Writers of Pro Football Prospectus 2008

Most Recent FO Features

GurleyTod16.jpg

» OFI: SEC Surprises

In an opening week where even the elite teams in college football looked mortal, the SEC had two big surprises in Texas A&M and Georgia defeating their South Carolinian opponents by big scores.

24 Jul 2013

2012 Pressure Plays, Defense

by Danny Tuccitto

Can you believe it? With the Cowboys and Dolphins having reported for training camp duty this past weekend, the long NFL offseason -- and your resulting affective disorder -- is officially over. But before we give our undivided attention to the 2013 campaign, there's one final piece of 2012 charting data that we would like to detail for you: performance on passing plays with pressure. These stats (and thousands more) are currently available in Football Outsiders Almanac 2013, which is now on sale via both digits and dead trees. And if you're the type of person who likes venturing into the belly of an Excel beast yourself, don't forget that our 2012 charting data is also for sale.

With several players making spirited runs at Michael Strahan's single-season sack record and Arizona's offensive line trying harder than Brett Favre to make it happen, one result last season that may surprise some people was that the leaguewide sack rate actually fell from 6.4 percent to 6.2 percent. What's more, it seems to be part of a downward trend that started around 1997. Of course, for all the skill that goes into a sack (or lack thereof when it comes to quarterbacks and offensive linemen), it also requires a lot of quasi-random things going right (or wrong). Aldon Smith could have bull rushed Jonathan Martin all the way to Mexico City in Week 14, but if Ryan Tannehill's first read was open, we would have never gotten to see this GIF.

On the other hand, pressure, as measured by our game charting, captures more of the skill part of the pass-rushing equation. Though less glamorous (and less GIF-worthy), hurrying a throw, hitting a quarterback during his throwing motion, or forcing a scramble often produces as much defensive success as a sack -- and sometimes more. In a typical season, there's also nearly twice as many plays with hits, hurries, or forced scrambles as there are plays with sacks, so we get the benefit of inferring skill from a larger sample size. For instance, in 2012, there were 1,169 sacks; but there were 4,003 plays with pressure if we add in hits, hurries, and forced scrambles.

Below is a table showing defensive stats for plays with pressure last season. The first three columns tell you about the defenses ability to get pressure (i.e., sacks, hits, hurries, and forced scrambles). The next three columns tell you their performance on plays with pressure. (Remember, negative DVOA means above average efficiency.) The next three columns tell you their performance on plays without pressure, while the final three columns tell you how much that pass defense's performance depended on pressuring (or sacking) the quarterback.

