How L.A. Rams Carried Matthew Stafford to Super Bowl Title

Los Angeles Rams QB Matthew Stafford
Los Angeles Rams QB Matthew Stafford
Photo: USA Today Sports Images

NFL Preseason Week 2 - Guest column by Ajit Kirpekar

The 2022 NFL season is upon us. As such, let us celebrate with yet another trip down recent memory lane as we visit the 2021 Rate of Adjusted Pressure Numbers, or RAP for short. These numbers show Matthew Stafford enjoyed the best pass protection of his career en route to a Super Bowl championship with the Los Angeles Rams.

Over the last couple of weeks, Football Outsiders has run basic pressure rates from 2021. Now we're going to make things a bit more complicated. RAP is an opponent-adjusted metric for measuring a team's success at generating or avoiding pressure. It's calculated using a machine learning model that first predicts the probability of pressure from a set of variables, and then it is re-rerun with the opponent-specific inputs set to average, with the difference between the two accounting for the opponent. For a more detailed explanation, see this link. And for last year's article, see here. Reminder, all data comes from Football Outsiders' game charting data (collected first by FO staff and volunteers, then in conjunction with ESPN Stats & Info, and since 2015 by Sports Info Solutions).

Every year the model goes through an additional iteration, which inevitably shifts the all-time rankings up and down some spots as new information is added. Last year, the big addition was incorporating Estimated Quarterback Time to Throw into the picture. This year, I added offensive line continuity as a variable. While Football Outsiders has its own offensive line continuity metric, it is broken down by year whereas I needed something that would be available at the weekly level. As such, I came up with my own metric, which basically measures how many consecutive weeks a team trots out the same five starting offensive linemen. In addition, I had a second metric to capture how many new members of the offensive line would be starting a given week compared to the previous week. Not only did these new additions impact the model, they were also part of the opponent adjustments for defenses.

Next, I like to examine major league changes that occurred during the season that might materially affect the RAP standings for the corresponding year. In 2020, the big one was Tom Brady finally departing New England for Tampa Bay and being replaced by Cam Newton. Last year, the big one involved the Super Bowl champions, as they underwent their own bit of quarterback plastic surgery.

Two Sides of the Stafford-Goff Trade
Team RAP Pressure
Rate
Rank
LAR 18.2% 18.4% 3
DET 23.8% 25.0% 10

There wasn't much to show last season for Detroit, but a (just barely) top-10 finish is nothing to sneeze at. Let's take a closer look at the two quarterbacks. The offensive line continuity variable here is better as it gets higher. Win probability, as listed, represents the average win probability across all of that quarterback's passing plays.

Matthew Stafford Pressure Rate by Season
Year Team Pressure
Rate
Estimated
Time To Throw
O-Line
Continuity
Win
Probability
2009 DET 29.6% 2.55 0.6 37.5%
2010* DET 24.4% 2.52 5.6 54.9%
2011 DET 25.1% 2.51 6.1 45.3%
2012 DET 18.7% 2.51 11.7 42.1%
2013 DET 20.7% 2.51 1.9 51.5%
2014 DET 24.7% 2.52 0.5 49.3%
2015 DET 24.6% 2.53 0.7 46.4%
2016 DET 28.8% 2.53 1.1 48.3%
2017 DET 32.1% 2.54 0.2 49.0%
2018 DET 30.2% 2.54 0.9 43.2%
2019* DET 29.3% 2.55 1.2 51.6%
2020 DET 20.8% 2.56 0.4 42.0%
2021 LAR 18.9% 2.57 1.7 50.0%
* injured
Jared Goff Pressure Rate by Season
Year Team Pressure
Rate
Estimated
Time to Throw
O-Line
Continuity
Win
Probability
2016 LAR 42.5% 2.67 5.1 36.2%
2017 LAR 32.4% 2.76 0.0 56.3%
2018 LAR 28.1% 2.79 0.0 55.3%
2019 LAR 32.0% 2.80 3.2 48.2%
2020 LAR 23.5% 2.80 1.3 50.5%
2021 DET 27.3% 2.80 2.3 37.6%

Squint hard enough and you can kind of see the similarities between the two quarterbacks beyond the fact that they are both former No. 1 overall picks. Both have generally had stable time-to-throw metrics and low offensive line continuity, with a one-year outlier for Stafford (not coincidentally timed with his best season).

