clock menu more-arrow no yes

Filed under:

Does pressuring the quarterback really lead to good defensive outcomes in the NFL?

Are modern NFL QBs becoming “immune” to pressure?

Syndication: PackersNews Wm. Glasheen/USA TODAY NETWORK-W via Imagn Content Services, LLC

There’s an old adage that “it’s hard for a quarterback to complete a pass when lying on his back with a defender on top of him.”

This would suggest that the more often a defense hits and/or sacks a quarterback, the more often a good outcome occurs for the defense. A good outcome would be forcing a throw early that might lead to an incomplete pass or an interception. Of course, a good outcome for the defense would also be a sack.

Sacks are happening less frequently than they ever have, but does that mean that defensive pressure on QBs is becoming less effective?

Three other passer stats that are trending up while sack rate is trending down are passer rating, completion percentage and TD:INT ratio.

To answer the titular question, I used pro-football-reference.com to look at defensive data. They have “QB hits” data per team going back to 2006. So I looked at what I call a “Watt” (which is a QBhit or a sack named for JJ Watt, who has been the best individual defender at this most years he has been in the league).

A QBhit is a play on which the defensive player hits the QB legally after he has thrown the ball or while he was throwing the ball. This is different from a pressure where the proximity of the defender to the QB makes the QB rush or alter his throw to avoid a QBhit or a sack. On a pressure the QB is not touched by the defender.

I looked at team “Watt percentage,” which is the rate at which a defense hits or sacks an opposing QB (Watts/dropbacks), and then I ran a correlation with passer rating allowed.

Theoretically there should be a negative correlation - as Watt percentage goes up, passer rating allowed goes down. The correlation was negative for every year from which I have data, but some years the correlation was much stronger than others.

For those who aren’t aware, a correlation of 40% or greater in sports statistics is a very strong indication that A and B are linked, but as any actuary or statistician will tell - correlation is not causation. In other words, just because Watt percentage is correlated to lower passer ratings allowed, does not mean that hitting or sacking the quarterback leads to lower passer ratings.

One explanation for this is that sacks do not factor into the formula for passer rating. That formula looks at completion percentage, yards per attempt, touchdown percentage and interception percentage. In a counterintuitive way, a QB who takes many sacks but still moves the sticks would actually have a higher passer rating than a QB who avoids sacks but doesn’t convert on long yardage downs.

Some QBs are better at throwing under pressure than others (see below) and there were four QBs who got pressured often (on 25% or more of their dropbacks), but had less than 16% of their throws counted as “bad throws.” Those four are Deshaun Watson, Russell Wilson, Kirk Cousins and Nick Mullens.

You might note the positions of Drew Lock, Teddy Bridgewater and Aaron Rodgers on the plot above - but let’s not veer off into that discussion here (we can save that for the comments if you wish).

I’m going to start by focusing on the last year when there was almost no correlation between Watt percentage and passer rating allowed - 2014.

From a league-wide view, the team that led the league in Watt percentage that season - the Jets at 26.9% - had one of the lower league-leading values of any team over the 15 seasons for which I have data. Another thing that should stick out to you about 2014 is that the team with the lowest passer rating allowed (i.e. the best pass defense) had a fairly high value of 74.1 relative to the other league-leading values over the 15 seasons. That team was the 2014 Browns.

2014 - the worst correlation year

2014 also happens to be the last time the the Broncos had a good passing offense. The 2014 Broncos faced the Jets, but not the Browns. In that game the Broncos won 31-17, but the contest was much closer than that score might indicate. The Broncos jumped out to 24-7 lead, only to have the Jets pull within seven in the fourth quarter. The game was sealed when Aqib Talib picked off Geno Smith in desperation time and took it back 22 yards for a touchdown.

Peyton Manning was hit a total of five times (two sacks) on 35 dropbacks. So the Watt percentage for the Jets this game was 14.7%, which was significantly lower than for most of their games in 2014. In case you forgot, Manning was really good at getting rid of the ball before he got touched by a defender.

