Clearing the Air: Addressing some misconceptions about the "conservative" Stanford offense

Because I am a contrarian, and I gain most of my writing energy from disagreeing in disagreeable fashion with wrong ideas, I’d like to clear up a bad idea advanced by Tom FitzGerald in the wake of the UCLA game. Here’s FitzGerald, triply wrong:

Nothing about UCLA’s defensive set-up led me to think Stanford’s run calls on three third-down plays would work. It’s fine that David Shaw loves the jumbo or "over" formation. But Ted Cruz wouldn’t have gotten this conservative. On 3rd-and-5 from the Stanford 25, Ricky Seale got just four yards in the first quarter. On 3rd-and-5 from the Stanford 30, Seale got just one in the second quarter. And finally, on 3rd-and-6 from the Stanford 36 in the fourth quarter, Gaffney was stuffed with 3:50 left and Stanford clinging to a seven-point lead.

It’s impressive that FitzGerald manages to be factually, strategically and philosophically wrong. (Well, mostly -- the fourth quarter 3rd-and-6 run was indeed really dumb.) But otherwise his writing and thoughts are leaky. Let’s start with the facts: the first two plays identified by FitzGerald were not jumbo formations. They were read-option plays out of a shotgun. This is a rather important distinction. Someone asked Shaw in his press conference why he’d chosen such conservative play calls, to which Shaw replied that he calls plays that he thinks will work.

What could have given him such a crazy idea?

Shit, maybe the UCLA game last year:

I mean,

What kind of dumb idiot,

Thinks if something’s successful

On four separate occasions

he should try it again the next year?

Now, it’s unlikely that UCLA would be so negligent in defending the read-option two years in a row, but the payoff is absolutely worth it: picking up first downs like free money abandoned on the sidewalk is a nice bounty. And here we come to a philosophical problem with FitzGerald -- and, to be fair, large swathes of fans and writers -- which is describing coaches as "conservative" or "aggressive." In most usages, the description has ceased to be helpful and instead describes the emotional state of the user: "conservative" describes outcomes or plays the commentator disapproves of; "aggressive" describes outcomes or plays the commentator approves of.

I’m not sure Google has this covered yet, but the ratio of negative-to-positive citations of conservatism in the context of football has to be more comically lopsided than a poorly baked cake. For me, "conservative" refers to attempting to decrease risk or variance, possibly at the expense of decreasing payoff; whereas "aggressive" refers to increasing payoff, at the expense of increasing risk or variance. So, for me, criticizing a coach as conservative or aggressive only makes sense in light of the tradeoff: how large is the decrease in risk as compared to the decline in payoffs?

Of course fans prefer aggressive teams -- they’re more fun -- but you shouldn’t confuse aesthetics with effectiveness. (Then again, David Romer has compelling evidence that coaches are too conservative with respect to fourth down, so it’s not as if fans are totally wrong.)

In the read-option case above, the decline in risk is pretty high, and yet the payoffs remain quite nicely-sized: it’s really excellent to be able to pick up cheap first downs by running the ball; it’s not exactly crazy to check and see if your opponents’ demonstrated weakness from your last meeting is still exposed. (For example, Brett Hundley reprised nearly all of his errors from last year.) UCLA has figured out the basic read-option; Shaw confirmed that and never called the play again after the second quarter. The results weren’t there, but then again you shouldn’t be results-oriented in your thinking.

Other notes about Stanford's win over UCLA:

1. Shaw is particularly likely to catch heat for being conservative; an image has been established, and as all good politicians and high schoolers know, it’s tough to shake an image and tougher to avoid having your actions being judged through that filter. Shaw was absolutely far too conservative in the fourth quarter; on the other hand, precisely no one is giving the coaching staff credit for rolling out a hurry-up, no-huddle drive in the first quarter. (It was mostly successful, too, were it not for Wilkerson’s option pitch adventures.)

The key to that drive, Shaw noted in his press conference, is Ryan Hewitt. If you -- like me! -- were wondering why Hewitt has been moldering through half the season, apparently the coaches have been watching, and waiting. Hewitt’s versatility is perfect for a hurry-up no-huddle offense: it allows the offense to position itself in several different formations with the same personnel.

2. Staying in contrarian mode, I remain worried about the defense. Statistically, excellent, of course. The odd trouble is that UCLA was insufficiently conservative, oddly enough. It came up with the original idea to trouble the Stanford defense schematically in the second tilt with Stanford: constant motion out of the backfield, serving either to unclog the middle (leading to Johnathan Franklin runs up the gut) or outflank the defense via swing passes. On Saturday, Stanford defended the latter and dared UCLA to try the former. In response, UCLA abdicated both.

Oddly, after Malcolm Jones gashed the Stanford defense the first drive, with 3 carries for 28 yards, the offensive braintrust decided that was too much success too easily gotten and passed for the rest of the day. (Hundley missed a ton of passes you’d think he’d hit and had a dumb pick in which he stared down a receiver -- you’d think better teams would exploit it. Then again, last year Hundley… missed a ton of passes you’d think he’d hit and had a dumb pick in which he stared down a receiver. Same stuff, different safety. The convenience of UCLA’s offensive penalties cannot be overstated either.)

3. All this aside, it’s a simple fact that, judging simply on the difference between the Pac-12 championship game and this year, the team is better. Last year, the team used a UW-esque great escape, winning off a de facto Ed Reynolds pick six and great field position -- yards per play were 4.92 for Stanford and 6.59 for UCLA. This year, it was 5.59 to 4.03. If anything, given the number of weird errors and uncharacteristic mistakes, the margin in yards per play and score could’ve been larger -- imagine what a blowout it’d be if Ty Montgomery or Devon Cajuste hangs on to the ball. (Or Wilkerson … or Gaffney …) The most powerful difference was in the offensive line, which was often dominated by the UCLA interior line in 2012 and instead did the dominating in 2013. In 2012: 3 sacks; in 2013: 0. Sam Schwartzstein, Kevin Danser and Khalil Wilkes were tagged for one penalty each in 2012; the offensive line accounted for zero in 2013. Obviously, that makes a big difference.