Bringing Statistics to Football

New York Times Link

Sabermetrics for Football

"...Belichick is known for his unorthodox strategies: being more willing than most to not punt on fourth down; running the ball far more than average in certain crucial situations; and eschewing two-point-conversion attempts in situations when orthodox doctrine recommends them.

Not coincidentally, experts in the world of football statistical analysis endorse all these strategies. For example, David Romer, an economist at the University of California at Berkeley, published a working paper arguing that conventional football wisdom led to far too much punting. Romer analyzed thousands of plays and calculated the chance of scoring from any position on the field. Based on that, he gauged the relative worth of the field position gained by punting against the lost opportunity to score. Romer found that football coaches punt far more than they ought to -- perhaps acting out of fear of the worst outcome (going for it on fourth down and failing), rather than rationally balancing risk and reward.

Romer's paper, ''It's Fourth Down and What Does the Bellman Equation Say? A Dynamic Programming Analysis of Football Strategy,'' is far from light reading, so it came as a shock to Romer when he learned that Belichick, who was an economics major at Wesleyan University, had read it.

Yet this is not the only example of the Patriots' willingness to turn to the academic world for guidance. A few years ago, Ernie Adams, the Patriots' football research director, asked Harold Sackrowitz, a Rutgers statistician, to give his opinion regarding the team's two-point conversion strategies. Sackrowitz concluded that the strategy was less than optimal -- and the Patriots subsequently did not try any two-point conversions in the 2003-2004 season.

Belichick also sticks to running the ball on third down in short-yardage situations, while other coaches pass reasonably often to try to surprise the defense. According to Aaron Schatz, founder of the sabermetrics-inspired Web site FootballOutsiders.com, Belichick's strategy, though predictable, is well warranted by a careful analysis of the risks and rewards involved in using it."

It should come as no surprise that one of the winningest coaches--known for his attention to detail--is also one of the most sophisticated. As Bill Parcell's once said: no matter how hard you work or how many hours you put in, if your opponents are using computers and you're still on legal pads, you're going to lose.

However, there is much you can do short of having a PhD in statistics or having huge sums of investment. Even HS and youth league coaches have access to programs like Microsoft Excel. Familiarizing yourself with some of the "tools" such as the Solver plug-in and data analysis (regression analysis, correlation and covariance, etc) can be easily done and result in lots of very telling information. Football is a statistician's dream, but the time has come to move beyond simple averages and aid your already difficult task of decision making with better analysis. In the future I hope to maybe make some worksheets or at least post some new ideas.