07-03-2013, 06:34 PM
One random comment on that article: The author uses the stat "recovered N of M fumbles in the game" two or three times. This is a pretty goofy stat if I understand what he means correctly. It makes a big difference how many of those M fumbles were by your own team and how many were by your opponent.
For one thing, fumbling is bad in itself no matter who recovers. For another, I'm pretty sure that on average it's easier (less of an accomplishment) to recover a fumble by your own team than to recover an opponent's fumble -- for example, think about fumbled snaps, fumbled handoffs, fumbled kickoffs. In all those cases the defense is usually not close enough to the ball to have a good chance to recover it.
As an example, if your team fumbled 5 times, you recovered 1, and the other team didn't fumble at all, that is a lousy performance. If the other team fumbled 5 times, you recovered 1, and you didn't fumble at all, that is a decent performance. But in both cases that article would say "recovered only 1 of 5 fumbles in the game."
For one thing, fumbling is bad in itself no matter who recovers. For another, I'm pretty sure that on average it's easier (less of an accomplishment) to recover a fumble by your own team than to recover an opponent's fumble -- for example, think about fumbled snaps, fumbled handoffs, fumbled kickoffs. In all those cases the defense is usually not close enough to the ball to have a good chance to recover it.
As an example, if your team fumbled 5 times, you recovered 1, and the other team didn't fumble at all, that is a lousy performance. If the other team fumbled 5 times, you recovered 1, and you didn't fumble at all, that is a decent performance. But in both cases that article would say "recovered only 1 of 5 fumbles in the game."