Bollettino

Streaks

LI intermittently tries to see things from the view of the Bush supporter. This isn't from any impulse to fairness, but from the same novelistic curiosity that makes a man slow down to look at a car wreck.

Now, here's the problem: Any non-Bush supporter looks at the news from, say, about January of this year, and asks: how can anyone support the contention that: a, the war was justified, and b., that it is going well, when all the evidence seems to be against it.

We think we know where to look for an answer: not in evangelical Christianity. Not in Dick Cheney's brainwashing powers. But in sports. That's right. We need to look at Bush's belief that the war is chugging along splendidly not in the light of information we have about the war, but in the light of the metaphysical belief, in sports, in the streak.

There’s an interesting paper by Bruce Burns, at Michigan State, on “heuristics as belief and behavior.” It is Burns contention that:

a, the hot hand belief in basketball – that is, the belief that there is a dependency between a player’s shots such that it makes empirical sense to talk about a “hot hand”, or a winning streak, has been disproven; and that

b., it nevertheless might be an advantageous strategy for a basketball team to operate with the belief that there are hot hands.

Burns states his thesis from the outset:

“The aim of this paper is to make this point by presenting an analysis of a
behavior that can be shown mathematically to improve the outcome for a decision maker, and thus is adaptive. The fact that the behavior may be supported by a false belief is irrelevant to whether the behavior is adaptive or not, though the false belief may actually be beneficial to the extent to which it helps to maintain the adaptive behavior.”

Burns makes a distinction between normative and adaptive views of heuristics. It is interesting, because the divide between them reflects almost exactly the divide between those who can’t understand how anyone could see the war as going well, and those who think that those who think the war isn’t going well are making the war go badly. The two sides are divided by mutual exasperation. Interestingly, when Gilovich, Vallone, and Tversky showed, in 1985, that hot hands don’t exist as statistically valid entities, the hottest objections they received were from basketball players and coaches. There is an ethos around sports that would make this kind of statement not only unbelievable, but, in itself, a kind of bad luck. It is that sports ethos that Bush, an owner, after all, of a baseball team, and a cheerleader in his prep school days, reverts to in his stump speeches, which the journalists all cover for their religious aspects.

Burns thesis, it would seem, would bulster the pro-war case. He divides his paper into five sections:

“The aim of this paper is to demonstrate how the hot hand behavior is adaptive, to
examine what general implications this analysis has for understanding people's reactions to streaks, and more broadly what implications this has for different approaches to decision making. Therefore the remainder of the paper is divided into five sections. The first section explains the adaptive approach to decision making and what it means to say that the hot hand behavior is adaptive. The second section supports the claim that the behavior is adaptive by showing this must be true if Gilovich et al's (1985) data are accurate. Most importantly, this is done by developing a Markov model of basketball shooting. In the third section this model is generalized to sequences of choices so as to determine the conditions under which following streaks should be adaptive. From these are generated empirical predictions regarding the conditions under which people will be likely to follow to streaks, and some empirical evidence regarding these is described. The implication of this analysis is that a belief in streaks may arise because following
streaks is adaptive, so in the fourth section is reported a study of people's attitudes regarding the hot hand belief in basketball and the hot hand behavior. This study examined the connection between the belief and the behavior as a function of basketball experience. The final section discusses how treating the hot hand as a belief or a behavior highlights the importance of a critical difference between the different approach to decisions making: that what is normative is not necessarily what people should do.”

The last sentence, of course, is the shocker. It is absolutely shocking to the liberal sensibility, which is built on precisely the opposite idea. We will explore that a little bit in another post.