Writers of Pro Football Prospectus 2008

Most Recent FO Features

McKinneyBen15.jpg

» 2017 Play-Action Defense

Our look at play-action pass in 2017 flips to the defensive side of the ball. Carolina was historically good, Houston was historically bad, and a long-standing question about year-to-year correlation gets cleared up.

03 Jul 2018

Defense and Pass Pressure 2017

by Scott Kacsmar

This is not breaking news, but all quarterbacks play worse when they are under pressure compared to a clean pocket. That's why teams place a premium on acquiring pass-rushers in the draft and free agency. The only players who can compare to quarterbacks in salary right now are pass-rushers. We are still waiting to see when the Rams will make Aaron Donald the first non-quarterback in NFL history to average $20 million per season. If it's not Donald, then it will be Khalil Mack in Oakland. Mack ranked third in quarterback hurries last season with 52, according to Sports Info Solutions (subscription required). Donald had 48.5 hurries, an incredible number for a defensive tackle.

We tend to groan and joke about how the latest "highest-paid quarterback in NFL history" usually hasn't been a player of the highest caliber. (Think Joe Flacco, Derek Carr, or Jimmy Garoppolo.) But teams are usually getting a desired impact when they ink a pass-rusher. Denver's Von Miller reset the market a few years ago with a deal that pays him just over $19 million annually. He finished second last season with 53 hurries. Demarcus Lawrence led the NFL with 55 hurries, and the Cowboys are set to pay him over $17 million this season on the franchise tag.

These star players are often the driving force behind their unit's pass rush, but their impact is still not on even footing with that of a franchise quarterback. Despite their best efforts last year, the pass defenses for Miller's Broncos (15th in DVOA), Lawrence's Cowboys (18th), and Mack's Raiders (30th) were nothing special.

Pressure is a desirable goal for every defense, but its presence alone does not guarantee a positive outcome. That's why we like to break things down by DVOA to see how a defense fares with pressure versus plays without pressure getting there. Our data for pressure comes primarily from Sports Information Solutions, further informed by the data we also collect from ESPN Stats & Info. You can read last year's results here.

Under Pressure … More Often

The 31.6 percent pressure rate in 2017 was the highest in any season since 2010. However, the trend of offenses getting better under pressure continued. Keep in mind that since this is for defense, the more negative the DVOA, the better for the defense.

Year Pressure Rate DVOA with Pressure DVOA no Pressure
2010 23.4% -76.8% 31.4%
2011 23.2% -80.0% 30.9%
2012 20.2% -89.7% 29.9%
2013 24.7% -75.2% 32.6%
2014 24.6% -76.0% 34.4%
2015 25.6% -67.3% 35.5%
2016 27.1% -60.3% 33.5%
2017 31.6% -55.5% 37.5%

Defenses used to post a DVOA of -75.0% or better when getting pressure, but for the third year in a row we saw that decrease, and it was down to -55.5% in 2017. The DVOA without pressure (37.5%) in 2017 was also at its highest since 2010. The skill sets of modern quarterbacks have probably never been better for handling pressure. Quarterbacks such as Aaron Rodgers and Russell Wilson can be magicians on the field, turning sure-sacks into big plays for their offenses. If you cannot go that route, then you better master the game pre-snap and get rid of the ball in a hurry, which is how Tom Brady and Matt Ryan have become two of the best quarterbacks under pressure over the years. Blindly throwing interceptions under pressure is an easy way to reserve a seat on the bench.

According to data from ESPN Stats & Info for 2010 to 2017, we found that teams won 40.3 percent of games when their quarterback was pressured at least 34.7 percent of the time, or one standard deviation above average. When the quarterback was pressured less than 15.4 percent of the time (one standard deviation below average), those teams won 61.7 percent of their games.

For as much as we talk about the importance of pressure, we have to remember that even the best defenses rarely ever have any week-to-week consistency of getting to the quarterback more than a third of the time. One really effective pressure that causes a sack or turnover can do more damage than three non-splash play pressures in a row.

