Statistically Speaking: Utah vs USC

By the morning of September 14, USC looked like a paper tiger. Lowly Boston College had just pounded the Trojans to a paste, behind 452 rushing yards from seemingly every conceivable BC player. 'SC hadn't turned the ball over or made big mistakes, they plain got whooped. They've played four games since then, however, and have looked like the ferocious, talented squad that we expected preseason. They've battened down the hatches and laid waste to the running games of the Sun Devils, Beavers, Wildcats, and Buffaloes, allowing 2.25 yards per carry in those games. Cody Kessler has been leading an excellent offense to big points and has only thrown one interception over 233 passing attempts. Have they righted the ship? Or will Devontae Booker and the Utes be able to replicate BC's performance and run the Trojans into the ground?

In my view, the question of what the Utes' rushing attack will be able to do to the Trojans is what will decide the outcome of this game. Basic statistics don't paint a rosy picture: the Utes 4.55 yards per carry puts them squarely in the middle of the pack, and USC is allowing 4.42 yards per carry; not great, but not abysmal either. Boston College has put up 5.7 yards per carry this season (15th nationally), so their performance against USC was not an aberration. We have some advanced metrics that can shine more light on what's really going on here, though.

Adjusted Line Yards is a statistic borrowed from the NFL that tries to isolate the line's contributions to a run game or run defense, and is adjusted for quality of opponent. Power Success rate measures a team's ability to turn a third or fourth and short into a first down, while Stuff Rate measures runs for a loss or that were stopped at the line of scrimmage.

O-Line Performance	Adj. LY	Power Success %	Stuff %
Utah	58th	72% (42nd)	20.2% (87th)
USC	38th	68.8% (63rd)	19.4% (73rd)
Boston College	26th	75.9% (27th)	17.4% (42nd)

The Utes don't compare favorably to BC here, the only impressive number is their ability to power ahead for a yard or two when it's needed. These numbers try to isolate offensive line performance and leave out the running back, but the teams we see at the top of the list are those which feature top-shelf NFL tailback prospects, exceptionally dynamic running quarterbacks, or both: teams like Georgia, Alabama, Auburn, Nebraska, and Oregon. Utah's offense has changed so much since entering Pac 12 play and giving Devontae Booker the starting job that while overall stats should be a concern, they aren't the end of the discussion. What hasn't changed are the defenses:

D-Line Performance	Adj. LY	Power Success %	Stuff Rate
Utah	38th	86.7% (125th)	19.7% (65th)
USC	29th	70.8% (79th)	18.2% (83rd)

USC's rush defense looks like the real deal, and Boston College more of an anomaly than a trend. USC hasn't been great at stopping short runs for first downs, and struggles to get into the backfield, but overall their line has been able to limit opposing runners and perform at an above-average level.

We've got one more set of numbers to look at which should hopefully illuminate the rushing offense question: like opponents. USC has played both Fresno State and Oregon State, a significant sample size. I've done my best to pull sacks out of the drive data and isolate true rushing yards.

Rush Defense, Like Opponents	Carries	Yards	Yards Per Carry
Utah	54	147	2.72
USC	51	246	4.82

These numbers show a sizeable difference between these two rush defenses. USC's rushing defense against Oregon State wasn't quite as sterling a performance as it first seemed; 2 sacks for 22 yards significantly skewed their numbers. Meanwhile, they really struggled to contain Fresno State's running attack, especially compared to the Utes.

Rush Offense, Like Opponents	Carries	Yards	Yards Per Carry
Utah	94	537	5.71
USC	99	502	5.07

These two teams both have excellent feature backs, and it showed in similar performances against like opponents. Javorius Allen is a heck of a player, and we should expect him to get his, even against the stiff Utah run D.

These numbers don't shed as much light as might have been hoped on this critical matchup, but they do suggest that USC's run defense has improved, but hasn't totally transformed since the BC game. That gels with expectations: they are playing mostly the same athletes at the same positions, and coaching only counts for so much. The Utes run defense has been successful in a totally different way than their pass defense. While the Utes specialize in creating havoc to disrupt the passing game, they have not excelled in creating similar plays against the run. Sound play by the secondary and swarming to the ball with speed has limited explosive plays for opposing offenses, which has kept the rushing yards totals in check.

The Utes probably aren't going to put up 400+ rushing yards; USC's loss to Boston College looks like a statistical anomaly. The weaknesses in the Trojan's run defense are real, however. Expect Devontae Booker to continue getting the ball on 65% of snaps, and to make lots of plays. Another 200 yard game isn't out of the question for Booker. If the Utes want to win, they are going to need that kind of night from The Beast.

