Writers of Pro Football Prospectus 2008

Most Recent FO Features

DarthRaider.jpg

» The Week In Quotes: December 15, 2017

As the regular season enters its home stretch, teams prepare for the playoffs, and everyone's busy with holiday activities, it's only obvious that the top item on anyone's mind is Star Wars. (No spoilers.)

13 Jul 2016

Defense and Pass Pressure, 2015

by Carl Yedor

Esteemed defensive coach Buddy Ryan famously had a playbook that described quarterbacks as "overpaid," "overrated," and "pompous," adding that they "must be punished." While blunt, it certainly had one thing correct: the ability to pressure the quarterback is a critical part of a successful defense. On Monday, we took a look at how quarterbacks were able to handle defensive pressure; today, we're switching sides of the ball and examining which teams were best at making opposing quarterbacks' lives miserable. While sacks will always be a driving factor in the discussion of pressure, sometimes merely forcing a quarterback off his spot and into an awkward throw is all it takes for a defense to win on a given play.

In a slight increase from 2014, defenses were able to successfully generate pressure on pass plays 25.6 percent of the time, compared to 24.6 percent in 2014. Interestingly, the league average DVOA on pressured plays "jumped" from -76.0% to -67.3% -- still very much in the defense's favor, but there was clearly some offensive improvement. This continues a trend that has seen offenses perform better under pressure since 2010. The three-year average for the 2010 through 2012 seasons was -82.2%. By contrast, league average DVOA when there was no pressure in 2015 was 35.5%, which was very similar to the 34.4% mark from 2014.

The table below focuses on team defense statistics for the regular season. Pressure plays include sacks, hurries, and forced scrambles. Aborted snaps and coverage scrambles/sacks are not included, but defensive pass interference penalties are. A negative DVOA is good for the defense while a positive one is bad. The defenses are sorted by the difference in DVOA between pressure plays and non-pressure plays. The higher the ranking, the more that a team's defensive performance improved when getting pressure on the quarterback.

As noted also in the article on how quarterbacks performed under pressure, this year's pressure numbers are a combination of data from ESPN Stats & Info and Sports Information Solutions.


2015 Defenses With and Without Pressure
Defense Plays Pressure
Plays
Pct
Pressure
Rank DVOA
w/Pressure
Rank DVOA no
Pressure
Rank DVOA
Dif
Rank
SD 567 143 25.2% 16 -88.7% 5 52.8% 30 -141.4% 1
MIA 594 150 25.3% 15 -78.7% 9 61.6% 31 -140.3% 2
PHI 696 173 24.9% 19 -99.0% 3 39.2% 19 -138.3% 3
CAR 726 191 26.3% 12 -114.4% 1 20.6% 5 -135.0% 4
GB 630 167 26.5% 10 -93.4% 4 27.6% 10 -121.0% 5
WAS 628 142 22.6% 28 -76.4% 11 44.0% 26 -120.4% 6
DEN 655 214 32.7% 1 -104.1% 2 12.6% 1 -116.7% 7
TEN 567 143 25.2% 17 -68.4% 17 47.6% 29 -116.0% 8
DAL 564 139 24.6% 22 -75.8% 12 37.9% 18 -113.6% 9
NO 606 129 21.3% 32 -40.5% 27 72.8% 32 -113.2% 10
NYJ 674 189 28.0% 6 -82.7% 6 29.5% 12 -112.2% 11
TB 591 144 24.4% 23 -64.3% 20 47.0% 28 -111.3% 12
DET 602 150 24.9% 18 -65.5% 19 43.3% 22 -108.8% 13
KC 694 195 28.1% 5 -82.6% 7 21.6% 6 -104.2% 14
STL 674 191 28.3% 4 -76.9% 10 25.9% 9 -102.8% 15
PIT 694 168 24.2% 24 -69.8% 14 31.1% 14 -101.0% 16
Defense Plays Pressure
Plays
Pct
Pressure
Rank DVOA
w/Pressure
Rank DVOA no
Pressure
Rank DVOA
Dif
Rank
CHI 587 145 24.7% 20 -56.5% 22 44.1% 27 -100.6% 17
NE 677 167 24.7% 21 -67.6% 18 31.9% 15 -99.5% 18
ARI 638 187 29.3% 3 -79.7% 8 19.8% 4 -99.5% 19
CIN 713 190 26.6% 9 -73.1% 13 25.0% 8 -98.1% 20
BAL 623 149 23.9% 26 -53.7% 24 43.7% 25 -97.4% 21
MIN 633 177 28.0% 7 -63.3% 21 32.3% 16 -95.6% 22
HOU 619 167 27.0% 8 -68.6% 16 18.0% 3 -86.6% 23
SEA 614 180 29.3% 2 -68.8% 15 17.1% 2 -86.0% 24
IND 646 171 26.5% 11 -54.3% 23 29.0% 11 -83.3% 25
CLE 567 134 23.6% 27 -35.0% 29 43.7% 24 -78.7% 26
NYG 684 177 25.9% 13 -34.9% 30 43.5% 23 -78.4% 27
ATL 610 131 21.5% 31 -42.7% 25 35.1% 17 -77.7% 28
SF 607 156 25.7% 14 -29.4% 31 43.1% 20 -72.6% 29
JAC 664 144 21.7% 30 -27.4% 32 43.2% 21 -70.6% 30
BUF 650 142 21.8% 29 -36.4% 28 29.8% 13 -66.2% 31
OAK 715 172 24.1% 25 -41.4% 26 23.6% 7 -65.0% 32

