Behind the Numbers: Putting Stanford's Passing Game Under the Microscope

Disasters have a way of obscuring perspective and creating hyperbole.

So, naturally, it’s filled the air following a dire offensive performance against Oregon State. While there are some problems for the Cardinal in both the passing and running phases of the game, it’s good to reassess relative to expectations -- what did we expect out of the offense, and what happened?

The headline results show improvement -- the offense is nearly 5 points per game better; it’s up to 6.23 yards per play, which lumps it in with the Pac-12’s vast second tier -- from Arizona State to Washington; it converts 3rd downs 10% more frequently... it is, simply put, a better offense than last year.

But the better question is: is the offense maximizing its potential? It’s hard to answer counterfactuals with certainty, but the answer seems to be no. Let’s start with the passing game and Kevin Hogan. We had roughly three expectations for Hogan and the passing game: first, that he increase yards per attempt; second, that the passing offense feature more explosive plays; third, that he be sacked less. Others might have more or fewer requirements, but this is a rough consensus. Anyway, the grade -- rather confoundingly -- seems to be good when viewed at a glance.

Yards Per Attempt: Hogan is averaging 8.5 yards per attempt this year; last year, he chipped in at 7.2 yards per attempt. Going into the season, I was hoping the offense could scratch out something close to 8 yards per attempt. Hogan’s yards per attempt actually ranks second in the Pac-12 through Week 9. So that's success -- stirring success, in fact.

Explosive Pass Plays: Last year only 8.77% of Stanford's pass plays went 20 or more yards. That’s very poor; it was probably the main offensive weakness, as it made drives a gruesome slog and didn’t dissuade defenses from stacking the box heavily. This year, explosive pass plays occur on 14.13% of dropbacks. That’s a big increase -- particularly since 2011’s rate was 12.2% and 2010’s was 12.66%. (Unfortunately, cfbstats.com does not have explosive play data before 2010; my guess is that 2009, during Chris Owusu’s prime, was the most productive year of all.) Again, success! So the mystery continues: where are the problems?

Sacks: And here’s where we start to see the problems. Hogan was very easily sacked in 2012: 6.75% of his dropbacks ended in sacks, as compared to 3.29% of Nunes’. The headline number shows progress -- only 4.66% of dropbacks have ended in a sack this year -- but it’s once you begin digging that the problems become obvious.

For one, for an offensive line expected to be elite -- if not the best in the country -- it doesn’t quite measure up to its predecessor’s standards: 2.57% of dropbacks ended in sacks in 2011 and, incredibly, only 1.56% ended in sacks in 2010. Still, perhaps those expectations were unrealistic (Andrew Luck covers a lot of mistakes, after all).

But let’s think more specifically here: under what conditions did Stanford's dire games occur? Of the really bad games during the past month -- that is, Washington, Utah, and Oregon State -- two of the three occurred on the road.

You may be guessing by the direction of this post that there’s a problem with sacks on the road, and you, hypothetical reader, are absolutely correct: 6 of Stanford’s 9 sacks have occurred on the road, meaning that 6.38% of Hogan’s dropbacks end in sacks on the road. If anything, this stat understates things. As you’ll see later, Hogan frequently had to contend with pressure in the Oregon State game, and the offensive line picked up two false start penalties.

So something seems to be happening on the road with the protection? What? I do not know; I’m not an offensive line expert. Still, it’s jarring to see Kevin Danser dump-trucked -- even by a player as good as Scott Crichton -- in the sack he gave up in the first quarter. And it’s odd when you think about the schedule: ASU, UW, UCLA... three sacks, total. Army, Utah, Oregon State... six sacks. But I think there’s more to it than that. The simple fact is that Hogan is the most volatile quarterback in the conference.