Since 2010, the correlation between a defense's pressure rate and DVOA is -0.33 -- the defenses that get pressure frequently tend to outplay those that don't. The correlation between a defense's DVOA with pressure and overall DVOA is stronger at 0.43 -- defenses that play well with pressure tend to play well without it too. But what is most telling of a defense tends to be how it defends when pressure doesn't get there. The correlation between DVOA without pressure and overall DVOA is 0.80, since that's how most of the game is played defensively. We also see higher year-to-year correlations when looking at DVOA without pressure. Year-to-year correlation for defensive DVOA is already low at 0.34 since 2010, but it's just 0.06 if looking at DVOA with pressure against next year's overall DVOA. That jumps up to 0.25 using DVOA without pressure.

Year-to-year defense (2010-2017) Correlation
DVOA to DVOA 0.34
DVOA with pressure to DVOA 0.06
DVOA no pressure to DVOA 0.25
DVOA with pressure to DVOA with pressure 0.23
DVOA no pressure to DVOA no presure 0.28

Pressure is an amazing tactic to bring chaos and randomness to the game, but good coverage matters too, as we will see with the 2017 splits.

The 2017 Data

The following table shows each defense's success with and without pressure in the 2017 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. The table is sorted by descending pressure rate.

2017 Defenses: With and Without Pass Pressure
Defense Plays Pressure
Plays
Pressure
Rate
Rk DVOA
w/Pressure
Rk DVOA
no Pressure
Rk DVOA Dif. Rk
WAS 609 233 38.3% 1 -74.3% 7 36.0% 14 -110.3% 8
DAL 632 228 36.1% 2 -36.1% 25 42.1% 19 -78.1% 23
JAX 593 209 35.2% 3 -98.8% 2 14.3% 1 -113.1% 7
CIN 646 225 34.8% 4 -67.8% 11 53.0% 30 -120.7% 3
PIT 580 202 34.8% 5 -75.4% 6 34.0% 13 -109.3% 9
CAR 637 220 34.5% 6 -60.9% 14 26.4% 9 -87.3% 19
DEN 553 186 33.6% 7 -31.7% 26 29.4% 11 -61.1% 31
PHI 671 225 33.5% 8 -82.5% 4 37.8% 15 -120.4% 4
IND 563 186 33.0% 9 -20.6% 30 52.8% 28 -73.4% 25
NO 633 206 32.5% 10 -67.3% 12 24.4% 5 -91.7% 17
LAR 635 206 32.4% 11 -69.2% 10 16.7% 2 -85.9% 21
TEN 672 216 32.1% 12 -47.0% 20 47.9% 26 -95.0% 15
SEA 626 200 31.9% 13 -60.7% 15 32.5% 12 -93.1% 16
MIN 613 195 31.8% 14 -86.1% 3 20.4% 3 -106.4% 10
CLE 589 186 31.6% 15 -17.9% 31 44.1% 24 -62.0% 30
NYJ 613 192 31.3% 16 -27.9% 27 39.2% 17 -67.1% 27
Defense Plays Pressure
Plays
Pressure
Rate
Rk DVOA
w/Pressure
Rk DVOA
no Pressure
Rk DVOA Dif. Rk
KC 639 200 31.3% 17 -39.1% 23 47.9% 25 -87.0% 20
LAC 599 186 31.1% 18 -72.2% 8 25.8% 8 -98.0% 13
ATL 622 191 30.7% 19 -52.9% 18 49.1% 27 -102.0% 11
ARI 626 192 30.7% 20 -71.8% 9 25.3% 7 -97.1% 14
BAL 640 196 30.6% 21 -100.2% 1 24.7% 6 -124.9% 1
CHI 598 183 30.6% 22 -42.3% 22 27.0% 10 -69.2% 26
SF 609 186 30.5% 23 -22.4% 29 43.2% 21 -65.6% 28
HOU 566 169 29.9% 24 -25.5% 28 39.1% 16 -64.6% 29
NYG 620 185 29.8% 25 -48.5% 19 42.6% 20 -91.1% 18
GB 575 171 29.7% 26 -62.1% 13 59.5% 32 -121.6% 2
DET 631 183 29.0% 27 -76.1% 5 43.2% 22 -119.3% 5
NE 659 190 28.8% 28 -36.3% 24 40.7% 18 -77.1% 24
OAK 587 167 28.4% 29 -55.4% 17 59.4% 31 -114.8% 6
MIA 581 165 28.4% 30 -46.4% 21 52.8% 29 -99.2% 12
BUF 627 175 27.9% 31 -57.3% 16 23.5% 4 -80.8% 22
TB 600 156 26.0% 32 -17.5% 32 43.3% 23 -60.7% 32
NFL 614 194 31.6% - -55.5% - 37.5% - -93.0% -

