Quarterbacks and Pressure, 2016
by Scott Kacsmar
Even though pass pressure only occurs on roughly a quarter of NFL pass plays, it creates havoc on the field, and can often lead to significant, game-deciding results. Have you heard the one about the time the Falcons blew a 28-3 lead? The two biggest plays in Super Bowl LI were not any of the 85 passes that left the hands of Tom Brady and Matt Ryan, but rather two sacks on ill-advised decisions by Atlanta. Ryan was strip-sacked on a third-and-1 deep in his own territory, then later went down for another sack on second-and-11 after the Falcons were in field goal range with a 28-20 lead in the final four minutes. Meanwhile, Atlanta's worn-out pass rush completely dissipated, and the rest was history. You could say Atlanta literally crumbled under pressure in the big game.
In our data for each season since 2010, every starting quarterback has had less efficiency when pressured than when throwing from a clean pocket. So any TV analyst who states "you must get after [quarterback X]" is actually just preaching an obvious truth for every game plan. No quarterback really likes to be pressured, though some do handle it with better grace than others on a consistent basis.
Today, we are looking at our charting results for quarterbacks and pressure for the 2016 regular season. These are always some of the most cited and noteworthy numbers in Football Outsiders Almanac (the 2017 edition will be available in July!). This data primarily comes from Sports Information Solutions charting, although we also have access to ESPN Stats & Information data and we use that to check against the SIS data. Pressure will always have a subjective element to it, but we feel that comparing the two sources has led to a more accurate measurement.
It is important to note that the DVOA numbers listed below do not represent each individual quarterback's passing DVOA, as seen here. Instead, this rating represents team offensive DVOA with this quarterback either passing (including sacks) or scrambling. Each quarterback listed had at least 200 pass plays, and they are listed in order by ascending pressure rate.
|Qualifying Quarterbacks with and without Pass Pressure, 2016|
It seems like Chicago always has some league-leading result in this table each year. In the past it had been Jay Cutler (2015) and Josh McCown (2013) leading the league in DVOA when pressured, but this year it was Brian Hoyer having the lowest pressure rate. Chicago's line also protected Matt Barkley (12th) fairly well, at least one part of the Bears offense was not a mess. (Jordan Howard makes it two parts.) Only Derek Carr (6.3 percent) had a lower knockdown rate than Barkley (7.9 percent). As a reminder that the quarterback plays a huge role in these numbers, Jay Cutler's knockdown rate of 21.3 percent would have been the third-highest in 2016. Hoyer only had 200 pass attempts, so he won't qualify for official NFL single-season records, but he threw zero interceptions and only took four sacks, including one he caused himself. It was an oddly mistake-free season from Hoyer, albeit in limited playing time. He will now try to excel in Kyle Shanahan's offense in San Francisco.
As for quarterbacks who did qualify for single-season rate records, only Colin Kaepernick and rookie Dak Prescott managed to avoid throwing a single interception while under pressure in 2016. The Cowboys controlled most games last season, but the 49ers often trailed; that's further evidence that Kaepernick, who remains a free agent, played better than his reputation. As for Prescott, he ranked just 24th in pressure rate despite the reputation of his offensive line. However, he had the fifth-best DVOA when pressured (-30.5%). He was also third when not pressured (77.4%), making him the only quarterback to rank in the top five in both categories last season. Tom Brady was the next closest at sixth with pressure, second without it. It's no surprise to see that Matt Ryan's MVP season was the best without pressure. However, as we discusssed in previews before the aforementioned Super Bowl, there was a bit too much pressure on him at times last year, and it proved costly in the end when he took five sacks in the big game.
No one extended the play quite like Aaron Rodgers last season. He had the best DVOA under pressure (almost positive at -3.3%) and smallest drop in DVOA when pressured. NBC had to put a timer on some of Rodgers' efforts against the Lions in Week 17, including a touchdown pass to Geronimo Allison that took more than eight seconds from the snap to completion. This type of effort almost won Rodgers another MVP, but for the Packers to get back on the pre-2015 efficiency level of offense, they'll have to start making more on-time plays.