Let’s examine two different ratings systems -- ESPN’s adjusted QBR and the NCAA’s pass efficiency rating and categorize QB games into "brilliant" or "dire" categories. Brilliant games are single-game scores that, if extrapolated across a season, would land the QB in the top ten in the country (so 82.2 on ESPN’s adjusted QBR; 164 on the NCAA scale). Dire games are single-game scores that, if extrapolated across a season, would put the QB below the median qualifying QB (so 60 and 135, respectively). Here’s how the Pac-12 QBs rank (again, through week 9):

Not bad, right? Hogan can be quite brilliant on his day. But let’s examine the bad games:

Hogan -- when he’s good he’s quite good indeed; when he’s bad, well… Put it this way: if you believe ESPN’s QBR this season, Hogan has either been brilliant or awful literally every game. What’s behind this? There’s the sack issue, mentioned above. That might play a role.

But I’m beginning to wonder whether the offense has embraced the explosive play too much. On one hand, it’s undoubtedly worked at its stated aim: there are many more explosive plays and defenses are much more reluctant to stack the box than in 2012. But look at the effect it’s having on Hogan’s efficiency. Hogan’s highest pass completion percentage this year was against UCLA, for 72%. Hogan’s pass completion percentage for 2012 was 71.7%. Deep passes are much more variable, and Hogan’s attempting many more of them. Let’s compare three games from this year, along with UCLA Round II from last year. (Remember, I track how far the ball goes in the air, rather than how far the play itself goes):

The 2012 Pac-12 championship game featured all of two shots downfield; Hogan took three shots downfield in Oregon State... a game in which he attempted 18 passes. So there’s been a marked shift to downfield passes, which will, inherently, create a more volatile offense. And the Oregon State game marked a particular extreme -- the 2013 UCLA game featured a much more balanced and preferable approach (though I may be letting the result bias me too much). And the use of personnel have, at times, accentuated some of the ineffectiveness and volatility in the passing game approach.

Unless I’m mistaken, the Washington State game was the first time Michael Rector has been targeted on a throw under 25 yards or so. That’s the type of statistic that underlines the occasional vertical monomania of the passing game. And who is getting targeted short? For some reason, Jordan Pratt was targeted three times and Charlie Hopkins once against Oregon State. While pass interference was a factor in their missing some of the catches, it’s nevertheless strange to be throwing out a converted defensive end and a walk-on wideout as targets in critical situations, absent some sort of injury crisis.

But there’s no crisis; there is, in fact, a pretty solid stable of playmakers, many of whom don’t get enough touches. The offensive bandwidth is being allocated incorrectly. (By the way, most people missed the egregious part of the Utah playcalls. It wasn’t calling a pass when a run should’ve been called. It’s not an illogical idea, really: you want the element of surprise from time to time. But one of the targets was Eddie Plantaric, another converted defensive end. If your 3rd-and-2 pass call to seize a victory and preserve an undefeated season hinges on passing to a converted defensive end, it’s time to rethink your play call.)

It’s hard to know at what level the passing game is the intention of the coaches and what’s the execution of the players on the field. But one element creating volatility that’s unquestionably under the control of the coaches is putting Hogan on the run. Hogan was bootlegged once against Utah and had zero runs called for him, which is the lowest total I’ve recorded for his career (though I haven’t exactly gone through every game.) This despite Hogan’s running being an effective weapon: two of Stanford’s touchdowns were aided by Hogan scrambles (which he did a lot of -- 6 scrambles leading to a run or pass, a pretty high number reflecting the effectiveness of the Oregon State pass rush).

As I’ve noted before, the number of Hogan run calls have been carefully husbanded, doubtless to create uncertainty in opponents and save hits on the quarterback. It’s a logical strategy, but it shifts risk around: it decreases long-term risk (by keeping your quarterback healthy) while increasing short-term risk (by increasing the risk of, well, losing a game). It’s a confident, aggressive play against the Beavers, especially since Hogan ran quite a bit against UCLA: Hogan ran five times, as we know, but that stat understates things -- he gave the ball on several zone-read calls.

From Hogan’s perspective, then, UCLA looks like the prototype game: a balance between explosive passes and safe ones; and plenty of opportunities to show off his best asset -- his running and improvisational ability. Can it be reacted on November 7? At least one element is directly under David Shaw’s control.