Washington may come as a surprise as the defense with the highest pressure rate (38.3 percent) last season. The Redskins were only No. 12 in pressure rate in 2016, but none of the top four teams here ranked higher than 12th in 2016. Washington has used a lot of high draft picks on edge rushers over the years, but old reliable Ryan Kerrigan led the way with 35 hurries and 13 sacks. Preston Smith chipped in 29 hurries, and Junior Galette (24 hurries) finally returned to action after tearing his Achilles two years in a row. Washington has not brought Galette back, but looks to get more than five games out of Jonathan Allen, a first-round rookie who was injured. What Washington really missed last season was better play in the secondary. Although cornerback Josh Norman was not thrown at very frequently, he was disappointing when he was targeted, ranking just 60th in adjusted success rate and 76th in adjusted yards per pass allowed, according to Sports Info Solutions charting.

Staying in the NFC East, the Super Bowl champion Eagles were solid across the board with the No. 8 pressure rate and fourth-best DVOA with pressure. The Eagles did not generate more than 4.0 sacks in any game last season, and only had one sack in Super Bowl LII, but it was the big strip-sack of Tom Brady by Brandon Graham late in the game. The Eagles also pressured Case Keenum into a big pick-six in the NFC Championship Game, so they took down the two best quarterbacks under pressure in 2017 in consecutive playoff games. The Eagles, Bengals, Panthers, and Broncos were the only four teams to finish in the top 10 in pressure rate in 2016 and 2017. The Bengals have quietly matched Denver in doing so in the last three seasons. (We try to find some positives to justify Marvin Lewis' never-ending tenure.)

No team improved its ranking in pressure rate from 2016 more than Dallas, climbing from 29th to second. A lot of that was Lawrence, but Dallas really needs to stop losing games to suspensions on the defensive side of the ball. The Colts (23) and Steelers (19) were the only other defenses to jump more than a dozen spots in the pressure rate rankings from 2016.

No defense fell farther in pressure rate in 2017 than Miami, which went from fourth in 2016 to 30th last year. Cameron Wake still led the way, but Ndamukong Suh saw his hurries drop from 35 to 20.5, and his 4.5 sacks are the fewest he's ever had in a full season. Miami added Robert Quinn from the Rams in a March trade, but could still expect to see more out of 2017 first-round pick Charles Harris, who had 19 hurries last year.

The Florida team to really pick on is Tampa Bay, which had the lowest pressure rate (26.0 percent), and even when the pressure did get there, the Buccaneers still had the worst DVOA (-17.5%). As Aaron Schatz already pointed out, nine of Tampa Bay's 22 sacks (a league-high 41 percent) were non-pressure plays (coverage sacks or plays where the quarterback was at fault). Tampa Bay's solution was to cut underwhelming veterans in the trenches and surround Gerald McCoy with Jason Pierre-Paul (40 hurries in 2017), Vinny Curry (28 hurries), and first-round rookie Vita Vea. Curry actually had more hurries as part of Philadelphia's rotation than anyone in Tampa Bay last year. That influx of talent should make a difference.

