Quarterbacks and Pressure 2018
by Scott Spratt
Rule changes in recent seasons have turned a lot of sacks into roughing-the-passer penalties and automatic first downs. But exceptional pass-rushers like Khalil Mack still command multiple first-rounders in trades and upward of $100 million in guaranteed money because pressure has a marked effect on offensive success even if that pressure does not end with a hand on the quarterback. Across the NFL, quarterbacks produced twice as many yards per play (7.8 vs. 3.9) and enjoyed a near-100% jump in DVOA when they worked in a clean pocket versus when they dealt with pressure. And only the best pressure-facing quarterback – Patrick Mahomes – performed better under pressure than the worst quarterback – Josh Rosen – fared without it.
|Qualifying Quarterbacks with and without Pass Pressure, 2018|
Two seasons don't necessarily make a trend, but Deshaun Watson has taken a couple of steps down the path Russell Wilson has followed his entire career. Wilson has been one of the five most pressured quarterbacks all seven of his NFL seasons, and he has finished in the top 10 in DVOA with pressure in six of them.
|Russell Wilson vs. Deshaun Watson Under Pressure|
|Russell Wilson||Deshaun Watson|
|1||27.5% (3)||-48.7% (2)||71.4% (2)||41.9% (1)||-36.0% (10)||99.3% (1)|
|2||36.6% (2)||-27.6% (4)||67.3% (6)||41.1% (1)||-27.4% (5)||56.2% (14)|
|3||39.1% (1)||-44.5% (5)||75.3% (3)|
|4||31.7% (4)||-32.4% (4)||86.4% (1)|
|5||34.9% (3)||-51.6% (14)||49.1% (13)|
|6||39.8% (2)||-21.0% (4)||64.6% (8)|
|7||36.7% (4)||-33.8% (7)||73.8% (3)|
The salary cap limits teams' resources and effectively requires them to sacrifice the quality of some position groups. That's especially true when teams sign a quarterback after his rookie deal. As much as it violates conventional wisdom, Wilson's and Watson's specific strengths and weaknesses make their offensive lines natural choices for their teams to neglect in favor of skill-position and defensive talent. Watson's freshman and sophomore seasons suggest the Texans are following that blueprint. They'll just have to hope that Watson also shares Wilson's preternatural ability to avoid hits and major injuries. Watson has endured a handful of injuries in his college and pro careers to date, but the bulk of his missed time has stemmed from a pair of non-contact ACL tears suffered in practices that don't necessarily portend future injuries.
The rookies Josh Allen and Josh Rosen handled their frequent pressure very differently. Allen acquitted himself well with a top-10 finish in DVOA with pressure, ahead of a handful of cool-headed veterans including Matt Ryan (13th), Drew Brees (14th), Aaron Rodgers (17th), and Tom Brady (22nd). In contrast, Rosen had the worst DVOA with pressure in football. But both Allen and Rosen were in the bottom three in DVOA without pressure, and multi-year trends suggest that this is a more consistent identifier of successful and unsuccessful quarterbacks than performance with pressure.
At first glance, Rosen can offer a few additional reasons for optimism. His limited mobility makes him more sensitive to the quality of his pass-blocking, and his trade rescues him from a Cardinals team with coaching instability and a bad offensive line. However, the Dolphins were one of just six teams with a higher adjusted sack rate than the Cardinals in 2018, and they lost who may have been their best offensive lineman, Ja'Wuan James, to a $51-million free-agent deal with the Broncos.
Regression may be all the Dolphins' pass-blocking needs to outperform that of the 2018 Cardinals, but the statistical evidence doesn't support a narrative that poor blocking sped up the game for Rosen with or without pressure. The correlation between pressure rate and DVOA without pressure is weak, while the moderate positive correlation between pressure rate and DVOA with pressure runs counter to expectation and suggests that the more pressure that quarterbacks see, the better they perform under pressure.
|Pressure Rate Correlations, 2010-18|
|Pressure Rate to DVOA w/Pressure||+0.27|
|Pressure Rate to DVOA no Pressure||+0.06|
I think that trend illustrates that quarterbacks share responsibility for their pressure rates with their blocking. Mobile quarterbacks like Watson and Wilson own a disproportionate share of the heaviest pressure seasons from the past decade, and the other end of the list is full of all-time great pocket passers. Peyton Manning owns the three best pressure-rate seasons since 2010, and he accomplished that across two different teams.
