowl losses by teams that just missed out on the Playoff.
One argument that’s been made against expansion lately is that the committee’s No. 5, 7, 8, and 9 teams (Georgia, Michigan, UCF, Washington) took bowl losses. Those losses are an example of the same thing No. 1 seeds’ repeated Playoff losses show: that college football teams are hard to figure out, even when we think they’re good.
Expansion’s not happening now. Before the Championship, Playoff Board of Managers chairman Mark Keenum said it was “way too soon” to expand out from the current four-team model.
No. 1 seeds continually losing just further strengthens the argument to expand the current model, whether that happens soon or not.