The best defenses at generating pressure were pretty much exactly who you would expect. Denver lead the way with a 32.7 percent pressure rate, the only team in the league above 30 percent. Following them were Seattle, Arizona, St. Louis, and Kansas City to round out the top five. It couldn't have been fun to play quarterback in the West division of either conference, and that does not look likely to change in 2016.

New Orleans had lowest pressure rate at 21.3 percent, closely followed by Atlanta at 21.5 percent, Jacksonville at 21.7 percent, and Buffalo at 21.8 percent.

When pass pressure reached the quarterback, Carolina posted the best DVOA in the league at -114.4%, ahead of Denver's -104.1%. Carolina and Denver both registered as top-five defenses both with and without pressure, which certainly does a lot to explain how we ended up with our Super Bowl 50 matchup.

On the other side of the coin, Jacksonville, San Francisco, the Giants, Cleveland, and Buffalo were the worst five teams by DVOA when they actually got pressure on the opposing quarterback. Seeing Buffalo down there is not a great sign for the Ryan family legacy, as Rex Ryan has made pressure a cornerstone of his defenses. We'll see if adding his brother Rob to the mix will end up helping turn things around in upstate New York. (The Buffalo chapter of FOA 2016 will be skeptical, certainly.)

It's important for teams like Buffalo to make some significant improvements in generating pressure because even the best defensive teams will not be successful without being able to rush the quarterback. The Broncos ranked first in DVOA when not generating pressure, but their mark of 12.6%, if applied to an overall defense, would have been the second-worst in the league, only ahead of New Orleans -- the worst defensive team in DVOA history. Eight other teams -- Seattle, Houston, Arizona, Carolina, Kansas City, Oakland, Cincinnati, and St. Louis -- also posted DVOA marks without pressure that were better than New Orleans' overall defensive DVOA for the season. The good news in New Orleans is that there is nowhere to go but up. The bad news is that the second-worst Bears were closer to New England, ranked 12th overall in defensive DVOA, than they were to New Orleans. Oh, and who began the year as defensive coordinator for the 2015 Saints, you might ask? Rob Ryan.

Part of what made New Orleans so historically futile on the defensive side of the ball was their last-ranked defense without pass pressure. The Saints allowed a DVOA of 72.8% when they were unable to get pressure on the quarterback. Miami, San Diego, Tennessee, and Tampa Bay rounded out the bottom five in terms of their defensive DVOA without pass pressure. Fortunately for the four other teams in the bottom five, they were able to get pressure at closer to a league average rate, which limited the degree to which their defensive ineptitude without pressure came to light.

