The score projection formula I use to come up with bowl outcomes had a mixed year. It did its best ever straight up, going 29-11 in just picking the eventual winners of the games. It also had its worst ever year in six total against the spread at just 17-23 versus the sharps in Vegas.
There is still one more bowl game to go: the national championship between Clemson and Alabama. The formula went 15-3 straight up in games between Power 5 conference teams, and one of the misses was picking Texas A&M over Louisville. There were some extenuating circumstances with the QB exodus from College Station, so the formula worked out splendidly over those kinds of games.
Of course, one of the other misses was projecting Oklahoma to beat Clemson 37-30. The reality was a 27-point swing the other way with the Tigers winning that one 37-17. It didn't come close with that game.
Then again, it didn't come all that close with the other semifinal either. It projected Alabama to win 26-19, a solid touchdown victory albeit with a Spartan cover of the 9.5-point spread. Instead, the Crimson Tide rolled all over MSU to the tune of 38-0. Turns out, it's not a good idea to fight Alabama by doing roughly the same thing Alabama does but with less talent.
So with the possibility that this is underestimating both teams, the formula sees a classic of a title game. It has Alabama winning 27-26, meaning Clemson covers the (presently) seven-point spread. I can project yardage the same as I do points, and it comes out as having the Tigers outgaining the Tide 391-348 despite the loss. That would be an impressive showing by Clemson, as only Ole Miss's 433 and Mississippi State's 393 beat that among past Alabama opponents on the season. The Rebels' 43 also is the only score greater than the 26 projected for Clemson.
Of course given how the formula has done straight up and against the spread this year, maybe the smart pick is to shout "Roll Tide" and take Alabama -7.