Writers of Pro Football Prospectus 2008

18 Jan 2012

Great Defense Makes a Long Awaited Return

There are a lot of places in this article where I take a Football Outsiders concept -- like the fact that the 49ers defense ranks first in Second Level Yards -- and explain it in raw data form. And of course, when talking about the Patriots defense I bring up the old "field position" discussion. But it is still pretty interesting, even though the field position differential for the 49ers is even more interesting.

Posted by: Mike Tanier on 18 Jan 2012

11 comments, Last at 22 Jan 2012, 3:06am by Intropy

Comments

1
by Mr. Guest to you (not verified) :: Wed, 01/18/2012 - 10:11pm

Surprising opinion from FO staff. So what is it, are turnovers mostly luck or, instead, are they skilled plays?

HOU: 4 TO
NO: 5 TO
GB: 4+ TO

That's why those teams lost. HOU defense was better, and neither GB or NO were actually stopped.

7
by turbohappy (not verified) :: Thu, 01/19/2012 - 12:39pm

The FO research has never shown that turnovers are mostly luck (AFAIK), just that fumble recovery is random. Causing fumbles and interceptions are non-random.

8
by JIPanick :: Thu, 01/19/2012 - 9:06pm

Non-random is a tricky word. It's more instructive, I think, to think of interceptions and forced fumbles as weighted random events, and fumble recoveries as unweighted random events (or weighted only be situation rather than team).

Let's consider a hypothetical QB who has a "true" 2% interception rate, constant across a 15 year career, and throws exactly 500 passes a year.

So, taking random.org, running 15 sets of 500 rolls of a d50, and counting the 1s as picks, we get:

14 picks
10 picks
5 picks
15 picks
11 picks
6 picks
10 picks
13 picks
11 picks
24 picks <- Hello, 2011 Philip Rivers!
15 picks
9 picks
9 picks
18 picks
8 picks

That's a lot of year to year variation. It's possible that this isn't a very accurate model (if he throws 250 passes in 3.5% situations and 250 in 0.5% situations, for example, it'd drive the number of picks closer to exactly 10 every year), but I actually think it's pretty solid.

It should also be as accurate or more when applied to defensive rather than offensive interception rates. I certainly think that teams with extremely good/bad TO +- should be though of as having been both good/bad and lucky/unlucky.

I have no idea what FO's research actually says about this.

9
by big_jgke :: Fri, 01/20/2012 - 1:00pm

Crikey! Nerd Alert!

10
by Andrew Potter :: Fri, 01/20/2012 - 2:15pm
2
by Subrata Sircar :: Wed, 01/18/2012 - 10:22pm

"...no team has been able to stop Tom Brady from cruising down the field with his precision short passing attack."

The Steelers haven't vanished from anyone else's memory, right? The Patriots didn't turn the ball over and got one interception themselves, but Brady throwing for <200 yards doesn't sound like an unstoppable juggernaut.

The Patriots had a single 3-and-out in Q1, while the Steelers moved the ball at will. Their next drive crossed midfield but then got thrown back by sacks. They then got the ball at the Steeler 8 and scored a TD off the interception. Next drive was a 2-minute drill for a field goal with 30 seconds left in the half. 2nd half was a 3-and-out, 50 yards and missed FG, 14 yards and punt, and then the 67-yd TD drive to get them to within 6.

8 drives, 4 punts; 1 that started on the opponent's 8. [The Steelers also held the ball for almost twice as long as NE and rolled up 400+ yards of offense, though. It's not clear the Ravens can do that - Ray Rice is much better than Mendenhall, and the Ravens line is better, but Flacco is no Big Ben.]

All of which is to say that the Ravens can beat the Patriots if they, too, can rush the passer while covering all the short zones so Brady can't hit his hot routes for YAC and doesn't have time for a deep route.

4
by Whatev :: Thu, 01/19/2012 - 2:50am

Well, yes, but if you can rush the passer while covering all the short zones so that the opposing quarterback can't hit his hot routes for YAC and doesn't have time for a deep route, you can pretty much beat ANYBODY.

5
by Intropy :: Thu, 01/19/2012 - 4:11am

I'm running a T-formation offense with Jim Brown, Jim Thorpe, and Doak Walker. Cliff Battles is my handoff specialist, and Whizzer White is my lawyer.

6
by dryheat :: Thu, 01/19/2012 - 9:58am

Alex Karras in charge of the stables?

11
by Intropy :: Sun, 01/22/2012 - 3:06am

Fitzchivalry Farseer

3
by artmaccuinn (not verified) :: Thu, 01/19/2012 - 2:44am

why am I not surprised that a discussion of great defense involves a Rex Grossman Factor? actually, RGF inclusion in any discussion of great defense is probably mandated by law.