We’ll focus on the 2014 Jets before turning our attention to the 2014 Browns. After that we’ll look at some teams on the other end of the spectrum to see if they might have skewed the data and “messed up” the correlation in 2014.

The 2014 Jets had 153 Watts which was fifth in the league behind the Texans (154), Lions (161), Bills (165) and Ravens (167). While the Jets were fifth in the league at hitting opposing QBs, they were one of the worst teams in the league at stopping the pass (from passer rating allowed). They allowed a passer rating of 101.5 which was 30th. League average that season was 88.9. They finished the season allowing 31 TD passes while only getting six interceptions. Interestingly they were almost exactly average in terms of passing yards allowed; they allowed 3746 and league average was 3789.

By passer rating allowed, the Browns were best in the league in 2014 at 74.1. Only two of the teams who were top five in Watt percentage were also in the top five in passer rating allowed - the Bills (2nd - 74.5) and the Texans (5th - 80.4). The other three teams who were top five in Watt percentage in 2014 came in at 9th (DET), 19th (BAL) and 30th (NYJ). So that would explain the lack of correlation between Watt percentage and passer rating allowed in 2014; two teams who hit and sacked opposing QBs regularly, the Ravens and the Jets, were both fairly bad at stopping the pass.

Looking at it from the other side, the other three 2014 teams who were in the top five in passer rating allowed had these ranks in Watt percentage: San Francisco (21st), Cleveland (28th) and Cincinnati (29th). Two of the best teams in the league at stopping the pass, were terrible at hitting and sacking the QB in 2014. This also explains the poor correlation for 2014.

Without doing a deep dive on 2009 and 2010, I would wager that the situation was the same. There were some teams that were good at hitting/sacking the QB who were bad at stopping the pass and there were teams who were good at stopping the pass who were bad at hitting and sacking the opposing QB.

2020 - the best correlation year

As it happens, the most recent NFL season was the one in which we saw the greatest correlation between hitting/sacking the QB and good outcomes for the defense on passing plays.

Negative 62% is very strong correlation in sports statistics. Let’s dig into why there was such a strong negative correlation in 2020. Keep in mind that the strong negative correlation in 2020 means that last season teams that hit and sacked the opposing QB at a higher rate also generally allowed lower passer ratings.

The top five teams in Watt percentage in 2020 were the Steelers, Eagles, Saints, Rams and Washington Football Team.

