How FEI and S+P see the game: Oregon at Utah 2014

FEI is the Fremeau Efficiency Index, created by Brian Fremeau. Brian Fremeau is an author at Football Outsiders,ESPN and BCFToys. FEI is an advanced statistical measure for college football that tracks drive efficiency instead of per-play success.
S+P is created by Bill Connelly. Bill Connelly is an author at SBNation, RockMNation, Football Study Hall and Football Outsiders. S+P is an advanced statistical measure which combines success rate, explosiveness per play and opponent adjustments.

Before we get into the game view, I think it's important to do this first.

This article will go into more depth about the depths of how wrong I was, but boy was I wrong. That being said, S+P was wrong, FEI was wrong, everyone was wrong. No one saw Oregon's domination coming - not even AlohaMeansTouchdown.

How S+P sees the game:

	OVERALL		When Oregon has the ball		When Utah has the ball
Category	Oregon	Utah	UO Off	Utah Def	UO Def	Utah Off
F/+ Rk	2 (35.2%)	34 (12.9%)	1 (24.1%)	13 (14.0%)	31 (8.7%)	87 (-5.4%)
S&P+	6 (251.7)	63 (201.8)	1 (135.1)	38 (109.2)	23 (116.6)	96 (92.6)
Play Efficiency			2 (150.8)	33 (116.5)	35 (114.0)	98 (89.6)
Rushing S&P+			1 (152.1)	33 (114.7)	46 (107.5)	85 (94.2)
Passing S&P+			3 (153.8)	36 (114.7)	31 (117.5)	105 (84.9)
Std. Downs S&P+			3 (137.7)	15 (121.9)	19 (121.4)	105 (88.0)
Pass. Downs S&P+			2 (182.4)	68 (99.8)	74 (97.0)	81 (95.5)
Drive Efficiency			6 (132.4)	60 (102.0)	9 (134.6)	67 (100.7)

Oregon went up 6 ranks after dominating Stanford, and became the best overall offense in the country as a result. Not only that, Oregon is the best offense Oregon has been...ever. Not 2012, not 2007 - this right here, right now, is the best Oregon has ever been on offense. Their defense isn't even that bad, either.

Similarity scores for Utah

Here's the first part of 'which statistical measure do you like more'. S+P didn't like the matchup last week vs. Stanford at all. S+P looooooves the matchup against Utah. Oregon has 40 point advantages across the board here. The one place Oregon doesn't dominate is on standard downs - where it is only a 16-point win for Oregon.

Long time readers might remember this sounding like something familiar, and they're right: it's somewhat like what I said about Stanford in 2012. The key difference is that S+P thought Stanford was a great defense but Oregon was otherworldly. This year, Oregon's offense is otherworldly - but Utah's defense is merely good.

That being said, Utah's offense is up first, and boy is it not great. The closest team to them we've faced is Wyoming (101st), with Washington (81st) being similar. Utah is worse in almost every way compared to both of those teams save in two areas - passing downs and drive efficiency. For whatever reason, they tend to get some yards every drive. But they don't string a lot together.

Utah's defense, as I said, is merely good by S+P. UCLA (32nd) and Arizona (47th) are the next closest teams to Utah. Utah is worse in all ways than those teams save in one odd place - standard downs. For whatever reason, Utah is good at making people less successful early in the count or on close down and distance drives. They are otherwise not insanely great at anything.

Similarity scores for Oregon

Oregon's offense is much, much better than anything Utah has faced by S+P standards. The closest team is USC (23rd), followed by Arizona State (34th) and UCLA (35th). USC is probably not a good comparison given how much they've improved as the season has gone along, but Arizona State and UCLA are both fair. Still, Oregon eclipses those teams by more than 20 points each.

Oregon is also a better team on defense - and amusingly has a better defense than Utah does. The nearest neighbors are UCLA (32nd)  and USC (36th). Oregon is worse against the run but much better against the pass, and is especially good at getting turnovers compared to any of those teams. Expect that trend to continue.

Oregon's offense vs Utah's defense

Oregon has an absurd 40 point differential almost everywhere against Utah. While Oregon's offensive prowess is probably inflated due to great games against a couple of good teams, it's still going to be a very good team. More importantly, Utah's defensive strength here really isn't that great. They're not significantly better than anything Oregon has faced this year and much worse than Michigan State, for instance. It will be interesting to see if Oregon gets stopped early in the count, as Utah only has a 16-point differential on standard downs. Not that it matters, as Oregon has a ridiculous 80-point differential on passing downs, one of Utah's biggest weaknesses on offense. As far as S+P is concerned Oregon's efficient offense isn't going to be stopped by Utah any time soon. Marcus Mariota should have a great game.