Defense Plays Pct
Pressure
Rk with Pass Pressure without Pass Pressure Difference
Yds DVOA Rk Yds DVOA Rk Yds DVOA Rk
DEN 630 25.7% 1 0.8 -123.8% 5 6.9 26.9% 15 -6.1 -150.7% 4
PHI 545 25.3% 2 3.9 -54.7% 32 7.9 51.1% 32 -4.0 -105.7% 26
STL 635 22.4% 3 2.7 -87.8% 16 7.1 20.0% 8 -4.4 -107.8% 24
SD 633 22.3% 4 2.5 -71.3% 26 7.3 35.0% 19 -4.8 -106.3% 25
CIN 636 22.0% 5 1.6 -93.6% 14 6.9 22.1% 10 -5.3 -115.7% 17
NYJ 542 22.0% 6 3.1 -56.4% 29 6.6 13.4% 5 -3.5 -69.9% 32
ATL 596 21.8% 7 4.0 -80.0% 21 7.5 21.2% 9 -3.6 -101.2% 29
NYG 589 21.6% 8 2.0 -100.3% 10 8.7 36.9% 20 -6.6 -137.2% 8
DET 616 21.4% 9 2.2 -64.1% 28 7.9 39.7% 25 -5.7 -103.8% 28
MIN 700 21.4% 9 1.6 -92.2% 15 7.5 45.2% 29 -6.0 -137.3% 7
ARI 568 21.3% 11 1.4 -129.2% 2 7.3 10.3% 3 -5.9 -139.5% 6
HOU 654 21.1% 12 1.8 -106.0% 8 7.1 15.4% 6 -5.3 -121.4% 13
CHI 658 21.0% 13 1.1 -130.5% 1 6.6 -1.6% 1 -5.4 -128.9% 9
DAL 561 20.9% 14 2.3 -85.0% 17 8.0 43.5% 27 -5.6 -128.5% 10
SEA 631 20.8% 15 1.8 -109.9% 6 6.6 7.7% 2 4.8 -117.6% 14
WAS 695 20.7% 16 4.0 -78.1% 22 7.5 26.5% 13 -3.4 -104.6% 27
Defense Plays Pct
Pressure
Rk with Pass Pressure without Pass Pressure Difference
Yds DVOA Rk Yds DVOA Rk Yds DVOA Rk
BAL 620 20.5% 17 2.1 -82.7% 19 7.4 25.9% 12 -5.3% -108.5% 22
MIA 680 20.3% 18 1.6 -83.3% 18 7.7 34.2% 18 -6.1% -117.5% 15
SF 628 20.2% 19 0.7 -94.9% 12 6.7 13.2% 4 -5.9 -108.1% 23
OAK 572 19.8% 20 2.2 -75.2% 23 8.2 51.0% 31 -6.0 -126.1% 11
TEN 634 19.4% 21 2.5 -82.3% 20 7.6 33.2% 16 5.2 -115.5% 18
PIT 584 18.8% 22 1.4 -94.8% 13 6.6 26.8% 14 -5.1 -121.6% 12
BUF 602 18.8% 23 0.6 -109.6% 7 7.8 45.6% 30 -7.2 -155.2% 2
NE 659 18.7% 24 2.4 -101.6% 9 8.2 38.5% 24 -5.8 -140.1% 5
GB 642 18.5% 25 1.3 -96.6% 11 7.0 16.8% 7 -5.8 -113.4% 19
JAC 591 18.3% 26 3.9 -66.6% 27 8.0 44.6% 28 -4.0 -111.1% 20
NO 652 17.5% 27 4.0 -56.0% 30 8.3 37.2% 22 -4.3 -93.2% 30
CAR 628 17.4% 28 1.6 -125.0% 3 7.0 25.7% 11 -5.5 -150.8% 3
TB 681 17.2% 29 3.5 -55.7% 31 8.2 34.2% 17 -4.7% -89.9% 31
IND 595 17.1% 30 2.7 -72.5% 25 7.7 37.2% 21 -5.0 -109.7% 21
CLE 660 16.7% 31 1.2 -124.4% 4 7.4 37.9% 23 -6.2 -162.3% 1
KC 517 15.5% 32 4.1 -74.1% 24 7.8 43.1% 26 -3.8 -117.1% 16
NFL 619.8 20.2% -- 2.2 -89.7% -- 7.5 29.9% -- -5.2 -119.6% --

Compared to 2011, last season produced a pressure drop of 3.0 percentage points leaguewide, which mimics the aforementioned dip in sack rate. However, defenses were actually slightly more efficient on both plays with pressure (-80.0% DVOA in 2011) and plays without it (30.9% DVOA).

In light of recent news about Von Miller potentially being suspended for four games, Denver's stats might be cause for concern. No one pressured opposing quarterbacks more often than did the Broncos, and they were one of the most dependent on pressure for their overall pass defense efficiency. As Andy Benoit detailed in Denver's FOA 2013 chapter, the 2012 Broncos featured the second-most efficient third down defense since 1991 (-47.8% DVOA). Against third down passes, their -66.8% DVOA was the best since 2001. Both of these stats suggest last year's fifth-ranked unit is due for some regression to the mean in 2013. No doubt much of last year's success on third down was due to Denver's pressure stats, so the possible loss of Miller and the actual loss of Elvis Dumervil may prove to be a handy explanation for whatever regression does occur. That duo combined for 117.5 sacks, hits, and hurries last season, which accounted for nearly 70 percent of the front seven's total.

Like the Broncos, the Bears had one of the best pass defenses in the NFL last year (No. 1 DVOA, in fact). Unlike Denver, however, Chicago's DVOA splits were highly consistent from down to down (they ranked in the top four across the board). One of the main reasons why: They were the best pass defense last year both with pressure and without pressure. It's as if Chicago -- say, against Arizona in Week 16 -- could have subbed in a defensive line from the Uncanny Valley and they still would have played above-average pass defense.