As dark a picture as the numbers paint for Rosen, he can still point to Jared Goff as the example of a player who defied the trends. Goff wasn't just bad as a rookie. He was standard deviations worse than the who's who of quarterback busts – Blaine Gabbert, Christian Ponder, and Jimmy Clausen, to name a few – on their plays without pressure.
|Best DVOA w/o Pressure, Rookies, 2010-18|
|2012||Robert Griffin III||WAS||473||66.7%|
It is an incredible story that Goff turned that around. But Sean McVay makes Goff's pressure splits the most fascinating part of this analysis. With McVay, Goff went from worst to nearly first in DVOA without pressure, but he has made only modest improvements to his DVOA with pressure. It's hard not to read those splits as evidence of McVay's brilliance rather than Goff's. McVay famously communicates his pre-snap reads to help simplify Goff's responsibilities, and when the Patriots switched defenses after communications cut off with fewer than 15 seconds on every play clock in the Super Bowl, Goff struggled. After throwing to the same receiver running the same route on the previous play, Goff put the final nail in the Rams' coffin when he panicked in reaction to the Patriots' fourth-quarter blitz and threw an easy pass for Stephon Gilmore to intercept. That decision punctuated a season full of similar mistakes.
Clearly, Goff's poor play under pressure is not a dealbreaker or the Rams would not have been in the Super Bowl in the first place. McVay has done an excellent job in limiting the amount of pressure Goff sees, cutting the team's pressure rate from a league-leading 40.4 percent the year prior to his arrival to 30.7 and 25.5 percent the last two seasons, both bottom-10 in football. And the offseason should provide McVay the time to create strategies like a hurry-up offense to counter the blueprint that Bill Belichick's play-switching provided for other teams.
Meanwhile, Goff has a more realistic role model for his potential improvement than Rosen does in Goff. Kirk Cousins was a bottom-third performer under pressure in each of his first two seasons as a starter and has climbed all the way to the top 10 as of 2018. One could argue that Goff is even ahead of that pace given that Cousins spent the first two years of his career on the bench.
|Kirk Cousins vs. Jared Goff Under Pressure|
|Kirk Cousins||Jared Goff|
|1||40.4% (1)||-126.5% (34)||-45.2% (34)|
|2||30.7% (22)||-56.8% (21)||83.4% (2)|
|3||17.6% (35)||-75.2% (20)||35.6% (25)||25.5% (30)||-67.9% (24)||69.1% (5)|
|4||20.5% (33)||-110.1% (35)||69.0% (4)|
|5||24.9% (22)||-54.5% (15)||65.5% (4)|
|6||35.6% (9)||-55.8% (18)||60.0% (13)|
|7||34.5% (5)||-34.3% (8)||39.2% (25)|
It has somehow taken me 12 paragraphs to mention that Patrick Mahomes finished first in DVOA both with and without pressure. But you didn't need an analysis of his pressure splits to know he's really good. As has been a running theme of this article, the latter accomplishment is more impressive than the former. Even restricted to full-year starters, the DVOA-with-pressure leaderboard from the last decade includes a couple of odd names in Case Keenum and Josh Freeman.
|Best Single-Season DVOA w/ Pressure, 2010-18|
|Minimum 500 Total Plays|
In contrast, the list of the best DVOA-without-pressure seasons exclusively features exceptional quarterbacks with a possible caveat depending on your opinion of Goff.
|Best Single-Season DVOA w/o Pressure, 2010-18|
|Minimum 500 Total Plays|
Mahomes was the silver medalist of the decade in both DVOA with pressure and without pressure, somehow finishing second in each to Tom Brady from two different seasons. But Brady never finished first in both in the same season. And no one else besides Mahomes has either.
The 2018 leaders in DVOA without pressure are about as unassailable as the full decade's list.
|Best DVOA without Pressure, 2018|
Then-35-year-old Ryan Fitzpatrick's sixth-place finish was undoubtedly an outlier. He owns just one other season this decade outside of the bottom 10. But 23-year-old Nick Mullens' eighth-place finish deserves greater consideration. The earlier leaderboard shows that Mullens' performance without pressure was of similar quality to Watson, Wilson, Dak Prescott, and Robert Griffin III in their rookie seasons. But Mullens' (non-) draft status isn't his only differentiator. Crucially, Mullens tallied his impressive total over just 201 plays without pressure (292 total plays). Most of the other solid rookie performers – such as Colt McCoy (262 plays), Cody Kessler (223), and 49ers teammate C.J. Beathard (256) – who failed to sustain that success did so over a limited sample in their freshmen seasons. Sample size is really important.
|Worst DVOA w/o Pressure, 2018|
Most of the DVOA-without-pressure trailers in 2018 were young players who will have a chance to improve as they gain experience. But Jaguars fans have to be concerned that the team replaced one veteran with consistently poor DVOA-without-pressure totals for another. Foles led all quarterbacks with a 94.8% DVOA without pressure in his breakout 2013 season, but he has finished 26th (34.8%), 35th (11.5%), and 31st (24.4%) in his three qualifying seasons since. Like Bortles, Foles has typically performed better by rank when under pressure than without it, and one could argue that is a positive trait for the quarterback of a run-and-defense-focused team that will want to minimize its major mistakes. That argument is just more palatable for a player who is earning less than $50 million in guaranteed money.
21 comments, Last at 13 Jul 2019, 11:12am
#1 by PirateFreedom // Jul 08, 2019 - 7:54pm
The suggestion that performance without pressure was more indicative of quarterback quality then performance under pressure was surprising and interesting to me.