For Stafford in particular, this past season was metaphorically going from the poor house to the penthouse, and his RAP numbers reflect the same. However, his career to this point was a bit of a surprise. Despite the reputation as a traditional pocket passer with a strong arm and a quick release, his overall pressure rate has been strange. After several years of positive trends, Stafford suddenly reversed course in 2016 and stayed that way until his last year in Detroit. I'll leave it to Lions fans to provide some perspective on why this might be.

As for Stafford's unfortunate counterpart, Jared Goff went from the ideal-if-awkward to the modern day NFL hellscape. Of course, Goff isn't blameless, but a look over his pressure percentage shows at least a positive overall trend. And although this past season's disaster was reminiscent of his ghoulish rookie season, the RAP numbers clearly indicate he has improved since then.

RAP by Offense

Top Offenses by RAP, 2021
Rank Offense RAP Pressure
Rate
Difference
1 PIT 14.7% 16.9% +2.2%
2 TB 15.7% 16.0% +0.3%
3 LAR 18.2% 18.4% +0.2%
4 GB 21.7% 21.4% -0.3%
5 LAC 22.3% 23.9% +1.6%
6 NE 22.7% 22.0% -0.7%
7 ARI 23.3% 25.4% +2.1%
8 CLE 23.3% 25.1% +1.8%
9 DAL 23.4% 23.2% -0.2%
10 DET 23.8% 25.0% +1.2%
11 TEN 24.2% 24.4% +0.2%
12 WAS 24.9% 26.5% +1.6%
13 BAL 25.3% 25.0% -0.3%
14 CIN 25.7% 27.0% +1.3%
15 LV 25.7% 27.3% +1.6%
16 NYG 26.2% 25.8% -0.4%
17 ATL 26.5% 30.1% +3.6%
18 SF 27.0% 24.5% -2.5%
19 CHI 27.2% 27.8% +0.6%
20 JAX 27.3% 26.1% -1.2%
21 IND 27.6% 28.1% +0.5%
22 MIN 27.7% 27.8% +0.1%
23 DEN 27.8% 27.4% -0.4%
24 KC 28.5% 25.3% -3.2%
25 MIA 29.0% 28.3% -0.7%
26 CAR 29.5% 28.8% -0.7%
27 PHI 29.8% 27.3% -2.5%
28 BUF 30.0% 27.8% -2.2%
29 NYJ 30.1% 30.5% -0.4%
30 NO 30.4% 29.5% -0.9%
31 SEA 32.2% 31.7% -0.5%
32 HOU 32.9% 31.2% -1.7%

Four teams finished in the top 10 in 2020 and then repeated the feat in 2021, with Pittsburgh and Tampa Bay once again finishing first and second respectively.

With the newest iteration of the model, Pittsburgh now has three teams finishing in the top 10 in the all-time offensive RAP rankings, with Ben Roethlisberger now in a tie with the great Peyton Manning for the most top-10 finishes. Curiously, Tampa Bay's repeat strong finish means the 2020 and 2021 teams also join the top 10 all-time, giving Brady two appearances with a chance to tie his great rivals this upcoming season.

Joining Pittsburgh and Tampa Bay as a repeat finisher in the top 10 is Green Bay, a team helmed by another NFL greybeard. Much like Stafford, Aaron Rodgers' pressure rate has bounced around a lot during his career, with steady improvement early in his career before entering a mid-career slump with pressure rates into the mid-30% territory before a sharp pullback in his last two MVP campaigns.