On the other end of the spectrum is a pass defense like Cleveland's. The 2012 Browns suffered from being heavily reliant on pressure, but not getting it anywhere near enough. We assume this is why they gave Paul Kruger $20 million in guaranteed money, and hired defensive coordinator Ray Horton, who coaxed pressure out of mediocre pass rushing talent in Arizona. Of course, both of Horton's seasons with the Cardinals saw his pass defense rank similarly regardless of pressure. (In 2011, they ranked 15th in DVOA with pressure and 19th without pressure.) Kruger's former team had a similar profile.

Speaking of the Ravens, we'll close by noting that 2012 saw a departure from the past couple of seasons insofar as the Super Bowl champion didn't rely as heavily on pressure to make their pass defense work. In 2010, Green Bay ranked seventh in DVOA with pressure. In 2011, the New York Giants ranked second in DVOA with pressure and fourth in DVOA difference.

That's all for now. Just as we did last year, part two of this series will take a look at how pressure affected performance from the quarterback's perspective.

Posted by: Danny Tuccitto on 24 Jul 2013

11 comments, Last at 30 Jul 2013, 4:46pm by jchavlik

Comments

1
by Karl Cuba :: Wed, 07/24/2013 - 8:06pm

OK, you know I love me some FO and am a big fan of Florida Danny but how can you fail to point out that most of the league's best pass defenses (Chicago, Seattle, Houston, San Francisco, NY Jets and Arizona) all rank as the best when not getting pressure?

This seems to be the most prominent result, it suggests that coverage rules OK.

3
by Danny Tuccitto :: Thu, 07/25/2013 - 1:00am

Because I was waiting for you to point it out!

In all seriousness, though, it's actually the expected result. If defenses(on average) only get pressure on 20% of their pass plays, then that means -- all else equal -- their pass defense DVOA is going to be weighted 4-to-1 towards efficiency on non-pressure plays. It's like saying, "Most of the best teams rank as the best when playing on Sunday." Yeah, that's because the vast majority of their plays come on Sundays.

p.s. Thanks for the kind words.

2
by BigWoody (not verified) :: Wed, 07/24/2013 - 10:03pm

Wow! I guess I wasn't paying attention to Philly last year. Their pass defense was bottom/bottom. I expected the Saints for that honor. Or am I reading this wrong?

4
by CBPodge :: Thu, 07/25/2013 - 4:16am

Philly's pass D was 24%, NO's was 20%. But hey, at least they were good at getting pressure on the QB!

5
by Joseph :: Thu, 07/25/2013 - 1:07pm

Actually, what that means to me (as a Saints fan) is that Philly couldn't cover anybody, or tackle anybody after the catch. They could pressure the QB, but then he still completed the pass. [or am I misunderstanding the table also?] No bones about it, my Saints couldn't defend the pass against nearly anybody.

6
by Led :: Thu, 07/25/2013 - 1:37pm

Plays with pressure include sacks, correct? If so, it would seem that a team's DVOA when getting pressure is heavily dependent on how often that pressure results in a sack.* Getting a lot of pressure is certainly better than not, but if you don't turn that pressure into many sacks (like the Eagles and Jets) then your DVOA in pressure situations is much worse than average.

* Except for the Bears, who had the best DVOA when getting pressure despite only a moderate number of sacks.

7
by Danny Tuccitto :: Thu, 07/25/2013 - 1:44pm

Very true. Talked about this a little bit in last year's piece.

8
by theslothook :: Thu, 07/25/2013 - 2:19pm

I think the next frontier in determining pass defense has to be which unit contributes most to it. Ie- is it corners, safeties, or the linebackers and in what proportion. And what are the intermingling effects.

9
by Danny Tuccitto :: Thu, 07/25/2013 - 3:38pm

Maybe this suggests a general baseline of 20% pass rushers and 80% coverage, with that 80% split out according to whatever our charting says about the average ratio of linebackers, safeties, and cornerbacks in primary coverage on non-pressure plays?

10
by theslothook :: Thu, 07/25/2013 - 4:03pm

That would actually be an interesting thing, I agree. But we'd have to start charting things like qbs first read and if he were covered or not. Maybe add that as a charting column for next year.

11
by jchavlik :: Tue, 07/30/2013 - 4:46pm

I was about to inquire as to whether base cover 2 teams such as the Vikings would be at a disadvantage (to look worse then they actually are) due to the "keep everything in front of you", "bend but don't break" approach until I realized that Chicago still runs a cover 2 as well. The Vikings badly need secondary help and have badly needed it for years. Is it safe to say that quality corners are more valuable than quality pass rushers in the modern NFL?