While every quarterback had a higher DVOA without pressure than with it, their ranks were often far apart in the two categories. For instance, Jameis Winston (15), Tyrod Taylor (14), Matthew Stafford (13), Blake Bortles (12), Brian Hoyer (11), and Andrew Luck (10) all ranked at least 10 spots better in DVOA with pressure than DVOA without it. Winston in particular had a very interesting season under pressure, which we will have to break down in further detail at a later date. For now, let's just say that ESPN's QBR liked his season under pressure a bit more than DVOA, but no system is really built to account for plays like this.
With 127 knockdowns, Carson Palmer gingerly took that crown from Winston (third with 120). He also held off another painful season for Andrew Luck (126), who has come frighteningly close to leading the league in knockdowns five years in a row. Injury, which is almost inevitable given all of this contact, is the only reason Luck did not achieve perhaps the most undesirable of quarterback records. Luck led the NFL in knockdowns in 2012, 2013, and 2014, but only had 58 knockdowns in seven games in 2015 before a lacerated kidney ended his season. His 16-game pace was 132.5 knockdowns, which would have beaten Winston (123) for the lead. In 2016, Luck missed one game against Pittsburgh, and backup Scott Tolzien suffered 12 knockdowns in his place for the Colts. Needless to say, Luck almost certainly would have been hit multiple times against the Steelers, giving him another black-and-blue award. Luck (20.1 percent) was one of three 2016 quarterbacks with a knockdown rate above 20.0 percent. The other two were rookies: Jared Goff (21.4 percent) and Cody Kessler (23.6 percent).
It is time that we talked about Goff, a true Hollywood horror film. With or without pressure, his DVOA was the pits of 2016. Now the fact that he had the highest pressure rate of any quarterback since 2010 (40.4 percent) might draw him some sympathy, but let's consider the fact that teammate Case Keenum's rate was only 21.2 percent, ranked sixth in the league. The Rams had four offensive linemen play at least 893 snaps last year, so there was not a massive decline in talent and continuity when Goff took over. Defenses blitzed Goff more than Keenum, but there was not a huge difference. It is more likely that the rookie just got a bit confused at times, mixing up where the next pass-rusher was coming from with where the sun was going to set.
Before we bring the sun down on Goff, we do have to reiterate that the No. 1 overall pick was just a rookie, and even Terry Bradshaw and Troy Aikman once recovered from horrific rookie seasons to make the Hall of Fame. The Rams are likely not building a dynasty beneath our noses, but it is worth mentioning again: you can't fully dismiss a player after one struggling season.
However, one of the absolute best stats in judging whether or not a quarterback can cut it in this league is DVOA without pressure. Some crazy results can happen under pressure, but you expect the quarterback to do well when given a clean pocket from which to operate. This is where Goff's season -- and granted, it was seven starts on an offense that has been barren for years -- gets really scary. Since 2010, no quarterback had a worse DVOA without pressure than Goff's -45.2%. In fact, it was not even close.
|Lowest DVOA without Pressure Since 2010 (Min. 200 Passes)|
|Rk||Player||Year||Team||Pressure Rate||DVOA no Pressure||Note|
|1||Jared Goff||2016||LARM||40.4%||-45.2%||Entering Year 2|
|2||Brady Quinn||2012||KC||23.6%||-6.7%||Never played another game|
|3||Blaine Gabbert||2011||JAC||26.4%||-4.5%||26 starts since (bust)|
|4||Christian Ponder||2011||MIN||26.4%||-1.6%||26 starts in 2012-14 (none since)|
|5||John Skelton||2012||ARI||20.2%||-0.7%||Never played after 2012|
|6||Jimmy Clausen||2010||CAR||29.5%||0.6%||Four spot starts since 2011|
|7||Derek Anderson||2010||ARI||25.3%||0.9%||Four spot starts since 2011|
|8||Matt Cassel||2012||KC||18.4%||1.4%||Left for MIN in 2013|
|9||Curtis Painter||2011||IND||19.6%||1.7%||Never started another game|
|10||Peyton Manning||2015||DEN||24.3%||5.5%||Retired after the season|
|11||Blake Bortles||2014||JAC||24.9%||6.5%||Still trying to be QB1|
|12||Terrelle Pryor||2013||OAK||34.3%||7.3%||Converted to WR|
|13||Matt Cassel||2013||MIN||20.5%||7.9%||12 spot starts since 2014|
|14||John Skelton||2011||ARI||23.4%||8.1%||Never played after 2012|
|15||Mark Sanchez||2012||NYJ||18.6%||9.9%||10 spot starts since 2013|
This is really discouraging. Peyton Manning knew it was time to retire after 2015. Curtis Painter (2011 Colts) and Jimmy Clausen (2010 Panthers) were so bad that their teams were able to draft Andrew Luck and Cam Newton with No. 1 overall picks. Matt Cassel and Mark Sanchez will only get work this fall if Mike Glennon or Marcus Mariota are injured. Terrelle Pryor ended up converting to a wide receiver, and the Cardinals had a hell of a time bridging from Kurt Warner's retirement to the acquisition of Carson Palmer.