At the bottom of the list are a pair of teams that are no strangers to last-place finishes. In fact, if you have followed these articles since the beginning, I have addressed the same two teams every single time. I am referring, of course, to the Seattle Seahawks and the Houston Texans. It is noteworthy, however, that this year both teams were missing their usual signalcallers and yet still ended up here. For Seattle, the problem wasn't just with Russell Wilson but with Geno Smith, whose poor pressure rates have dogged him going back to his days with the Jets. For Houston, David Mills was actually OK for a rookie, but it was Tyrod Taylor's stint at quarterback that was the biggest problem, with a 40% pressure rate.

As for the rest of the bottom, I want to focus on Buffalo and Kansas City. This year marked their best finishes in this stat since Patrick Mahomes and Josh Allen have been full-time starters. In fact, a look over their statistics shows a similar pattern. For starters, they share the distinction, along with Russell Wilson, for being the only quarterbacks who have a career pressure rate above 30% while also having a career win probability above 50%.

Quarterbacks, 30% Pressure Rates, 50% Win Probability
QB Team Estimated
Time to Throw
Win
Probability
Pressure
Rate
Josh Allen BUF 2.95 53.1% 31.7%
Patrick Mahomes KC 2.85 56.1% 31.6%
Russell Wilson SEA 2.97 52.7% 38.6%

Looking over the two quarterbacks:

Patrick Mahomes Pressure Rate by Season
Year Estimated
Time to Throw
Win
Probability
Pressure
Rate
2018 2.86 56.7% 36.2%
2019 2.86 57.8% 32.5%
2020 2.85 56.9% 29.6%
2021 2.84 53.7% 26.7%
Josh Allen Pressure Rate by Season
Year Estimated
Time to Throw
Win
Probability
Pressure
Rate
2018 2.95 42.9% 37.3%
2019 3.02 51.7% 31.5%
2020 2.95 57.3% 29.8%
2021 2.90 56.2% 29.0%

The estimated time to throw numbers suggest that Mahomes' and Allen's playing styles are not only similar, but have stayed relatively constant over the years. The improvement in pressure rate suggests they have gotten better over time at mitigating pressure. It will be interesting to see how the rest of their careers play out.

RAP by Defense

Top Defenses by RAP, 2021
Rank Defense RAP Pressure
Rate
Difference
1 NE 32.5% 28.6% +3.9%
2 TB 31.9% 28.5% +3.4%
3 MIA 31.7% 33.2% -1.5%
4 KC 31.4% 28.3% +3.1%
5 HOU 29.9% 24.2% +5.7%
6 CAR 29.8% 28.0% +1.8%
7 LAR 29.6% 28.8% +0.8%
8 LV 27.7% 26.3% +1.4%
9 SF 27.3% 27.2% +0.1%
10 BAL 27.2% 25.2% +2.0%
11 TEN 27.0% 27.1% -0.1%
12 DAL 27.0% 26.8% +0.2%
13 DEN 26.6% 24.9% +1.7%
14 PIT 26.5% 25.5% +1.0%
15 BUF 26.5% 31.7% -5.2%
16 LAC 26.3% 25.9% +0.4%
17 JAX 26.2% 24.1% +2.1%
18 NYJ 26.1% 22.3% +3.8%
19 CHI 25.4% 26.4% -1.0%
20 GB 25.1% 26.6% -1.5%
21 IND 23.9% 25.8% -1.9%
22 NYG 23.6% 23.3% +0.3%
23 PHI 23.5% 23.1% +0.4%
24 WAS 23.4% 22.9% +0.5%
25 CLE 23.0% 24.4% -1.4%
26 NO 21.9% 24.8% -2.9%
27 ARI 21.3% 23.1% -1.8%
28 DET 20.8% 22.9% -2.1%
29 SEA 20.7% 24.9% -4.2%
30 CIN 19.8% 24.0% -4.2%
31 MIN 19.7% 24.1% -4.4%
32 ATL 17.1% 17.8% -0.7%

The 2021 winner in Defensive RAP is New England, the first time the Patriots have ever finished atop these rankings. I have in the past commented about the uniqueness of New England and how seemingly indifferent they have been with respect to pass rush despite all of their success. As it turns out, that's not quite the case, as a look over their history since 2006 paints a rather interesting picture.