In concluding our trip around Florida, Jacksonville was the defense to celebrate last season. The Jaguars finished third in pressure rate, second in DVOA with pressure, and first in DVOA without pressure in their return to the playoffs. Calais Campbell was an excellent signing and led the team with 37 hurries, and Jalen Ramsey had a fantastic second year in leading a secondary that provided the best coverage in the league. The Jaguars racked up two road games with 10 sacks apiece, against the Texans (Week 1) and Colts (Week 7). There has only been one other game with 10 sacks by a defense in the NFL's last five seasons, when Washington taking down Jacksonville's Chad Henne 10 times in 2014. Jacksonville can likely expect to see regression in its sack numbers in 2018 given those two big performances came against the likes of Tom Savage (benched) and Jacoby Brissett (backup). Still, it's impressive that Jacksonville did something twice last season (10-plus sacks in road games) that the rest of the NFL has done twice in the last 20 years. The 2017 Jaguars are the first defense since the 2006 Ravens to have four players with at least 8.0 sacks.

Speaking of the Ravens, that was the only defense to have a better DVOA with pressure (-100.2%) than Jacksonville last year, which contributed to the Ravens having the largest DVOA drop (-124.9%) without pressure. Baltimore was still very good (No. 6 at 24.7%) without pressure, but it's important for that defense to keep developing young talent. Terrell Suggs (36 in October) can't do this forever. He was still great last year with 47.5 hurries and 11 sacks, but Za'Darius Smith (29) was the only other Baltimore defender with more than 16 hurries.

Green Bay faces a similar dilemma in trying to find a pass rush beyond just Clay Matthews and Nick Perry, who tied for the team lead with only 23.5 hurries each last year. Long-time defensive coordinator Dom Capers is gone. Mike Pettine returns to the NFL and will hope to have a big impact on Perry much in the way that he did for Jerry Hughes in Buffalo in 2013. The Packers also added Muhammad Wilkerson in a low-key free agency signing that could play huge dividends if he returns to his 2015 form. The Packers had the worst DVOA (59.5%) when pressure didn't get there last season, so that young secondary needs all the help it can get.

We briefly mentioned the Rams earlier, but lest we forget that this defense has added Marcus Peters, Aqib Talib, and Ndamukong Suh. The Rams were already No. 2 last year in DVOA without pressure, so imagine the potential of Suh and Donald getting in quarterback's faces and forcing passes with Talib and Peters in coverage. The experiment may not produce a Super Bowl right away, but look for the Rams to stomp the lesser opponents on their schedule this year.

Later this week, we'll continue our look at pass pressure by checking out each team's splits in how often it brought different numbers of pass-rushers: the standard four, a blitz of five or more, or a smaller pass-rush of just two or three.

Posted by: Scott Kacsmar on 03 Jul 2018

12 comments, Last at 12 Jul 2018, 9:16pm by The Ninjalectual

Comments

1
by The Ninjalectual :: Wed, 07/04/2018 - 12:06am

I think it's ironic that the 2011 draft was the one time Washington did the "sensible thing" (traded down in the first round for extra picks, still getting a player they like). They even ended up with 3x pro-bowler Ryan Kerrigan. But the guy they passed on is J.J. Watt.

2
by Jimmy Oz :: Wed, 07/04/2018 - 2:58am

What's the definition of "pressure"?

EDIT: I know what I think pressure is, but wondering if there's an FO definition...

3
by Bryan Knowles :: Wed, 07/04/2018 - 3:25am

A pressure is a passing play where the quarterback is hurried or hit by the pass rush. If the quarterback is forced to throw the ball early because a defender is racing towards him, or ends up scrambling forward as the pocket collapses, or gets knocked to the turf, or panics and throws the ball straight up in the air (the "Peterman Special") it counts as a pressure.

There isn't a strict mathematical definition or something; there's not a precise distance or time requirement for a defender to cause a pressure. It's a charting stat -- if the play was disrupted by a pass rusher, it's a pressure.