Advanced Metrics Primer:

This is one of the most even matchups by advanced stats we will see all year. For every strength of the Trojans, there's a strength of the Utes to counter, and each USC weakness is up against a Utes team that just hasn't made it click, except in one area: special teams.

ESPN FPI

	FPI (Rank)	Off. Eff. (Rank)	Def. Eff. (Rank)	SpcT. Eff. (Rank)	Overall Eff. (Rank)
Utah Week 7	14.1 (22)	3.45 (62)	7.09 (25)	7.87 (1)	19.00 (14)
Utah Week 8	14.0 (26)	2.78 (61)	7.91 (21)	7.53 (1)	18.21 (18)
USC Week 8	19.3 (13)	15.09 (12)	2.83 (43)	0.26 (58)	18.18 (19)

FPI gives USC a 55.6% chance to win this game.

The Utes took a very tiny dip in the FPI rankings last week, although they showed improvement on both offense and defense, other teams appear to have improved more. The Trojans have an impressive FPI but lag behind the Utes in efficiency, riding a potent offense to their high rating. I place more weight on efficiency in these metrics because ESPN is very opaque about what goes into FPI, and their overall numbers look a bit too much like fan power polls for my taste.

FEI

	FEI Rank	Off. FEI Rank	Def. FEI Rank	Spc. T. FEI Rank
Utah	8th	83rd	8th	1st
USC	3rd	11th	36th	80th

FEI is a metric which favors teams that have played tough schedules and tend to win with turnovers and special teams. Predictably, it favors both the Trojans and Utes. The story it tells is the same as ESPN's numbers: can the Utes defense slow down USC enough, and the special teams give enough short fields, that the Utes can keep up?

FEI closely tracks special teams performance, and ranks the Utes as being in the top 10 in Field Goal Efficiency, Punt Return Efficiency, and Punt Efficiency. USC has struggled in the punting game, allowing opponents to start on the 30 yard line. They struggle to get touchbacks or cover punts, and only 17% of opponent drives start inside the 20. The Trojans rank 116th in yards per kick return, and 113th in yards per punt return. They have yet to give up a special teams score or mammoth return, but they have yet to match up against a team like Utah, either.

S&P+

	S&P+ Rank	Off. S&P+ Rank	Def. S&P+ Rank
Utah	60th	92nd	36th
USC	25th	19th	37th

These numbers do not favor the Utes. You will notice an absence of special teams here, and it's worth noting that these numbers place a premium on record and success in traditional ways. This week's top 25 in S&P+ include such gridiron titans as Virginia Tech, Marshall, Tennessee, and our old friends the Michigan Wolverines, who are 23rd, 37 spots ahead of the Utes team who walloped them in every phase. S&P is useful in that it tells you that USC has been winning in a traditional way, while Utah has relied upon an unusual formula (special teams and sacks) to get to 5-1. It doesn't say much about who the best team is.

Massey and the Bookies:

The Massey aggregate rating has the Utes 23, and the Trojans at 21. Vegas opened with a pick 'em, although USC was favored by a point or so in a few places. That line hasn't moved, although ESPN's PickCenter suggests that Utah is the smart money bet. The experts can't seem to get a handle on this game, and neither can I.

Stat-Head Prediction:

This game is as tight as they come. The Utes' defense looks good enough to put a stop to any offense, but Cody Kessler, Nelson Agholor, and Javorius Allen have been among the best players in the nation. Meanwhile, the Utes offense has struggled to find success since entering Pac-12 play; uncertainty at quarterback, dropped passes, and poor pass protection have all been serious problems for the Utes, and it shows in their advanced metrics. At the same time, USC's defense has been decent but has shown some exploitable flaws in their two losses this year. If these two areas balance out, the Utes have an undeniable advantage on special teams: USC struggles to create touchbacks or cover returns, and don't have a player like Andy Phillips on the roster.

I'm going to pick the Utes, but it's a homer pick. This game is likely to come down to a handful of big plays, unless Travis Wilson can find his old form and put together a special performance. I can't imagine that the Utes are going to go another game without catching a well-thrown ball in a big moment, and if you read my article on the QB controversy, you know that I think the Ute's offense is much more efficient and explosive with Wilson taking the snaps. I expect some big plays from the Utes returning punts and kickoffs, 150 yards or more from Booker, and a few big 3rd down conversions or TDs from Wilson to be the difference.

30-27 Utes