[ad placeholder 4]
Christian Ponder and Blaine Gabbert may have started more than a dozen games each after their 2011 rookie seasons, but the Vikings and Jaguars moved on to draft Teddy Bridgewater and Blake Bortles in 2014. Bortles might be on his last legs in Jacksonville too.
If we dropped the requirement to 100 passes, that would bring up a holy trio of awfulness in Ryan Lindley (2012 Cardinals), Tyler Palko (2011 Chiefs), and Caleb Hanie (2011 Bears). It also includes Kirk Cousins in his 2013 season, and he has at least made something of himself in these last two years for Washington. It is worth mentioning that new Rams coach Sean McVay took over as Cousins' offensive coordinator in 2014. However, one name out of two dozen is a major problem. If a quarterback can't reach at least 10.0% DVOA without pressure, he just might not be capable of starting in the NFL. Ambiguous Hollywood ending: Goff has a long way to go.
For those curious about some of 2016's non-qualified quarterbacks, Denver fans may be interested to know that Paxton Lynch had a huge split of -166.8% DVOA with pressure and 57.2% DVOA without pressure. He'll battle Trevor Siemian for that starting job this summer. The Jets' latest quarterback, Josh McCown, is still the only qualified quarterback with a positive DVOA with pressure (8.0% for the 2013 Bears) in a season since 2010, but his -86.7% DVOA in Cleveland last year would have ranked 28th. Bryce Petty is likely not the answer for the 2017 Jets either -- his DVOA without pressure (-40.1%) was in Goff territory. Meanwhile, some AFC East backup quarterbacks did well, albeit on small sample sizes. Jimmy Garoppolo (7.0% DVOA on 19 pressure plays) and Matt Moore (13.7% DVOA on 31 pressure plays) each posted a positive DVOA under pressure. Garoppolo also had a Brady-like 84.3% DVOA when not pressured, though we only saw him in action against the Cardinals and Dolphins.
The Moore numbers are interesting given that Ryan Tannehill had the third-lowest DVOA when pressured (-108.8%) and the largest drop-off in DVOA from non-pressured plays to pressured ones (-163.7%). This has been a problem throughout Tannehill's five-year career. He has never ranked higher than 25th in the last five seasons in DVOA when pressured. While a quarterback like Brock Osweiler ranked 31st in DVOA whether he was pressured or not, Tannehill finished No. 8 without pressure, so he is a quarterback capable of doing damage with a clean pocket. He made some jaw-dropping throws in San Diego under pressure last season, but to say that he has been consistent at doing so in his career would simply not be true.
Tannehill is one of 18 quarterbacks to qualify for our pressure stats in each of the last five seasons. The following table shows where those 18 quarterbacks rank amongst each other in average pressure rate, DVOA with and without pressure, and pressure drop since 2012. These are not averages weighted by attempts.
|Quarterbacks: Pressure Splits (2012-2016)|
|Quarterback||Pressure Rate||Rk||DVOA w/Pressure||Rk||DVOA no Pressure||Rk||DVOA Difference||Rk|
It is not surprising that Tannehill has the lowest DVOA when pressured, and the largest drop. Andy Dalton has the lowest rate of pressure, but hasn't handled things too well when the heat has gotten there in Cincinnati.