Rk Tm Cmp Att Cmp% Yds TD TD% Int PD Int% AY/A Rate Sk QBHits ANY/A Sk% Drpbks Watts Watt%
1 Pittsburgh Steelers 298 526 56.7 3110 22 4.2 18 84 3.4 5.9 76.7 56 137 4.7 9.6 582 193 33.2%
2 Philadelphia Eagles 364 530 68.7 3798 27 5.1 8 43 1.5 8.1 102.4 49 115 6.9 8.5 579 164 28.3%
3 New Orleans Saints 333 557 59.8 3472 28 5.0 18 84 3.2 6.3 83.3 45 113 5.4 7.5 602 158 26.2%
4 Los Angeles Rams 347 548 63.3 3051 17 3.1 14 68 2.6 5.7 80.4 53 102 4.6 8.8 601 155 25.8%
5 Washington Football Team 330 529 62.4 3068 21 4.0 16 66 3.0 5.8 81.4 47 97 4.8 8.2 576 144 25.0%
6 Tampa Bay Buccaneers 426 617 69.0 3945 29 4.7 15 70 2.4 6.8 94.3 48 116 5.8 7.2 665 164 24.7%
7 Baltimore Ravens 380 596 63.8 3536 22 3.7 10 67 1.7 6.4 87.2 39 116 5.6 6.1 635 155 24.4%
8 Arizona Cardinals 365 570 64.0 3623 26 4.6 11 57 1.9 7.0 91.5 48 100 5.9 7.8 618 148 23.9%
9 New York Giants 388 574 67.6 3807 22 3.8 11 75 1.9 7.1 93.2 40 107 6.1 6.5 614 147 23.9%
10 Miami Dolphins 343 545 62.9 4024 21 3.9 18 76 3.3 7.3 87.0 41 99 6.2 7.0 586 140 23.9%
11 Kansas City Chiefs 349 557 62.7 3779 29 5.2 16 64 2.9 6.9 89.4 32 99 6.2 5.4 589 131 22.2%
12 Denver Broncos 374 567 66.0 3807 21 3.7 10 64 1.8 7.2 92.1 42 93 6.2 6.9 609 135 22.2%
13 Seattle Seahawks 450 674 66.8 4560 23 3.4 14 80 2.1 6.9 90.2 46 109 6.1 6.4 720 155 21.5%
14 Buffalo Bills 369 573 64.4 3726 23 4.0 15 76 2.6 6.5 86.9 38 92 5.7 6.2 611 130 21.3%
15 Chicago Bears 350 547 64.0 3705 28 5.1 10 71 1.8 7.4 94.9 35 88 6.6 6.0 582 123 21.1%
16 Green Bay Packers 346 536 64.6 3539 23 4.3 11 74 2.1 7.0 91.1 41 80 6.1 7.1 577 121 21.0%
17 Indianapolis Colts 369 562 65.7 3866 24 4.3 15 78 2.7 7.0 90.5 40 81 6.1 6.6 602 121 20.1%
18 Houston Texans 377 541 69.7 4104 30 5.5 3 50 0.6 8.9 109.6 34 80 7.9 5.9 575 114 19.8%
19 Los Angeles Chargers 340 543 62.6 3578 29 5.3 12 54 2.2 6.9 91.4 27 85 6.3 4.7 570 112 19.6%
20 Dallas Cowboys 328 513 63.9 3642 34 6.6 10 46 1.9 7.9 100.4 31 74 7.1 5.7 544 105 19.3%
21 Carolina Panthers 398 585 68.0 3825 28 4.8 7 58 1.2 7.3 98.4 29 89 6.6 4.7 614 118 19.2%
22 Atlanta Falcons 425 625 68.0 4697 34 5.4 12 51 1.9 8.1 101.7 29 92 7.4 4.4 654 121 18.5%
23 Cleveland Browns 371 585 63.4 3962 31 5.3 11 74 1.9 7.4 94.8 38 75 6.6 6.1 623 113 18.1%
24 New York Jets 418 605 69.1 4409 34 5.6 10 67 1.7 8.0 103.2 31 84 7.3 4.9 636 115 18.1%
25 San Francisco 49ers 341 537 63.5 3327 25 4.7 12 62 2.2 6.5 88.6 30 72 5.8 5.3 567 102 18.0%
26 Jacksonville Jaguars 370 533 69.4 4231 34 6.4 12 60 2.3 8.4 105.6 18 77 7.9 3.3 551 95 17.2%
27 New England Patriots 333 494 67.4 3557 22 4.5 18 52 3.6 6.8 89.3 24 65 6.2 4.6 518 89 17.2%
28 Minnesota Vikings 358 541 66.2 4141 30 5.5 15 56 2.8 7.7 97.0 23 70 7.2 4.1 564 93 16.5%
29 Las Vegas Raiders 370 580 63.8 4212 28 4.8 10 74 1.7 7.7 95.3 21 77 7.2 3.5 601 98 16.3%
30 Cincinnati Bengals 340 541 62.8 3859 32 5.9 11 80 2.0 7.6 96.1 17 73 7.2 3.0 558 90 16.1%
31 Tennessee Titans 423 630 67.1 4439 36 5.7 15 81 2.4 7.4 97.5 19 72 6.9 2.9 649 91 14.0%
32 Detroit Lions 383 557 68.8 4558 38 6.8 7 51 1.3 9.3 112.4 24 56 8.6 4.1 581 80 13.8%
Avg Team 367.4 563.1 65.2 3842.4 27.2 4.8 12.3 66 2.2 7.2 93.6 35.5 90.2 6.4 5.9 598.6 125.7 21.0%
League Total 11756 18018 65.2 122957 871 4.8 395 2113 2.2 7.2 93.6 1135 2885 6.4 5.9 19153 4020 21.0%

Looking at where those five teams ranked in terms of passer rating allowed, they ranked: Steelers (1st), Eagles (28th), Saints (4th), Rams (2nd) and Washington (3rd). The only team who was good at getting to the opposing QB in 2020 that was not good at stopping the pass was the Eagles. This might have been because their defense had the fewest passes defended (PD) in the league last season. The Saints the Steelers had almost twice as many passes defended as the Eagles.

