Previewing the USC defense

USC is under new management, with Steve Sarkisian and Justin Wilcox in town. How much has the USC defense changed under a new coordinator? Not a ton. It's a luxury to coach a team like USC, because you don't necessarily have to do anything fancy scheme-wise. There's a certain justified arrogance when you can say 'Your 11 vs. my 11: we'll beat you one on one."

Obviously it's an oversimplification to say that USC plays a completely vanilla defense. Below you'll read about some interesting positional flexibility that can confuse somebody trying to read a defense. But ultimately this is a team that doesn't need fancy stunts or unexpected blitzes to get you off the field.

Starters

DE: Jr. Leonard Williams ; NT: Jr. Antwaun Woods ; DT: Jr. Delvon Simmons OR Jr. Claude Pelon ; RUSH: So. Scott Felix

The focus will obviously be on Leonard Williams, the best defensive lineman in the conference outside of Seattle, Washington. He's likely heading to the NFL when the season ends, and will likely go in the first round. He's not necessarily crazy disruptive (4.5 sacks, 6 total tackles for loss) but he demands a ton of attention, has great instincts and speed, and rarely gets blocked out of a play. I would expect Cal to double team him on most every passing play, and to run away from him whenever possible.

The defense is technically a 4-3, but in a lot of ways Scott Felix is a pass rush specialist outside linebacker. He's basically in a platoon with J.R. Tavai, and both will see the field, particularly on clear passing downs. Woods and Simmons are both solid if unremarkable clogs in the middle. Pelon must be coming on in practice, because he hasn't has much playing time so far this season to be listed as a co-starter.

SOLB: So. Su'a Cravens ; MILB: Sr. Hayes Pullard ; WILB: Jr. Anthony Sarao

Cravens is a big confusing question mark for the purposes of this preview. For one thing, he's recovering from an injury that likely won't inhibit him but still adds a bit of doubt. But beyond that, he has positional flexibility that might come into play. He's listed as a starting linebacker, but is also 3rd on the depth chart at strong safety . . . behind a starter that might not play. Does USC move Cravens to safety because they prefer their depth options at linebacker over their depth options at safety? I have no clue, and we won't find out until game time.

Regardless of where he plays, Cravens is at least the 2nd best defender USC can offer. Leonard Williams gets the plaudits (and the attention from offensive linemen) but Cravens has more sacks, tackles, and tackles for loss to go along with his ability as a cover safety.

Pullard and Sarao are less explosive than the disruptive Cravens, but still rack up plenty of tackles. It might be that USC goes with a sort of 4-2-5 formation with Cravens acting as a linebacker on some plays, and a safety in others. Simply stated, it's nice to have a guy with enough strength to rush the quarterback and enough athleticism to cover pass receivers.

CB: Sr. Kevon Seymour ; SS: Sr. Leon McQuay OR John Plattenburg ; FS: Sr. Gerald Bowman OR So. Leon McQuay ; CB: Fr. Adoree' Jackson

You will note that Leon McQuay is listed as possible starter at both safety positions, which is odd.If Plattenburg is indeed out with his thigh injury, then Bowman will likely start at free safety and McQuay at strong safety, with Cravens and backup cornerback Chris Hawkins available for nickel packages.

There are no big names in the USC secondary this year, which is odd for two reasons. First, you just expect USC to have a big name in every unit on the field. Second, USC's pass defense has been good enough (more on that below) that you would expect somebody to get some buzz.

Stats seem to indicate that teams throw at Seymour more often than Jackson, who is evidently holding up just fine despite playing as a true freshman. USC can, when fully health, rotate 4 safeties and none of them have collected a ton of stats. Still, USC's pass rush hasn't been overwhelming (a bit more than 2 sacks/game), so this is clearly a unit that can maintain coverage for a bit of time.

Season So Far

4.6 yards allowed/play in a 52-13 win over Fresno St.
6.1 yards allowed/play in a 13-10 win over Stanford
7.4 yards allowed/play in a 37-31 loss to Boston College
3.2 yards allowed/play in a 35-10 win over Oregon State
8.0 yards allowed/play in a 38-34 loss to Arizona State
4.7 yards allowed/play in a 28-26 win over Arizona
4.2 yards allowed/play in a 56-28 win over Colorado
4.5 yards allowed/play in a 24-21 loss to Utah
4.6 yards allowed/play in a 44-17 win over Washington State

USC has played 9 games so far this year, and in six of those games they have held their opponent below 5 yards/play. That's excellent, and it's pretty hard to lose a game when your defense plays that well (lol Utah game).

In three games this year, USC has allowed 6 or more yards/play, which is pretty bad. The Cal defense allows 6.2 yards/play on the season as a basis of comparison. That's bad, and it's pretty hard to win a game when your defense plays that poorly (lol Stanford).

So what is it about those three games? One came against Stanford*, as traditional and pro style as they come. One came against Boston College, a heavy run based, pistolish offense featuring a dual threat quarterback. One came against ASU, who runs a pretty standard spread offense.

In other words, there aren't many patterns. USC has played good offenses and gotten torched (ASU, Boston College) and has played good offenses and excelled (Arizona). USC has played bad offenses and gotten torched (Stanford) and played bad offenses and excelled (Oregon State, Utah, Fresno St.). Style doesn't seem to matter, patterns are infrequent. Utah and Boston College found success running the ball (albeit in completely different ways) while ASU and to a lesser extent, Stanford, found success passing.