There is an art determining just what qualifies as a pressure, so if you want to be pedantic about it, this is technically tackling "plays which were marked as a pass pressure by one of our two game charting companies, Sports Info Solutions and ESPN Stats & Info, with disagreements being determined by us on a case-by-case basis." In general, though, it basically means what you think it means.

4
by Noah Arkadia :: Thu, 07/05/2018 - 10:45am

It's clear what a bad DVOA without pressure means (bad coverage), as well as a bad pressure rate (bad pass rush), but what does it mean (if anything) when there's good DVOA without pressure but bad DVOA with pressure (Denver), or the opposite (Cincinnati, Green Bay)?

...Never mind, answered my own question, as from the tables above it seems that DVOA with pressure is more a matter of luck than anything else. Likely who the opposite QB is, as well.

5
by dank067 :: Thu, 07/05/2018 - 2:27pm

Wonder if there's just some missing context, such as if there was a correlation between DVOA with pressure with how often teams blitz a particular number of pass rushers. It would make sense if teams that are good at generating pressure with just 4 rushers put up a better DVOA with pressure since they still leave 7 in coverage. On the other hand, teams that disproportionately send 5-6 players might be able to generate a high pressure rate, but too often leave the secondary exposed and take a DVOA hit. Sounds like some of those numbers might be coming next week.

8
by Noah Arkadia :: Fri, 07/06/2018 - 2:18pm

Good point, let's wait and see.

6
by jtr :: Fri, 07/06/2018 - 10:20am

>The 31.6 percent pressure rate in 2017 was the highest in any season since 2010. However, the trend of offenses getting better under pressure continued.

Could this be a trend in the charting rather than in the game? If the charters are tending to call more plays "pressures" than in the past, we would expect to see both a rise in the number of pressure snaps and a rise in QB performance under pressure as more fringe pressure plays got charted that way. Plus the fact that pressure rate hung steady from 20-25% from 2010-2016 and has now suddenly jumped to 31% in two years has me suspicious.

I don't buy the argument here that pressure performance is rising because of the current QB's. It's not like 2010 was lacking for pocket magicians, with Roethlisberger playing wild backyard football, Mike Vick doing his Mike Vick thing for Philly, Tony Romo spinning away from free rushers, Manning and Brady shredding defenses with pre-snap reads and quick releases...

7
by Aaron Schatz :: Fri, 07/06/2018 - 12:52pm

No, it absolutely could be a trend in charting. Although I believe both ESPN S&I and SIS are finding the league-wide pressure rate increasing. My guess is that it's partly related to the poor quarterback play overall last year, the same way that scoring and yardage dropped a little in 2017. There were just too many good injured quarterbacks and too many bad quarterbacks forced into lineups. I would not be surprised to see the pressure trend reverse in 2018.

9
by jtr :: Fri, 07/06/2018 - 8:43pm

I always appreciate that FO is straightforward in acknowledging what they don't know. Football is complicated and messy and hard to analyze, and you guys don't pretend otherwise.

I'm not sure if the backup QB theory holds since DVOA against pressure is up; we would expect that to go down if this was the result of a bunch of Nathan Petermans Petermanning up the joint.

Maybe it's just that the NFL has an embarrassment of riches at the pass rushing positions. Or a new batch of charters who are more willing to call a play a pressure. We'll have to see what next year's data shows.

10
by ChrisS :: Mon, 07/09/2018 - 1:05pm

It is probably at least partially a trend. It seems to me that pass rushing was historically a bit undervalued and I think it is now moving towards being overvalued, along the lines of Left OT suddenly becoming a marquee position and over the last few years not so much.

11
by Bright Blue Shorts :: Mon, 07/09/2018 - 2:36pm

And I'll add that the CBA limiting practice time was introduced in 2011 and we've been hearing that offensive lines have been struggling to gel since then.

12
by The Ninjalectual :: Thu, 07/12/2018 - 9:16pm

A lot of that is that there isn't nearly as much of a "blind side" in shotgun. Look at how many teams send their best pass rusher against the RT these days.