Every team suffers a drop in defensive performance when they fail to get pressure on the opposing quarterback, but San Diego had the largest drop-off from pressure to no pressure at -141.4%, with Miami (-140.3%), Philadelphia (-138.3%), and NFC Champion Carolina (-135.0%) close behind. In rescinding Josh Norman's franchise tag and subsequently letting him depart for Washington, Dave Gettleman and company made the decision to focus their defensive spending on their front seven. Having already set the market with Luke Kuechly's contract extension at the start of the 2015 regular season, Carolina now has defensive lineman Kawann Short expecting a big raise too. Given that their defensive pressure on opposing quarterbacks is what made the Panthers so scary in 2015, it makes sense that they would want to double down on their strengths and live with letting Norman walk.

At the other extreme, Oakland (-65.0%), Buffalo (-66.2%), Jacksonville (-70.6%), and San Francisco (-72.6%) had the smallest drops in DVOA when pressure was removed from the equation. Of those four, only Oakland had an overall defensive DVOA in the top half (at 15th), and their position at the bottom of the drop-off list can largely be explained by their relative strength in DVOA when unable to get pressure on the quarterback. The other three had a smaller drop-off because their DVOA ratings with pressure were among the worst in the league.

Teams can manufacture pass pressure in a variety of ways, whether by relying on their front seven or by sending defensive backs blitzing from the secondary. The Rams have invested a lot in their defensive line, with five former first-round picks suiting up for them in 2015. However, according to ESPN Stats & Info, they also sent defensive backs on blitzes more than anyone else in the league, 20.7 percent of the time. Behind them were the Steelers (19.0 percent), the Jets (18.2 percent), the Cardinals (17.1 percent), and the Bills (14.9 percent).

On the other hand, New England called the fewest number of blitzes using a defensive back, doing so on only 4.1 percent of all pass plays. They were closely followed by Philadelphia (4.9 percent), Seattle (5.1 percent), Chicago (5.3 percent), and Atlanta (5.4 percent). After the Chandler Jones trade, I would not be surprised if the Cardinals and Patriots finished closer to each other in defensive back blitz rate in 2016.

Deciding that you want to send extra rushers is all well and good, but it doesn't mean a whole lot if you aren't successful at the end of the play. The table below shows teams' frequencies of using defensive backs on blitzes and how successful they were at generating pressure when they did so. Defenses are sorted by their frequency of sending a defensive back on a blitz. Success here is defined as preventing 45 percent of needed yards on first down, 60 percent of needed yards on second down, or 100 percent of needed yards on third/fourth down.


DB Blitz Rates, 2015 Defenses
Defense DB Blitz Pct Rank DB Blitz Suc Rank
STL 20.7% 1 64% 12
PIT 19.0% 2 61% 16
NYJ 18.2% 3 67% 7
ARI 17.1% 4 63% 14
BUF 14.9% 5 66% 8
GB 14.3% 6 57% 20
SD 12.7% 7 58% 19
DET 12.1% 8 56% 26
CLE 11.6% 9 53% 28
BAL 11.4% 10 52% 29
TEN 11.0% 11 60% 17
SF 10.7% 12 56% 25
NYG 10.4% 13 48% 30
CIN 10.1% 14 57% 23
NO 9.9% 15 54% 27
JAC 9.9% 16 70% 5
Defense DB Blitz Pct Rank DB Blitz Suc Rank
CAR 9.7% 17 70% 4
TB 8.7% 18 40% 32
IND 8.7% 19 60% 18
MIN 8.1% 20 65% 11
KC 8.0% 21 71% 3
DEN 7.1% 22 64% 13
DAL 6.5% 23 76% 1
HOU 6.2% 24 68% 6
WAS 6.1% 25 63% 15
MIA 6.0% 26 56% 24
OAK 5.5% 27 57% 22
ATL 5.4% 28 47% 31
CHI 5.3% 29 57% 21
SEA 5.1% 30 75% 2
PHI 4.9% 31 66% 9
NE 4.1% 32 65% 10

Tampa Bay had the lowest success rate on plays where they blitzed a defensive back, at 40 percent. The Buccaneers were joined in the bottom five by Atlanta, the Giants, Baltimore, and Cleveland. Atlanta may not have had a good defense in Year 1 under Dan Quinn, but at least they had the good sense to not send additional defensive backs as rushers when they weren't having a lot of success on plays where they blitzed a defensive back.