Now on the other side, here are the top five teams in passer rating allowed in 2020 (with values):

  1. Steelers (76.7)
  2. Rams (80.4)
  3. WFT (81.4)
  4. Saints (83.3)
  5. Bills (86.9)

The only team from the top five in passer rating allowed who was not in the top five in Watt percentage was the Bills. They ranked 14th in Watt percentage.

By the same token, let’s look at the teams who were the five worst at hitting/sacking opposing QBs:

  1. Lions (13.8%)
  2. Titans (14.0%)
  3. Bengals (16.1%)
  4. Raiders (16.3%)
  5. Vikings (16.5%)

We find that these teams were also generally bad at stopping the opposing passing offense as they ranked 32nd, 24th, 22nd, 21st and 23rd respectively. All five teams were in the bottom tier of the league for passer rating allowed.

You might think that these five teams would have drafted pass rushers in the 2021 draft, but none of these five teams used first-round picks on pass rushers. The Lions, Vikings and Raiders all drafted offensive tackles, while the Titans drafted a cornerback, and the Bengals drafted a wide receiver. In fact, if we look at day one and day two picks in 2021 from these teams, we find that only five picks were used on edge defenders (which is who you draft when you want to improve your pass rush). Those five picks were:

  1. Levi Onwuzurike (41st overall) - Lions
  2. Joseph Ossai (69th) - Bengals
  3. Malcolm Koonce (79th) - Raiders
  4. Patrick Jones II (90th) - Vikings
  5. Monty Rice (92nd) - Titans

So while none of the teams used a first-round pick on a pass rusher, they all used a second- or third-round pick on one.

I should note that this was supposed to be a weak draft class for pass rushers. Some drafts see a pass rusher/edge defender taken first overall and a handful taken in the top 10. This year the first edge defender drafted was not taken until the 12th pick, Micah Parsons, but we did see nine players taken in the first round who are (or could be) edge defenders in the NFL.

  1. Micah Parsons (12th) - OLB - Cowboys
  2. Zaven Collins (16th) - DE/OLB - Cardinals
  3. Jaelen Phillips (18th) - DE/OLB - Dolphins
  4. Jamin Davis (19th) - ILB (OLB) - debatable since it’s tough to play OLB in the NFL at 232 lbs - WFT
  5. Kwity Paye (21st) - DE - Colts
  6. Payton Turner (28th) - DE - Saints
  7. Gregory Rousseau (30th) - DE/OLB - Bills
  8. Jayson Oweh (31st) - DE - Ravens
  9. Joe Tryon (32nd) - OLB - Buccaneers

Thank you for getting nearly to the end of this fairly dense article.

I should also note that while I have been looking at QBhits and sacks, the one piece of data that PFR does not have until the 2018 season is QB pressures that defenses have each year. Since they do have that data for the past three seasons I can evaluate the correlation. The correlation between pressure percentage and passer rating allowed for the last three seasons is below:

2018: -33% (stronger than the -19% correlation to Watt percentage)

2019: -37% (weaker than the -49% correlation to Watt percentage)

2020: -60% (ostensibly the same as the -62% correlation to Watt percentage)

It would be interesting to find a site that tracks defensive pressure percentage back to 2006 (or beyond) to see if this same strong correlation between pressuring the QB and good defensive outcomes on passing plays can be found.

Poll

Did you learn anything from this?

This poll is closed

  • 51%
    Yes - thanks, this was informative
    (141 votes)
  • 13%
    Yes - but you could have made it clearer
    (37 votes)
  • 16%
    No - you lost me at correlation
    (46 votes)
  • 10%
    No - TLDR
    (30 votes)
  • 6%
    No - this was terrible and you suck because you threw shade on Drew Lock in this article
    (19 votes)
273 votes total Vote Now