I did mention last week that S+P gets weird when it gets too high on values, and that's true this week as well; I don't think Oregon's offense is insane. I do think they can play at that level some times - that's probably their ceiling. However, unlike last week where it had a great offense against an absurd defense, this week has a great offense against a good defense - and more often than not, S+P gets that right.

Oregon's defense vs Utah's offense

Oregon still has almost as much of an advantage on their defense as they do on offense. When these types of comparisons happen it tends to be the case that Oregon wins in a massive blowout; this looks like a comparison between Oregon and, say, WSU in 2012 or Cal last year or Colorado pretty much every year. The weakness of Oregon's defense is their run defense; expect Devontae Booker to get his yards as he is exactly the kind of bruising back that Oregon has some trouble with from time to time. Still, Oregon had an advantage here as well.

How FEI sees the game:

Beating down a rival in a methodical, efficient execution that makes Robespierre proud is the best way to FEI's heart.

	OVERALL		When Oregon has the ball		When Utah has the ball
Category	Oregon	Utah	UO Off	Utah Def	UO Def	Utah Off
F/+ Rk	2 (35.2%)	34 (12.9%)	1 (24.1%)	13 (14.0%)	31 (8.7%)	87 (-5.4%)
FEI Rk	1 (.302)	10 (.228)	2 (.873)	2 (-.694)	36 (-.284)	84 (-.174)
Field Position	22 (.537)	8 (.562)
Raw Efficiency	10 (.223)	30 (.092)	1 (.963)	17 (-.439)	75 (.116)	104 (-.331)
First Down rate			2 (.817)	40 (.620)	118 (.776)	95 (.620)
Available Yards rate			1 (.641)	29 (.370)	94 (.519)	112 (.363)
Explosive Drives			3 (.256)	19 (.076)	53 (.118)	116 (.065)
Methodical Drives			20 (.183)	99 (.174)	119 (.212)	60 (.141)
Value Drives			2 (.603)	15 (.259)	92 (.450)	101 (.308)
Special Team rank	17 (1.654)	5 (3.049)
Field Goal efficiency	95 (-.268)	4 (.709)
Punt Return efficiency	18 (.088)	4 (.246)
Kickoff return efficiency	83 (-.178)	74 (-.168)
punt efficiency	20 (-.229)	9 (-.319)
kickoff efficiency	11 (-.302)	106 (-.034)

FEI loves Oregon even more after that Stanford beatdown, and it's not that surprising. Given 9 offensive drives, Oregon had 6 TDs, 1 FG, 1 interception and 1 punt. That's about as dominant as you can hope against a top 10 defensive team. It was quite possibly the best performance I've ever seen from any Oregon offense.

Game Factors: the best and worst of the teams

I'm going to try breaking out the graphs by type and use the actual values instead of rankings to make it a bit more apparent what's going on.

Utah's defense is very good, and has had good games every week this year. They're also doing better as the year goes on. Meanwhile, Oregon has some fairly big peaks and valleys, but is still very positive every week.

Oregon also is the owner of the two best offensive games by FEI standards this year - against Stanford and against Washington.

Both of these teams are consistently good. Oregon is consistently better.

This one, however, is all over the map. It's clear that they're both 'okay' as far as that goes, but the variation for both teams is large - and is especially large for Utah. Oregon, meanwhile, hasn't had a whole lot of highs on defense ever. I don't think that'll change this game either.

What's interesting to me is to see how similar these graphs actually look to each other. Both teams are trending up in game efficiency as the season goes on, both are good, and Oregon is just a bit better everywhere overall. That, right there, is likely the difference.

From a game split perspective, Oregon's win over Stanford was their best offensive performance (no surprise) and an okay defensive one. Utah has been absurdly consistent on game-to-game performance, where their offense does enough and their defense does a bit more. Neither team has been exceptionally reliant on special teams or turnovers to win any game, though Utah probably should have lost to Washington State by a bit more.

Similarity scores for Utah

Utah is not on the good end of offense - they're worse than Stanford (66th) and better than Washington (108th). They're a far cry from the good teams in the PAC-12 and are going to be the easiest offense we'll have to face for the rest of the year.

I literally copied and pasted this sentence and replaced Stanford with Utah from the previous stats article because it's exactly the same:

Not particularly surprising, Utah will be the hardest defense we've faced this year. The next closest is Michigan State (19th), though they differ by more than 200 points. Washington is also close in comparison (24th), though even bigger of a gulf there (300 points).

