. . . and more on football
A couple of weeks ago the New England Patriots' head coach, Bill Belichick, decided to "go for it" (i.e., attempt to get a first down) rather than punting on fourth and two at his own 26 yard line. He didn't make it, and lost the game. Last Saturday Yale, which is presumably a smart if not terribly talented football team, tried the same thing against Harvard. They failed, and also lost.
The Patriots and (to a lesser extent) Yale losses spawned a small cottage industry of statistical analyses, many or most of which concluded that the decision to go rather than kick was logical. Yet most football people will tel you the decisions were stupid . . . and the fact remains both teams lost. What's going on?
I think this is a good example of the limits of statistical analysis and the benefits, often derided, of plain old common sense. One problem, I suspect, is that it's harder than you might think to make a statistical analysis. For example, do you look at all cases in which previous teams went for it on fourth and two, or only those where they did so in the fourth quarter? Do you look at all possible contests, or only those involving the same two (or comparable) teams? These kinds of questions come up all the time in statistics, but rarely with so much on the line.
Another interesting question is what the decision to "go for it" communicates and how the decision itself affects the psychology of both sides. Both the Patriots and the Elis (Yale) had long losing streaks against the relevant opponents: in the Patriots' case, at least, the decision reflected a lack of confidence in the team's defense as much as confidence in its offense. Did the recklessness of the choices subtly or not so subtly convey a sense of desperation, which inspired their opponents to thwart the coaches' plans?
I am reminded here of William F. Buckley's famous assertion that he would rather be governed by people chosen at random from the Boston phone book than by the Harvard faculty. Buckley, of course, went to Yale.