Writers of Pro Football Prospectus 2008

16 Aug 2012

Blitzing Drew Brees Not a Bad Idea

NFL.com's Gregg Rosenthal goes over his takeaways from FOA 2012.

Posted by: Rivers McCown on 16 Aug 2012

10 comments, Last at 21 Aug 2012, 9:13pm by chemical burn


by Marko :: Thu, 08/16/2012 - 3:00pm

"2. Cam Newton's had a sensational year passing, which often shows up in Carolina's running stats. The team averaged 6.3 yards-per-carry on shotgun handoffs, partly because defenses were so worried about Newton."

So the presence of Cam Newton inflates the stats of Carolina's RBs and makes them seem better than they are. The Panthers therefore shouldn't sign their RBs to lucrative contract extensions. Wait, what?

by milo :: Thu, 08/16/2012 - 5:03pm

Rosenthal: Blitzing Drew Brees worked in 2011. This goes against conventional wisdom, but it's also undeniably true. According to the Football Outsiders' Almanac, Brees produced an NFL-high 8.3 yards-per-pass play when opponents rushed three or four defenders last year. The average dropped to 7.2 yards (still good) against five defenders, and down to 6.8 yards against a big blitz of six defenders or more.

This is not what we expected. It is not what opposing defensive coordinators expected either, or they would have blitzed more. No quarterback saw fewer "big blitzes" in the NFL than Brees. This is not a new trend either; Brees has been better against four rushers than a blitz for three years running.

(Aaron Rodgers, on the other hand, was more effective last year the more pass rushers you sent at him.)

On August 1 Danny Tuccitto wrote an FO article showing Brees to be the best QB by DVOA when "pressured".

I'm interested in how the two stats have interacted in this way. The first stat measures what the defense sent, the second measures whether the defense achieved the goal of "pressuring" the QB which could explain most everything.

by tuluse :: Thu, 08/16/2012 - 5:08pm

It was still the 2nd largest difference.

by milo :: Thu, 08/16/2012 - 5:21pm

I read it as the second smallest difference.

by tuluse :: Thu, 08/16/2012 - 5:27pm

Do'h! You're right.

by chemical burn :: Fri, 08/17/2012 - 1:42pm

Well, the whole thing is sorta funny in that the analysis linked to here as the extra point doesn't seem to have noticed that literally every QB in the league is worse when pressured. What the FO article you linked to doesn't show is whether Brees got some of the least pressure in the league because teams were avoiding the risk of getting burned by him or whether he's so good at recognizing a blitz and getting the ball away before pressure even arrives that it doesn't even register as a pressure(or whether his excellent line was causing blitzes to get stone-walled.) That he has the 2nd smallest difference seems to suggest the analysis forming the extra point is wrong. Blitzing him has the least effect on him as it does as any effective QB in the league. Also, it's likely a lot of blitzes aren't even getting pressure. I'd say: don't blitz Brees. Or do. He's going to be one of the best at what he does either way. I suspect if you looked Brees numbers by the number of blitzers and not the percentage of pressure, you'd see that bringing blitzers will cause him to tear you up. Does FOA have those numbers? QB DVOA according to number of blitzers?

by chemical burn :: Fri, 08/17/2012 - 1:47pm

Looking at the FO article, it's not even about blitzing. These guys could have been pressured when only 3 line-men rushed the passer and everyone else dropped into coverage. Theese stats might even indicate that you don't want to give Brees the time to scan the field and let something develop because even if you get pressure eventually, he can still be effective.

Anyway, the linked article being about pure traditional stats doesn't tell us that much - if you bring the blitzers on 3rd and 6 and Brees averages "only" 6.8 yards on those plays, that's still very effective QBing...

by MJK :: Mon, 08/20/2012 - 4:40pm

Right. Breaking it down into a flowchart would look like this:

A. Defense sends some number of rushers --> B. Those rushers have some amount of success getting pressure --> C. The QB responds to the pressure or lack thereof with some amount of success.

What this XP article is driving at is what the overall correlation between A and C are, whereas the FO article was looking specifically at the linkage between B and C.

The FO article indicated that Brees is better than average at doing something (C) when B succeeds... and yet the XP stat implies that Brees is worse than average when the defense sends more rushers. That could be because Brees was worse than average at spotting the blitz but better than average at reacting to it when it happened, or because the Saints offense was bad at picking up blitzes but Brees still could succeed when they didn't, at least better than average QB's.

What the XP article should really say is not "Blitzing Drew Brees worked". It should say "Blitzing the Saints offense when Brees was QB worked".

by chemical burn :: Tue, 08/21/2012 - 9:13pm

I don't think you can even go that far because he uses traditional stats like 6.8 yards per pass to prove his point. If Brees were blitzed 90% of the time on 3rd and 4 and throwing for 6.8 yards per pass on average, the blitz might be causing his average yards per pass to drop, but it sure as shit isn't effective in any meaningful sense of the word...

by Danny Tuccitto :: Fri, 08/17/2012 - 11:45pm

I think chemical burn fleshes it out well here (puns!).