This period encompasses all of Ben Roethlisberger's time in Pittsburgh with Todd Haley as his offensive coordinator. The changes from Bruce Arians' vertical days to a quicker passing game, along with an improved offensive line (talent and Mike Munchak's coaching) have helped Roethlisberger to the second-lowest pressure rate. He's still good when things get chaotic, as his DVOA with pressure (-36.7%) and pressure drop (-88.9%) are the best here. It is also not surprising to see that Russell Wilson and Aaron Rodgers handle pressure very well by being creative.
[ad placeholder 3]
There's a good argument that Russell Wilson, since he entered the league in 2012, has the best overall stats of all these quarterbacks. His DVOA with pressure is second only to Roethlisberger, and his DVOA without pressure is the best in the league, even surpassing Brady's. The problem is that Wilson is easily the most-pressured quarterback in the league, facing it on over a third of his plays for the Seahawks. Wilson has never finished higher than third in DVOA in a season, but if his distribution of pressure to non-pressure plays was more like the league average, then his overall stats would certainly improve. If the Steelers were able to improve their pass-protection issues over time, then there is still hope that Wilson will one day have an adequate line in front of him too. He'll just likely be in his early-to-mid 30s, and one of the most veteran quarterbacks in the league as time winds down on this golden age of passers.
Later this week, we will look at the success of pressure by defenses in 2016.
23 comments, Last at 23 Jun 2017, 4:48pm
#1 by justanothersteve // Jun 20, 2017 - 7:18pm
Really good article but I don't see the point of the first table. Why is the table organized by pressures of individual QBs? If you are judging the line, shouldn't this be a team stat? The QB multiple-season table is much better as it emphasizes the difference between pressure vs no pressure.
#2 by Jimmy Oz // Jun 20, 2017 - 8:32pm
It would be good to see this with splits on the number of rushers.
#3 by galactic_dev // Jun 20, 2017 - 11:06pm
Better Jameis Winston highlight video link:
#4 by Karl Cuba // Jun 21, 2017 - 4:56am
Ah ha! Trick question, the pressure doesn't arrive at the quarterback, the pressure stays still and the quarterback rotates into it!
#10 by Chuckc // Jun 21, 2017 - 10:56am
You've been watching Kaepernick too much
#12 by dbostedo // Jun 21, 2017 - 1:18pm
#17 by Chuckc // Jun 21, 2017 - 7:35pm
The difference being, you can't watch too much Futurama
#5 by Rocco // Jun 21, 2017 - 9:28am
Every article that mentions Goff is more depressing and bleaker than the last one.
#6 by Aaron Brooks G… // Jun 21, 2017 - 9:49am
Couple of questions:
There are two kinds of pressure -- line busts and unsafe blitzing. What I would expect a QB's DVOA to be in response is very different for the two kinds.
This yields four possible outcomes:
Intentional absence of pressure --
In the case of Hoyer and Bortles, was it just that defenses were sitting back, because even without pressure, they weren't great QBs (or were in bad offenses)?
Unintentional presence of pressure --
Contra Taylor and Wilson, who clearly played behind poor pass-blocking lines.
Intentional presence of pressure --
Winston and Newton. They cratered when under heavy pressure, but you got killed if it didn't arrive.
Unintentional absence of pressure --
Carr and Dalton. Teams were trying to get there, but couldn't. Basically Winston and Newton, but with better lines.
And our wildcard: ¯\_(ツ)_/¯ --
Aaron Rodgers. It didn't really matter what you did.
Do you have any sense for how defenses treated these guys? We track pressure, but not whether it's a consequence of defensive tactics versus linemen incompetence. Or in the case of Roethlisberger and Wilson, sometimes just serial dilly-dallying.
#11 by BigRichie // Jun 21, 2017 - 12:11pm
As a Washingtonian who sees plenty of Wilson, I've thought his problem was the opposite. At a certain time point, he just takes off. His eyes go down (they do go back up if he breaks contain) and he looks for a running alley. I've seen him try to bolt from many clean pockets, so creating 'pressures'.