New England Defense RAP, 2006-2021
Year RAP Pressure Rate
2006 33.6% 34.5%
2007 25.3% 32.2%
2008 32.8% 31.8%
2009 20.0% 23.7%
2010 24.9% 24.0%
2011 22.8% 23.2%
2012 22.3% 24.7%
2013 24.7% 23.9%
2014 20.3% 23.4%
2015 25.7% 24.8%
2016 20.7% 22.7%
2017 22.2% 23.4%
2018 22.7% 26.9%
2019 29.7% 30.3%
2020 29.2% 29.7%
2021 32.5% 28.6%

The 2006 to 2008 seasons were sort of the last vestiges of the original dynasty with 2009 a clear point of demarcation. From there, New England entered its pass-rushing fallow period, consistently hovering below average throughout their second dynasty. Lately, however, there has clearly been an uptick which began in 2019 with their infamous larded schedule of Smurfs and continued to the present. It even persisted amid all of the COVID defections of 2020. I'll leave it to Pats fans to comment about why this might be.

The Patriots are joined by a number of other teams that repeated their strong finish atop these rankings from the year prior, including the Buccaneers, Chiefs, Dolphins, and Rams. I have extolled the Rams consistency in the past so I won't belabor that here. I do want to highlight the Dolphins, who climbed out of the proverbial RAP grave in 2019 and have now posted two straight strong years as a pass rush. Whatever you want to say about Brian Flores' tenure in Miami, the results on defense were clearly positive. After finishing dead last in 2019, in the last two years the Dolphins have ranked seventh and now third.

Tampa Bay's coaching staff also deserves praise for a strong two-year performance, ranking first in 2020 and second this past year. And of course, Kansas City's defense has quietly been an impressive pass rushing unit the last few years as well. Given the rough start, I was curious to see if their miraculous second-half turnaround was in part powered by a strong uptick in pass rush, but it turns out that they posted fairly even numbers in the first half and the second half of the season.

On the opposite end of the spectrum, it was a miserable year for Atlanta. This year's Football Outsiders Almanac (now available!) noted that despite finishing with a respectable win-loss record, this was an abominable year for the Falcons and it certainly showed up in the RAP numbers.

As for Seattle, their history during the Wilson era oddly mirrors that of New England.

Seattle Defense RAP, 2006-2021
Year RAP Pressure Rate
2013 39.1% 35.8%
2014 24.3% 26.0%
2015 30.0% 28.6%
2016 28.6% 27.0%
2017 31.7% 26.6%
2018 24.5% 25.0%
2019 19.8% 20.2%
2020 23.6% 24.5%
2021 20.8% 24.9%

After winning the Super Bowl in part powered by an all-time great defense, Seattle's core defense has slowly aged until the wheels truly came off in 2018 and the pass rush has limped along since then. With Wilson officially departed, the last vestiges of the great DVOA darling have come to an end.

Ajit Kirpekar is a data scientist based out of San Francisco, California. Despite this, he roots for the Indianapolis Colts.

Comments

14 comments, Last at 22 Aug 2022, 4:58pm

1 Outhouse to Penthouse seems…

Outhouse to Penthouse seems to be overstating things, as Stafford went from the equivalent of 4th in RAP to 3rd. 
The Goff comps and the Mahomes/Allen discussion strongly suggest pressure is mostly QB with a component of line and receiving corps quality.

As for 2016, Calvin Johnson retired at the end of 2015, the line got younger, and the bottom fell out of the receiving corps.

5 I may end up having to add…

I may end up having to add some variable to measure receiver quality. Not sure exactly how yet.

What's odd is I thought under Jim Bob Cooper, the Lions were throwing short a lot which would depress his pressure numbers.

2 Wanted to quickly address a…

Wanted to quickly address a point Noah brought up in the pressure articles. Miami's Defensive pressure rate in the first half of the season was a healthy 30%. The second half it surged to a ridiculous 38%.

