Thomas Schelling and Gameplanning vs. Playcalling

Thomas Schelling, along with Robert Aumann, was honored this week with the nobel prize in economics for his work in Game Theory--or "interactive decision theory". An amazing mind, his books are also extremely readable (i.e. unlike much game theory, they do not overwhelm you with the mathematics).

Since he won I browsed around the web a bit, and, thanks to Marginal Revolution, here is a lecture he gave on the theory of self-restraint.

I suggest reading the essay, as it touches on a variety of interesting and at times troubling issues of self-restraint and individual and public choice. It addresses the boundaries of human capacity and consequences of our abilities and limitations when making decisions.

Schelling begins his lecture:

A few years ago I saw again, after nearly fifty years, the original Moby Dick, an early talkie in black and white. Ahab, in a bunk below deck after his leg is severed by the whale, watches the ship’s blacksmith approach with a red-hot iron which, only slightly cooled by momentary immersion in a bucket of water, is to cauterize his stump. As three seamen hold him he pleads not to be burnt, begging in horror as the blacksmith throws back the blanket. And as the iron touches his body he spews out the apple that he has been chewing, in the most awful scream that at age twelve I had ever heard. Nobody doubts that the sailors who held him did what they had to do, and the blacksmith too. When the story resumes there is no sign that he regrets having been cauterized or bears any grievance toward the men who, rather than defend him against the hot iron, held him at the blacksmith’s mercy. They were not protecting him from an involuntary reflex. And he was not unaware of the medical consequences of an uncauterized wound. Until the iron touched him he knew exactly what was afoot. It was a moment of truth. He was unmistakably all there. He made his petition in clear and understandable language. They had neither personal interest nor legal obligation to subject him to torture. And they disregarded his plea. When the iron struck he went out of his mind, still able, though, to communicate with perfect fidelity that all he wanted was the pain to stop. While the iron was burning his body we might declare him to have been not fully present, but until that instant it is hard to claim that he didn’t understand better than we do what the stakes were.

Ahab and his wound dramatize a phenomenon that, usually not so terrifying, all of us have observed in others and most have observed in ourselves. It is behaving as if two selves were alternately in command. A familiar example is someone who cannot get up when the alarm goes off. [He also mentions examples like how we do not keep candy or alcohol in the house because of expectation of our own future weakness. A poignant example he uses is someone who attempts suicide.]

I don't want to misconstrue Schelling's lecture--it gets much further afield than anything here--but while reading it I was struck with the classic two-self dichotomy that every offensive coordinator in football must deal with: the gameplanner and the playcaller. (We could even probably say the 1st quarter coach vs. the 4th quarter coach, but that is a different discussion.)

Posed with similar circumstances and problems, even the exact same level of information, the in-game you and the weekday you would likely give very different answers. Bill Walsh has often talked about the advantage of scripting plays and playcalling in detached, relaxed circumstances where logic can dictate vs. the insanity of a game situation.

One of the implications is that we must take turns as Ahab and as the crew, holding each other down to see us through what our immediate self vocally rejects. Even going so far as to override us when we are at our weakest. While not as dramatic as Moby Dick or even some of Schelling's other scenarios, it is an interesting view.

Further, a gameplan can be seen as a form of contract with our football team and the other coaches that guarantees that we will stick to what has been decided collectively and with the most information possible. While it is a blueprint for attacking your opponent it is also there to guard against you suddenly becoming Hal Mumme and throwing 50 times when, as a staff, you'd decided to be a lot more like Woody Hayes that game.

The obvious caveat is that a gameplan is contingency based, much of it depends on what your opponent does. But, a well crafted gameplan can still handle these scenarios and be created in a detached setting. In Schelling's language, created by your week-day self.

Notes on Running With the Football

From a press Conference with Notre Dame Coach Charlie Weis:

[In response to a question on running back Darius Walker's running style]

But I can tell you what I used to tell [Patriot's receiver] Deion Branch after I had a big research study on (Marvin) Harrison a few years ago when Deion was a second year [in New England]. I noticed that Marvin, with all of his production, any time the hits were coming, he was going down. So after I thought about it for a while. I thought, this isn't the stupidest thing in the whole world to have your best guy, when he's about ready to get crunched, go ahead and make sure that doesn't happen.... there are times to take the hit and there are times not to take the hit.

The first important point is explicitly made: If you're talking about your best guy, there is logic to letting him go down or go out and bounds and avoid the big crunching hits. While we don't want to coach pansies, you also want to have the kid for the whole season. Injuries are a bigger threat to receivers than say offensive linemen because it is more difficult for them to play through injuries due to the nature of the position.

Second, which can be gleaned since his example was Marvin Harrison, most all production, at least a receiver's, is done before contact is made. This does not limit yards after the catch, but instead tells you that the focus is on running away from defenders rather than at them, either to run them over like Earl Campbell or try to individually juke every guy out, which is idiotic.

Instead, great receivers catch the ball, get upfield immediately, and try to score by splitting defenders. What do I mean by splitting defenders? Quite simply: run inbetween them. There are circumstances when you need to take the hit to the defender (sometimes on a slant all you can do is deliver the "forearm of doom" to the safety after the catch) but, usually, you score by running directly upfield right inbetween the corner and the safety for the long TD.

It's an often missed point. If you watch Sportscenter you will see that almost every short pass that goes for a TD involves a receiver bursting through a seam rather than trying to juke guy X, run through guy Y, spin off guy Z, and then finish by jumping over guy A. Notice I ran out of letters because doing this, even if successful, takes so long that the whole defense has time to show up. It happens occaisionally, but don't make it a habit.