Extra Points
News and commentary from around the Web

FO Mailbag: Rex Grossman

Mactbone: My impression is that Rex Grossman performs well when the team is up and plays poorly when the Bears are losing. I'm curious if his DVOA/DPAR is higher when the Bears are already up by 7/14 and lower when they're down 7/14 points. Of course, he could just play well when he has a short field and poorly when he actually has to put a drive together. Just trying to get a handle on the reason the Bears will win or lose in the playoffs.

Aaron Schatz: Thanks for the excellent question, Mactbone. It turns out that you are absolutely correct. When I measure scoring gap, I generally use 8 points as the guideline, since that's the maximum score you can get on just one drive. Here are Grossman's stats when the Bears are up by 9+ points, losing by 9+ points, or within 8 points either way.

Lose >8: -17.0 DPAR, -70.7% DVOA, 4.3 ypa, 4 INT
Within 8: 1.8 DPAR, -10.6% DVOA, 5.9 ypa, 7 INT
Win >8: 39.3 DPAR, 82.2% DVOA, 9.4 ypa, 0 INT
Wow, that would concern me if I was a Bears fan. Evil Rex seems to come out whenever the Bears are losing, and they're more likely to fall behind a good team they'll meet in the playoffs than they are to fall behind some team like Detroit.

How does this compare to other quarterbacks? I ran two other random guys with somewhat similar total value this year:

Here's Tom Brady:

Lose >8: 10.2 DPAR, 22.4% DVOA, 5.5 ypa, 0 INT
Within 8: 15.4 DPAR, 5.0% DVOA, 6.5 ypa, 8 INT
Win >8: 18.5 DPAR, 36.3% DVOA, 6.1 ypa, 1 INT
Brady seems to play looser in close games -- more yards, more turnovers -- but better overall when he either is coming back from behind or has a nice lead. Here's Steve McNair:

Lose >8: -0.2 DPAR, -14.0% DVOA, 6.5 ypa, 3 INT
Within 8: 11.8 DPAR, 2.0% DVOA, 5.7 ypa, 5 INT
Win >8: 5.0 DPAR, 2.8% DVOA, 5.0 ypa, 1 INT

I have no idea if this is actually a real trend that represents the Chicago offensive style, or Grossman himself, or if it is just random chance. One of the projects I hope to do for next year's book is a study of which "splits" represent actual skills and which are basically just random from year to year.


There are no comments yet.