As we used to say, every defense looks good in the PAC-12 until they've played Oregon. Stanford dropped to 27th after their game against Oregon. FEI thinks this is going to be a great challenge - but FEI thought that precisely about Stanford last week. FEI was very, very wrong about how the game was going to turn out, and while it thought Oregon would win, it thought Oregon would win with small advantages on offense and big ones on defense.

Similarity scores for Oregon

Oregon rocketed up the FEI charts to #2 this week, and as a result they're the top in the PAC-12. Utah has faced some demons, however - UCLA (7th), Arizona State (8th) and Washington State (14th) are all quite good. None are quite as good as Oregon, which has a 240 point advantage over the next closest team. Oregon also has the highest raw offensive efficiency in the country.

Oregon is a peer of UCLA (34th) and Arizona State (40th) - and in particular UCLA and Oregon are very similar in the kinds of teams they've faced. Both teams are reasonably comparable.

Oregon's offense vs Utah's defense

Oregon is, once again, surprisingly decent at methodical drives this season. Utah is shockingly bad. I would expect Oregon to get some fairly long scoring drives here and there. Given the numbers, I would expect Oregon to have decent advantages across the board, though Utah is especially stingy on big plays.

Oregon's defense vs. Utah's offense

Oregon had a 300 point differential vs. Stanford last week; they have over 400 points this week, and that's without Dres Anderson as well. Oregon should really be able to win almost everywhere; even Oregon's major weakness (explosive drives) is hampered by Utah's inability to be explosive.

Special Teams

Utah is one of the best in the nation in special teams - but this year, so is Oregon. Oregon and Utah are very closely matched everywhere. Utah does great at punt returns but Oregon is great at punting coverage. Utah is great at punting but Oregon is great at returns. Oregon is much better at kicking the ball off than Utah is returning it, and has a small advantage when returning it themselves (mostly because Utah is really bad). Utah is great at field goals, so if it comes down to a kick Utah might have a small advantage there.

Utah is also one of the best in the nation in getting good field position - but so is Oregon. This is surprisingly a wash.

So what does this all mean?

This is what I said last week:

I think Oregon loses. If Oregon wins it will likely be because of a high turnover ratio, a big play on special teams and an offense getting a couple of explosive plays and a couple of 10+ play drives. If Oregon loses, it'll look a lot like 2012 did - with Kevin Hogan scrambling to get first downs, crappy penalties on Oregon, poor execution and Marcus Mariota running for his life.

What I want is for Mariota to have a true Heisman moment in this game - some absurdly long scramble, an amazing pass play or two, and taking over this game. I want this with all my cold, calculating heart. But the stats don't think it's going to happen.

I was so very, very wrong.

Oregon played their best offensive game ever from an advanced stats perspective. They did give up a lot of yards but didn't give up all that many points. They were ruthlessly efficient on offense. They were good enough on defense.

Marcus Mariota had a true Heisman moment. Typical of Stanford to have a kind of crap season this year and ruin it, because if this was against a 7-0 Stanford everyone in the world would be noticing how great Mariota was. I mean, look at this run again:

Everything is great here. Everything. Mariota does a great job on the fake. Everyone blocks well. Receivers are pancaking their targets. He could have been running for days. He doesn't even hit full speed.

Stats don't predict games like this all that well. They certainly didn't last week. This was a game that clearly Oregon had thought about and planned for many weeks. While I was more negative than the stats would say (Bill Connely had this as an 81% Oregon win) none of the stats thought that Oregon would win by 29 or have the most dominant offensive game thus far this season across all of college football. Oregon jumped up absurdly in advanced stats rankings because of this game. That just happens.

So I was wrong, and the stats were wrong, and the prognosticators were wrong. The stats have just not been good when it comes to predicting how Stanford is going to go. This year it went a different way.

For the Utah game, both systems think Oregon is going to win. FEI likes Oregon a bit on offense and a lot on defense. S+P likes Oregon absurdly on offense and a bit on defense. I think both are somewhat overrating how good Oregon is, but Oregon should do well no matter how you look at it. As always, nonpredictive events such as injuries, fumble luck and the like can hurt Oregon and make it a closer game than it should be. That being said, the line is currently Oregon -9.5, and I don't see with these stat differentials how Oregon can win by less than 2 scores. Oregon typically has problems with teams who can run decently but pass well, and Utah can run well but pass exceptionally poorly. While I suspect that Utah will control the ball for long stretches and run consistently, I don't see them being able to do this all day. My gut feeling is that FEI is more right, again, and the key will be Oregon's defense being excellent against a bad Utah offense.

I think Oregon wins, and I think Oregon covers.