There is an advantage when comparing samples to having a sample set where subjects have fewer variations in circumstance. I just would have guessed that playing under pressure was an important skill in itself that would be a huge part of a QB's 'final score'
The lists sure make it look like the unpressured numbers are more predictive though.
#3 by mehllageman56 // Jul 09, 2019 - 12:24pm
The majority of passing plays end up not having any pressure on the quarterback; only Watson suffered a pressure rate above 40%. Therefore, a quarterback's play without pressure ends up more indicative of their overall play. The corollary to this is that NFL teams should be devoting more resources to their secondary and coverage linebackers than to pass rushers. It might even make sense to keep one less rush linebacker for a 3-4 team, when you can just blitz anyone from the secondary, a la Rex Ryan's defenses. Which, by the way, didn't create a whole lot of sacks, but did create a lot of interceptions, incompletions, and failed completions.
#20 by Noahrk // Jul 13, 2019 - 11:12am
Yes, but don't forget what is stated in the article, that pressure creates considerably better results than no pressure. What is more interesting to me is that there is surely a correlation between coverage and pressure. After all, pressure as a stat does not indicate how long it took for it to be created. Good coverage allows more time for pressure to get there. I think the biggest difference between investing in coverage and passrushers is that you need an entire group of good players to have good coverage, while a single passrusher can create pressure without any help, or at the very least create more opportunities for his teammates.
#4 by Scott Spratt // Jul 09, 2019 - 4:01pm
It may be the case that a good quarterback handles pressure by avoiding it rather than by performing well when it happens. To me, that's what the positive correlation between pressure rate and DVOA with pressure suggested.
#8 by sbond101 // Jul 10, 2019 - 7:29am
There is definitely something more going on here relating to the correlation between pressure rate, finding receivers open "in rhythm", and pressure DVOA. Looking at the opposite extremes on one side you have Brady, who always ranks low in pressure rate in part because he so often has the ball out quickly - usually if QB's in this mold hold the ball (and these snaps correlate with recorded pressures more often) something bad has, or is going to, happen. On the opposite extreme you have Wilson, Mahomes, Watson, old-time Rothesburger, etc... that intentionally hold the ball to force opposing secondary's to cover longer knowing that they are great at escaping under the belief that the big play is worth the risk - this strategy necessarily exposes them to greater pressure rates and would be really foolish if their DVOA on those snaps didn't justify the approach. - None of this is intended to say pressure rate is not useful - But there is definitely something else going on that makes pressure snaps less comparable than non-pressure snaps.
#9 by Scott Spratt // Jul 10, 2019 - 8:44am
I agree with your point here. I think it would be interesting if we could come up with a way to measure a player's tools to handle pressure, e.g. speed, agility, or size in the case of someone like Roethlisberger. I suspect that and pressure rate would relate to one another.
#10 by sbond101 // Jul 10, 2019 - 9:02am
I think the way to start is to begin to track time to pressure & time till QB release; With that information it should be possible to create a series of data points; pressure rate at 1, 2, 3, 4, & 5+ seconds, time-adjusted pressure rate etc... and start to look at this problem (and also look a lot more intelligently at OL pass protection performance). The biggest problem I foresee in this path is the decline of sample size as the ~600 dropbacks a QB might have start to get sliced up.
#2 by Theo // Jul 09, 2019 - 12:24pm
"but the bulk of his missed time has stemmed from a pair of non-contact ACL tears suffered in practices that don't necessarily portend future injuries."
Eh, I wonder how does this works.
Because he wont tear any more ACLs?
#5 by Scott Spratt // Jul 09, 2019 - 4:07pm
I'm sure Thomas Davis would love if you could only tear each ACL once! But no, this was more the case of me trying to explain a complicated thought in a few words. My perspective is the biggest injury fear you have for a quarterback comes from the hits he takes. And because Wilson and Watson run the ball more often than a typical quarterback, they open themselves up to more hits. Wilson has done a great job of minimizing those hits by getting out of bounds and sliding. We'll see if Watson can follow suit. I don't think his non-contact injuries say much either way for how often he'll be hit and how well he'll handle it.
#18 by Aaron Schatz // Jul 12, 2019 - 11:25am
If Extra Points has disappeared from your sidebar, please do us a favor and e-mail FOfirstname.lastname@example.org so we can figure out what's going wrong and fix it. Please send along your contact information so we can reach out to you and fix the problems. Thank you.
#15 by ChrisS // Jul 11, 2019 - 2:00pm
Scott, good first article. Does pressure include sacks? Keenums 2017 DVOA with pressure sure adds certainty to the flukiness of his performance that year. I remember many passes being flung under pressure and Thelen or Riggs making a great grab.
#17 by Scott Spratt // Jul 11, 2019 - 5:38pm
Thanks, Chris. The pressure rate page in the 2019 premium charting data will give full explanations of what is and isn't included in those stats when it goes live in the coming days. But, yeah, pressure rate is (non-coverage sacks + hurries) / plays, and then sometimes I use the term pressure in text to just mean hurries. Hope that helps.