As a former Wisconsinite, I see almost just as much of Rodgers. He was amazing starting right around Halloween of last year. Have never seen the like. Sometimes he'd bolt quickly to the left or right, always successfully, having had his eyes downfield yet knowing that it had sprung wide open for him to roll thataway. (He has said he 'scouts' the opposing rush) Other times he sat and sat and sat in the pocket, with half the rushers now behind him so he just had no way of knowing where they were, yet trusting that his lineman had sealed the entire circle around him. Have never enjoyed watching a QB as much as I did Rodgers the last 10 weeks of '16-'17.
#14 by Vincent Verhei // Jun 21, 2017 - 3:59pm
Do you have any sense for how defenses treated these guys? We track pressure, but not whether it's a consequence of defensive tactics versus linemen incompetence.
We actually do have data that breaks down pressure rates and yards per play by number of pass-rushers. We just don't have time to publish all our data in one article. Some of that data will be in FOA2017. Maybe will have time to post more here later this summer after the book is done.
To answer questions others have asked, we do not have data on how much time quarterbacks spend in the pocket or hang on to the ball.
#7 by Hoodie_Sleeves // Jun 21, 2017 - 9:53am
Does the SIS data that this is based on do anything to quantify whether the pressure is due to pass rush, or due to the quarterback simply taking too long to do things?
A lot of the guys with decent "under pressure" DVOAs also seem to have high pressure rates, which makes me think they're guys who are slow to throw the ball away (and that certainly fits with Roethlisberger and Wilson)
#8 by Noahrk // Jun 21, 2017 - 10:42am
This analysis looks like it could help bridge the divide between Tannehill supporters and haters. Makes sense to me.
#9 by Raiderjoe // Jun 21, 2017 - 10:46am
will reads if have time alter tongith. may be notable interets
#13 by dbostedo // Jun 21, 2017 - 1:20pm
Is it possible to align this data with how long the ball was held on each of the plays with pressure?
#15 by Scott Kacsmar // Jun 21, 2017 - 4:54pm
There's data on that, but we're not in a position to publish it here.
#16 by fb29 // Jun 21, 2017 - 5:38pm
#18 by Theo // Jun 22, 2017 - 11:46am
"Before we bring the sun down on Goff, we do have to reiterate that the No. 1 overall pick was just a rookie, and even Terry Bradshaw and Troy Aikman once recovered from horrific rookie seasons to make the Hall of Fame. The Rams are likely not building a dynasty beneath our noses, but it is worth mentioning again: you can't fully dismiss a player after one struggling season."
So you're saying there is a change?
#19 by justanothersteve // Jun 22, 2017 - 12:10pm
I'm not sure I get that either. I was thinking maybe Goff can recover from a horrific rookie season. I no longer watch the Rams because they've (thankfully) left St Louis, so I didn't realize just how bad his playing was. But when he sucks without any pressure whatsoever, it's hard to be optimistic. I could suck just as bad as Goff for less money, though I don't think my 61-year-old, short, fat body could take the beating.
#20 by Theo // Jun 22, 2017 - 12:33pm
I misspelled "chance".
#21 by Scott Kacsmar // Jun 22, 2017 - 5:28pm
I just get the sense that if we had charting data for Bradshaw or Aikman, then we'd see them have horrible DVOA without pressure too as rookies. So a massive reversal has been done before, but yeah, we might be in the "1 in 500 chance" territory here that Goff recovers.
#22 by jtr // Jun 23, 2017 - 10:04am
I wonder if Tannehill's lousy pressure numbers have to do with the fact that he spent most of his college career playing WR instead of QB. I would think that learning to react to pressure would be the QB skill that would require the most game reps. Non-pressure skills like reading defenses or making accurate throws can be practiced in the film room and practice field. But there's really no way to simulate trying to make a play while several giant men try their hardest to knock the snot out of you.
#23 by Vincent Verhei // Jun 23, 2017 - 4:48pm
I was patient with Tannehill coming out of college for that very reason, but now he's a 29-year-old five-year pro with 77 starts. I think we can stop using inexperience as an excuse for him now.