About the only thing predictable about anything related to USC since Carroll left is how they play against Cal. God I hate USC.

*I know what you're saying. "Nick, Stanford scored 10 points in that game. Please don't tell me you actually think the Stanford offense played well?!" I actually do, sorta. Stanford moved the ball with ease between the 30s, and their red zone failures were a combination of bad luck, bad special teams, and Cro-Magnon football strategy from David Shaw. USC's defense gets a little credit for coming up with red zone stops, but just as many demerits for allowing a bad Stanford offense to move the ball so easily in the middle of the field.

Against the pass

2013: 5.8 yards/attempt allowed, 7th in the country
2014: 6.2 yards/attempt allowed, 28th in the country

Above I noted that USC's pass rush isn't going gangbusters, as they are 59th in the country in total sacks. However it's worth noting that USC very very rarely blitzes, and it's easy to understand why. When you have the type of talent USC typically does, you can count on winning one on one battles without risking blitzes that might leave your players in coverage exposed. And when you have Williams to always occupy two blockers, all the better.

So yeah, I would expect a team that can routinely leave 7 men in coverage because their 4 man rush can eventually get to the quarterback to fare well, generally speaking.

Against the run

2013: 3.95 yards/attempt allowed, 46th in the country
2014: 4.01 yards/attempt allowed, 53rd in the country

Boston College gained 452 yards on 54 carries - 8.4 per carry! - and if you remove that extreme outlier performance, USC only allowed 3.1 yards/run. Obviously, when you take the worst performance out of any team's resume they tend to look better. But I think it's necessary to do in this case, in part because the BC game was so outlandishly cartoonish compared to everything else USC has done this year, and in part because it seems unlikely Cal can replicate what the Eagles did anyway.

It is fair to note that USC hasn't exactly faced a running quarterback since that game, as it's a down year for Pac-12 running quarterbacks, and USC hasn't faced the only 2 that are particularly dangerous in Marcus Mariota and Brett Hundley. Does that mean that Tony Franklin should break out the Rubenzer machine? Maybe, maybe not.

Cal fans were rightly excited with the showing of the running game against Oregon State. But that doesn't mean we should expect anything near the same level of success tonight. OSU has the worst defensive line in the Pac-12, while USC might have the best, depending on how you feel about Washington and Stanford.

Advanced stats

2013 S&P: 4th in the country
2014 S&P: 38th in the country

2013 FEI: 6th in the country
2014 FEI: 24th in the country

The advanced stats show a slight but meaningful decline from last year that more or less matches a slight but meaningful decline in USC's traditional rate stats. When you look at the nitty gritty data, S&P thinks that USC does better against the pass, and against passing downs, than against the run and in standard downs. I'm a bit skeptical - I think the Boston College weirdness might be unduly influencing the numbers - but the larger point is sound: Stay out of obvious passing downs. That's important in every game but even more important against USC. I like our chances of converting 3 and 8 against Washington St. or Colorado. Not so much the Trojans.

FEI component data indicates that, ASU 4th quarter comeback aside, USC doesn't tend to allow a ton of big plays. Cal's recently developed ability to move the ball incrementally down the field will likely be critical, although we're all excited to see what kind of impact the return of Davis and Lawler have.

Turnovers

2013: 23 forced turnovers (17 interceptions, 6 fumbles), 47th in the country
2014: 17 forced turnovers (11 interceptions, 6 fumbles), 36th in the country

What is interesting about USC this year is how evenly distributed all of their interceptions are. Nine different players have picked off a pass, but nobody has more than two. We're reaching broken record territory, but a huge part of the battle is keeping Williams and Cravens away from Goff when he's in the process of throwing. Williams has forced 3 fumbles, intercepted a pass, and has likely played a critical role in a few other interceptions aided by pressure.

Conclusions

I wasn't terribly optimistic before I started this preview, and after diving into the numbers and really looking at USC's performances, I'm actually a bit more pessimistic. The Trojan defense has played good or better defense for pretty much the entire year outside of five quarters: the entire game vs. Boston College and the 4th quarter against Arizona State, when USC allowed Mike Bercovici to throw for 243 yards on three straight touchdown drives.

If you're looking for the best reason for hope, it's that 4th quarter against Arizona State. Right now, USC's secondary is a bit stretched. One starting safety might be out. Another, they would prefer to play as rush linebacker. But they are playing a team that deliberately tests your two deep. If Cal can keep the game close for 3 quarters, run lots of plays, and keep USC on the field, then there's the chance that USC's depth gets exposed and Cal can have their way in the passing game late.

The path to an upset victory would almost certainly have to include at least the semblance of a running attack. I'm very, very encouraged by what Lasco and the line have put together of late, but USC is a tall order. If Cal gets any traction on the ground it would be a huge win and a great sign of things to come both this year and next.

But the biggest test of all? Two interconnected battles:

• Every single one on one battle at the point of attack between Cal's offensive linemen and USC's defensive linemen on passing plays
• Jared Goff's ability to read the defense and quickly make the right read and throw