Varsity Numbers: F/+ vs. AP

by Bill Connelly

As is the case each season, we have made one final update to our college football F/+ projections as the season begins. (The season has technically already begun, but as proof that we're not doing anything after the fact, Brian Fremeau posted his final numbers on Aug. 30 here, and I posted my final numbers on Aug. 31 here.)

These numbers mostly mirror what was presented in Football Outsiders Almanac 2015 (still available!), but there are a few differences based on both changes to the experience numbers and my endless tweaking of weights in formulas.

Depending on how you look at it, either the AP did a pretty good job with its preseason polling, or the numbers pass the eyeball test. Or both. As you see below, of the teams in the F/+ top 15, 13 are also in the AP top 15. That's a little more harmony than normal, I believe.

There are some differences, though, and I thought this would be a good time to discuss them.

AP likes 'em more

TCU (AP No. 2, F/+ No. 10)

(Note that while we still have TCU below where conventional wisdom puts them, they have moved up, after being ranked No. 15 in Football Outsiders Almanac 2015.)

Why the media likes the Horned Frogs: They looked amazing in last year's Chick-Fil-A Bowl (thereby winning the retroactive "team most spurned by the Playoff" award), they return a Heisman favorite in quarterback Trevone Boykin, they finished 12-1 last year, and when they looked good, they looked amazing. Seriously, what their defense did to Ole Miss in the bowl was not safe for work.

Why the numbers don't (as much): They were absurdly lucky from a turnovers perspective last year; going by the way I measure turnover luck, bounces benefited them to the tune of more than six points per game, the most in the country. And while they certainly had some incredible wins -- 42-3 over Ole Miss, 30-7 over Minnesota, and of course 82-27 over Texas Tech -- those bounces came in handy. They were 3-1 in games decided by one possession.

Also: their offense wasn't amazingly efficient -- when Boykin struggles, it's because his completion rate is dipping quickly -- and they have to replace quite a few big-time contributors on defense. We assume a Gary Patterson team will always be fine defensively, but the numbers aren't necessarily going to see that.

Auburn (AP No. 6, F/+ No. 13)

Why the media likes the Tigers:

Gus Malzahn offense + Will Muschamp defense = success!

Muschamp returns to his defensive roots after his ouster as Florida's head coach, and he inherits a unit that has plenty of athleticism and was rock-solid over the first half of last season. Throw in Malzahn's offensive wizardry, and you've got yourself a hell of a team.

Why the numbers don't (as much): For one thing, the stats don't care who the coordinator is. I'd love to change that, but it's nearly impossible. The numbers also take note of Auburn replacing its quarterback, two leading running backs, two of three receivers and three starters on the offensive line. Malzahn has recruited quite well, and again, it's really hard to assume poor play from an offense powered by Gusoline. But there's quite a bit of turnover here, and the new pieces aren't guaranteed to immediately click.

Clemson (AP No. 12, F/+ No. 19)

Why the media likes the Tigers: Quarterback Deshaun Watson really might be amazing (and at this specific moment in time, he's healthy, too!), and he's got most of his skill position guys back, as well.

Why the numbers don't (as much): The offense stunk last year (73rd in Off. S&P+), mainly because Watson was suffering from countless unlucky ailments. Also: the Tigers were horrendous on the ground (107th in Rushing S&P+). Oh yeah, and the defense, an incredibly disruptive force that helped the Tigers to overcome offensive woes and still finish 10-3, was decimated by graduation and attrition: six of the top seven linemen are gone, as are the top two linebackers, the top safety, and the top cornerback.

That's ... a lot to lose. Watson's potential is off the charts, and maybe his presence is enough to carry the Tigers through a lot of close games. But that's a lot to put on one player, and if he gets hurt again, Clemson's ceiling drops from top 15 to barely top 40.

F/+ likes 'em more

Ole Miss (F/+ No. 5, AP No. 17)

Why the numbers like the Rebels: Simply put, they were most likely the best team in the country for the first two months of last season. The Land Shark defense was incredible in every way, from big hits to big-play prevention, and a timely, efficient passing game was able to make plays at the right moments.

Through October 18, Ole Miss had beaten Alabama (No. 2 in F/+ at the end of the year) and destroyed Boise State (No. 21) and Tennessee (No. 24). And a lot of the key pieces to both the defense (tackle Denzel Nkemdiche, tackle Isaac Gross, linebacker/end C.J. Johnson, rover Mike Hilton, nickelback Tony Conner) are back. Plus, star receiver Laquon Treadwell is healthy and leads a potentially fantastic receiving corps.

Why the media doesn't (as much): Who cares about the first two months? The Rebels lost a heartbreaker to Auburn when Treadwell suffered a gruesome injury and fumbled as he was about to score the go-ahead touchdown, then fell into a funk: they got pummeled by both Arkansas and TCU and finished the season losing four of six.

Seriously, this was a calamitous fall. From my Ole Miss SB Nation preview:

Oh yeah, and they have to replace their quarterback now.

If TCU got a bowl bump for crushing Ole Miss, Ole Miss gets an anti-bump. But the numbers don't really care about how the season ended; they just care about the overall averages.

UCLA (F/+ No. 7, AP No. 13)

Why the numbers like the Bruins: Steady year-to-year improvement? Check. The Bruins were 75th in F/+ in Rick Neuheisel's final season (2012) and have improved to 31st, then 13th, then 12th under Jim Mora. Excellent recruiting? Check. They're fourth in two-year recruiting rank. Returning experience? Check. They return eight starters on each side of the ball, from 1,500-yard rusher Paul Perkins, to 131 career starts on the offensive line, to all but one player in each level of the defense.

That's a team built for the numbers to love.

Why the media doesn't (as much): One of the only players UCLA has to replace is quarterback and NFL early entry Brett Hundley. And the new starter, Josh Rosen, is a true freshman. He's a very well-decorated true freshman, but ... RED FLAG RED FLAG RED FLAG RED FLAG RED FLAG.

Georgia (F/+ No. 4, AP No. 9)

Why the numbers like the Dawgs: Georgia is consistently good; the Dawgs have finished 14th or better in F/+ for four consecutive years and were fourth last season. They return maybe the best running back in college football (Nick Chubb), a well-seasoned offensive line, and an active, super-athletic back eight on defense. Mark Richt recruits well and wins 10 games each year when the injury bug isn't particularly cruel. From the perspective of recent performance and returning talent, they are easily the surest bet in the SEC East.

Why the media doesn't (as much): They dropped the ball last year, laying one of the season's biggest eggs in a blowout loss to Florida and blowing the SEC East title (to Missouri, which the Dawgs beat by 34 points on the road) in the process. They struggled defending powerful run games last year and must now replace three linemen. And yes, they have a new quarterback.

One more thing: when Richt lost offensive coordinator Mike Bobo to the head coaching job at Colorado State, he replaced him with Brian Schottenheimer. When fans (in this case, those of the St. Louis Rams) lose their offensive coordinator and celebrate wildly, that's a scary sign.

Of course, all Schottenheimer has to tell his new quarterback is, "Put the ball in Chubb's (or his awesome backup's) belly."

For five of the six teams here, part of the difference between perception and projection might come because of the quarterback position. It's interesting, and it makes sense, as simply using returning starters doesn't give extra weight to the quarterback position.

So here are the official, last-second projections. F/+ rankings will be posted here weekly, starting next week. And you can find this week's F/+ projections/picks post here. To say the least, the season started well for the numbers yesterday. (Had to say that now before the tide inevitably turns.)