Smart Football has moved!

Please check out the new site, All future updates will be made there.

Tuesday, May 26, 2009

Assorted links and notes - May 26, 2009

1. I guess the lesson is "Never turn down the President": Steelers linebacker James Harrison, in capping one of the unlikelier weeks in recent memory, had to put down his pet pitbull after it attacked his son.

2. NFL outlaws the "wedge formation" for kickoffs. It will be interesting how teams respond:

On a whiteboard in Westhoff’s Jets office, the sketched-out kick return looks markedly different now. In one version, the wedge has essentially been cleaved in half, two men separated from another pair by a couple of yards, heading upfield in lockstep. He also has a version in which two players are together with a lone blocker running alongside, until all three converge on the player who once would have been the wedge buster.

"You're going to see more man schemes," Toub said. "Everybody will have a man. There won't be any more zone blocking with the wedge."

Yet, the story is still sort of complicated:

Don't be surprised if some clubs try taking advantage of a loophole in how the banning of the wedge will be enforced. April recently spoke with NFL director of officiating Mike Pereira, who explained that officials will watch how the blocking forms at the time the return man fields the ball and determine whether there is a violation at that point. A flag won't be thrown if only two blockers are within two yards of each other, on the same plane, in front of the returner. And it wouldn't be a violation if two additional blockers were positioned in front of the other set of two, turning the wedge into sort of a box, when the ball is caught. That is a conclusion that special teams coaches have drawn from watching a DVD that league officiating crews have been showing to coaches as examples of what will and won't be penalized.

4. Facebook for football?

5. The New York Giants' David Diehl does a spot on impression of Brett Favre:

6. Phil Birnbaum points out that sometimes coefficients need not be "statistically significant" to be, well, significant.


Jeremy said...

That article by Phil Birnbaum is just ridiculously wrong. It's junk science at his best. His hypothesis is probably wrong: triples do not have a large impact on runs scores. But this doesn't mean that triples are worthless or less than a double (since doubles are statistically significant in increasing runs scored). Rather, it has to do with the small number of triples hit. A team that hits 20 triples versus 10 can only score 10 more runs (due to the triples), but this is not statistically significant over 162 games. However, we can modify the hypothesis and retest. For example, we can say, all things being equal, a triple leads to a run more than a double. You cannot use negative results as proof of a positive result. Otherwise, no one is ever wrong and science is a worthless endeavor.

Anonymous said...


Stan said...

I think Birnbaum's point should be understood in the context of fumble recoveries in football. Smart football coaches know that there is a proper way to try to recover a football. If you teach it and practice it, your players will get better at holding on to a fumble.

Yet, stats geeks "know" that fumble recoveries are totally random. The problem is that legitimate recovery opportunities are rare for any particular player. And the way the ball bounces has more to do with the recovery than the technique.

Statistically, the sample size is too small.