A long night in Los Angeles: Examining Stanford's loss to USC

It's remarkable how just one week changes things so much in college football.

And after Stanford's 20-17 loss to USC, the timbre has once again changed around the Cardinal.

It's too easy to attribute the loss to the random number generator that is Kevin Hogan coming up "blarrrghhhh," but this is obviously a factor, and probably the most important factor. But the loss is a wicked problem on various fronts: way overdetermined, with the missed field goal, sundry penalties, drops by Ty Montgomery, and the continuing curious road struggles of the offensive line.

Let's start with the simplest thing, though: close game prowess. As any stat person will tell you, over the long run it's something of an illusion. Good teams will win more often than not, but they don't dominate close games. That's why a focus for Stanford was -- or should have been -- in blowing more teams out this year and hence winning more comfortably. Stanford's 3-2 in one-score games this year after going 8-2 last year. Regression to the mean, they call it. But that's the macro look. What about the micro -- what happened in this game, right now?

Whereas previously on the road the offensive line struggled by giving up too many sacks, the offensive line did not do superbly in getting push on base running plays in conventional situations. Stanford running backs averaged 3.09 yards per carry for 34 yards on first down, a sign of how well the USC defensive front did in limiting the running game for Stanford.

Containing Stanford's run game so well on first down puts the coaches in a bind, and has ever since the 2009 Big Game, when the offensive brain trust decided unwisely to continue pushing it with Toby Gerhart and away from Andrew Luck. Contrary to popular memory, in which Harbaugh did A Very Bad Thing by giving the game over to his quarterback, the mistake of the staff was in not riding Luck in unconventional situations -- throwing more on 1st down, contrary to what the defense was anticipating.

The coaches learned that particular lesson in this game against USC, seeking out alternative plays on early downs. Hogan was 5 for 6 for 6.3 YPA when throwing on first down, hardly spectacular but better than his overall average, and certainly better than base running plays. Non-base running plays called on first down averaged 6.75 yards per carry.

So on some level the ability of playcallers to mix up their plays was successful. But the offense still runs into two basic constraints, one imposed and the other conceptual. The first is that, as you've noticed, Hogan is much too erratic -- at least in the way the current offense sets him up to play. Hogan was much less monomaniacal than in the past in seeking out downfield strikes, but the fact remains that Hogan, right now, is too erratic on the road to be depended upon to produce points reliably. While he incinerated the Washington State defense, that appears to be an outlier. The question for next year is: will Hogan become more reliable on the road, or does he need to be replaced by someone who is reliable?

The conceptual problem with the offense remains the same. It's not in the playcalling, as some too easily say. (And, please, stop it with the calls for Pep Hamilton. Watch a Colts game and identify the playcalling genius at work there. If you can find it, congratulations -- they should've put you on the Higgs Boson search team.) I hope I'm not too easily convinced by my prior opinions on the offense, but issues still lie with the personnel.

Today we saw Michael Rector targeted on actual, live intermediate routes. The man whose sole role seemed to be "run down the field really fast, see what happens" got the ball in his hands on actual intermediate routes -- and guess what? Fast people turn out to be effective all over the field. A strange concept, I'll grant you, but Rector's previous per-catch average of 40 is sinking, with a deplorable dink-and-dunk average of 22 this game. There's still low-hanging fruit there, should anyone bother to pluck it. (And don't forget the not-so-bad block on Tyler Gaffney's first TD.)

Rector's quality performance underscores the mystery of the offense, which is why certain players are designated for certain roles. Whatever the depth chart says, Jordan Pratt has gotten the lion's share of wideout snaps after Montgomery in the post-Cajuste world. Why? He had a couple solid catches, but Whitfield, Rector and Kelsey Young all offer far more in athleticism. What's up with Barry J. Sanders, who was used as a solid decoy against Oregon but, to my knowledge, did not enter the field of play as an offensive player against USC?

The offensive performance bears a decent relationship to the UCLA game, in which passages of good play were speckled with enough errors to ruin the whole thing. But it's time to ask: what's gone wrong when so many errors are allowed to dominate to such a great degree?

Altogether, the offense is in a weird place. But it doesn't have to be. And that's probably the most frustrating part.