3 the lack of correlation of…

the lack of correlation of defensive RAP to total sacks is astounding to me. At least at first perusal. The second worst defensive RAP team, Minnesota, was second in 2021 sacks. Pitt and Chi are middling by the RAP measure. whereas New England was well below average in actual sacks.

Certainly there's something to be said about having the guys who can actually "get home" I guess. Might some QBs have a tendency to hold the ball longer against the teams with guys who don't get home, and thus inflating their "pressure" rates? Whereas is Robert Quinn or TJ Watt are lurking, a swim move away from a strip sack, the internal clock ticks faster and the pass goes quicker even on those occasions they're not getting through? [and then there's atlanta, who should be deweighted, as they may yield a significant correlation just due to the heavy point on the far lower left...]

4 Total sucks is a cumulative…

Total sacks is a cumulative stat, while RAP is a rate stat. Usually, the two rate stats are well correlates I think? I haven't run the correlations, but in the past when looking at things it's usually been a 1 to 3 ratio. Meaning, you get 1 sack for every three pressures.

There's a ton of debate whether a player who gets a lot of sacks and less overall pressure is just good at converting pressures to sacks or just lucky. I tend to be on the side of he got lucky.

6 I guess the correllative…

I guess the correllative rate stat would be sacks/snaps, or sacks/pass play. So I guess the correlation should be RAP to sacks/pass play. Still, unless Minnesota faced a historically high number of pass attempts, their placement as number 2 in sacks, and 31 in RAP stands out. Maybe there is a correlation, with Minnesota being a major outlier, but there are several others that fall out of alignment as well, such as New England, as I mentioned.

I don't think high rates of conversions to pressure is a matter of luck, at least not purely. For instance, if you're scheming pressure for your DBs and LBs, versus getting pressure from your designated pass rushers, you're less likely to get a pressure to sack conversion (I think, generally) in the former case, both because of skill/talent in actually taking down the QB, the QB being more averse to taking the hits from the big guys so just kind of giving up, hot routes less likely to be open and/or in sight lines, etc.

9 Minnesota was 9th in…

Minnesota was 9th in adjusted sack rate and by raw pressures was somewhere below average. Opponent adjustments hit them harshly, along with the rest of the NFC North. 

New England was actually 14th in adjusted sack rate - so above average and very good just in raw pressure rate. I guess that's one where you can maybe debate if they were just a bit unlucky with sacks.

Your latter comment about DBs makes some sense, although I wonder how much of the total pressure plays are distributed to non-pass rushers. I don't think its that large. 

 

8 I think this is particularly…

I think this is particularly instructive. In season, you'd rather have the cb who intercepts a lot of passes because ints swing games far more than a corner who gives up a million first downs but the defense is able to hold them to a fg in the red zone.

But int rates are super wild and almost no corner has shown the ability to remain a turnover magnet; with the exception of Peanut Tilman who was a savant for punching the ball loose. 

10 Ouch

Next, I like to examine major league changes that occurred during the season that might materially affect the RAP standings for the corresponding year. In 2020, the big one was Tom Brady finally departing New England for Tampa Bay and being replaced by Cam Newton. Last year, the big one involved the Super Bowl champions, as they underwent their own bit of quarterback plastic surgery.

This is gonna be interesting, but not fun to do next year I guess. 

11 This headline is misleading…

This headline is misleading as the Rams seemingly had nothing to do with San Francisco upsetting Green Bay and preventing Stafford from losing at Lambeau. 

12 Should have shown Wilson's career as well

to get an idea of what the future might hold for Mahomes and Allen. They all play the same style of football. It's a style Pete Carroll wants out of his QB. That's why the Seahawks tried to get Allen and wanted Mahomes.

14 Brady too

This same concept can be applied to Brady and the 2020 Buccs as well. 

Save 10%
& Support the Writers
Support Football Outsiders' independent media. Use promo code WRITERS to save 10% on any FO+ membership and give half the cost of your membership to tip the team of writers.