The most successful team when sending a defensive back on a blitz was Dallas, with a success rate of 76 percent. Closely following them were Seattle at (75 percent), Kansas City (71 percent) and then Carolina and Jacksonville (both at 70 percent).

If anyone was wondering whether there is a relationship between a defensive back blitz rate and defensive success rate, the short answer is no -- in 2015, at least. The correlation coefficient of the two rates was -0.04, small enough to be basically meaningless. Pressure can be the key to success for a team's defense, but there is no one "best" way to generate pressure without regard to personnel.

Posted by: Guest on 13 Jul 2016

11 comments, Last at 17 Jul 2016, 10:43am by Noah Arkadia

Comments

1
by Jerry :: Wed, 07/13/2016 - 8:40pm

Looks like there's a tagging issue at "The bad news is that the second-worst Bears".

2
by Vincent Verhei :: Wed, 07/13/2016 - 9:38pm

Fixed. Thank you!

3
by BroncosGuyAgain :: Wed, 07/13/2016 - 10:19pm

s/b "Bills", yes?

And thank you for the article. An unusual analytical angle.

4
by skepticismissurvival :: Thu, 07/14/2016 - 5:26am

For the DB blitz %, does that mean percentage of total defensive plays or percentage of blitzes that a DB was sent on?

8
by Aaron Schatz :: Thu, 07/14/2016 - 1:07pm

Percentage of pass plays.

5
by ChicagoRaider :: Thu, 07/14/2016 - 10:17am

I think it is interesting that the Raiders, with their poor secondary, benefitted the least from quarterback pressure. Isn't the conventional wisdom that pressure helps a bad secondary?

6
by Noah Arkadia :: Thu, 07/14/2016 - 11:20am

It's pretty clear that a bad DVOA without pressure describes the (in)ability to cover. But what does a bad DVOA with pressure mean? Bad luck? Or is it a reflection of blitzing and putting more pressure on your DBs to cover one on one? Especially when it's coupled with a relatively good DVOA without pressure, as in the cases of Oakland, Indianapolis and Houston.

10
by Vincent Verhei :: Fri, 07/15/2016 - 7:12pm

Here's how I'd rank these stats in terms of accuracy and relevance.

1) PRESSURE RATE: Every single defense is much better when they get pressure on the quarterback, so the frequency of pressure is pretty obviously more important than the specific results of said pressure. Does not account for blitzes; doing so would probably make it more accurate

2) DVOA WITHOUT PRESSURE: Probably the best way to measure each team's secondary (and linebackers in coverage) separate from the pass rush. Again, accounting for blitzes would make it more accurate.

3) DVOA WITH PRESSURE: This is largely redundant with sack rate. The five worst defenses with pressure were JAC, SF, NYG, CLE, and BUF. Four of those teams were in the bottom five in the league in sack rate; the Jags were 22nd. Things aren't quite as consistent at the top of the table, but the best defense with pressure, DEN, was also first in sack rate. Of the next top four teams in defense with pressure, GB was 7th in sack rate, CAR was 11th, SD was 18th, and PHI was 24th.

The correlation between DVOA with pressure and sack rate was -0.567, which is pretty strong as far as football stats go.

11
by Noah Arkadia :: Sun, 07/17/2016 - 10:43am

Thanks for the reply, Vince. That's interesting about the correlation between sack rate and DVOA with pressure.

7
by IlluminatusUIUC :: Thu, 07/14/2016 - 11:52am

It looks to me like Buffalo's statistics were pretty good when they sent a DB on a blitz but generally horrendous overall. I guess that in some way validates the criticisms of our big four DL, in that one got hurt, one started a personal foul parade, and another took a paid vacation.

9
by cjfarls :: Thu, 07/14/2016 - 3:11pm

While some of this is pretty obvious (DEN and CAR's defenses being very good, etc.), without some linkage to the number of rushers total, it gets hard to interpret this. E.g., if you can only get pressure by sending 5+ folks, it may not matter if you didn't blitz a DB because there is still likely to be a coverage hole.

Might explain some of the weird questions (like the OAK secondary one above) that these stats raise.