Paul plays analytics referee between Aaron Schatz and Mina Kimes.
Many NFL analysts are predicting that the Green Bay Packers will take a major step forward on defense next season. Among those people is ESPN’s Mina Kimes, who recently tweeted this:
I’m sure this will go well for me on here
But I stand by it! https://t.co/d63QBUTtGJ
— Mina Kimes (@minakimes) June 13, 2022
However, Football Outsiders’ founder Aaron Schatz quickly jumped in to yuck our collective yum, analytically speaking.
All due respect to Mina… I don’t get the popular view that the #Packers will have a top D this year.
GB was 22nd in DVOA last year. They have not been top 10 since 2015. Alexander back is good! The rookies are promising. You can tell a D improvement story. But it’s unlikely. https://t.co/as6OfPHqDg
— Aaron Schatz (@FO_ASchatz) June 14, 2022
Aaron followed this up with a few additional tweets, which I will quote,
More detail on why GB D unlikely to be top 5 this season.
1) GB D was 15th/17th with healthy Alexander in 2019-2020.
2) Rookies are nice but Walker is ILB and GB often plays only 1 ILB anyway.
3) GB was already 7th in TO/drive last year. That’s unlikely to improve.
— Aaron Schatz (@FO_ASchatz) June 14, 2022
OK, but then we compare it to last year… when they were worse than 2019 and 2020. Is Alexander that good to make up the difference between 22nd and 2nd? Especially given that CB are notoriously inconsistent based on either coverage stats or PFF grades?
— Aaron Schatz (@FO_ASchatz) June 14, 2022
I think it’s worth taking Aaron’s arguments point by point. Before we get into that, I want to make it clear that Aaron is fundamentally correct about the Packers’ defense in 2021. By basically any advanced metric, it not only wasn’t as good as most of us thought, it was actually worse than Mike Pettine’s final defense that got him fired, despite having arguably better talent. Green Bay was 22nd in DVOA in 2021 (+3.6%, with an important reminder that a higher percentage indicates a worse defense), but 17th in 2020 (+0.5%) and 15th in 2019 (-0.9%). And before you jump all over Football Outsiders’ methodology, EPA more or less agrees with this assessment.
2021: 18th, 0.018 EPA/Play
2020: 12th, 0.026 EPA/Play
2019: 12th, -0.030 EPA/Play
All of that said, I agree with the general consensus that the Packers sure seemed better last year under Joe Barry, even without Jaire Alexander for most of the season. So, what exactly is going on? Should we be optimistic, or is there some secret downward trend in the background of these numbers, destined to doom Green Bay to yet another season of defensive mediocrity? One of the dumbest things anyone can ever write is “sure, regression to the mean happens pretty much every single time, but it’s not going to happen to my team.”
Sure, regression to the mean happens pretty much every single time, but it’s not going to happen to my team.
With apologies to Aaron Schatz, I do think he’s wrong in this instance, and I think some local knowledge of the team actually helps with context. I’m not 100% confident that the Packers will be a great defense this year, and Mina’s projection as the 2nd best defense in football seems hugely optimistic. However, I think a top-10 finish (by DVOA and/or EPA) is likely and a top-5 finish isn’t out of the question.
First, and most importantly, defenses are VERY tough to predict in general, and while Aaron cites turnovers as an area likely to hurt Green Bay, turnovers are just one little nugget in the big bucket of randomness we call defense. The fact that they ranked 22nd despite having good individual parts would, on its face, indicate regression towards being better, not worse.
The big thing though, which will be obvious to any Packer fan, is the secondary. Aaron states that corners vary immensely from year to year, and that Jaire may not be enough to move the needle alone; while that’s true, it also misses the point. Let’s take a look at where the Packers struggled last season and why they struggled there. Here is a chart of the Packers’ defensive DVOA rank versus receiver type. “Other Receiver” is primarily slot receiver, but is also lumped in with the occasional 4th and 5th receiver. I’ve also included the two best defenses (by DVOA) from last season for context.
This is all over the map, but the important takeaway is that under Barry in 2021, the Packers excelled at stopping all wide receivers. That’s not a bad way to be, especially given how bad most tight ends in the league are. Of course, tight ends really did eat them up, especially on short passes:
Here we get to the most important part about defense in general and why a quantum leap from 22nd into top-5 is entirely possible. The Packers may have had good individual players at each level of defense, but they still had their weaknesses, and weaknesses on defense are worse than weaknesses on offense. No one forces a quarterback to target a terrible wide receiver, but a quarterback absolutely can target a poor corner or safety over and over. More often than not for the 2021 Packers, that corner was Chandon Sullivan, though it was also frequently Darnell Savage or Henry Black.
Here, Aaron glosses over the other important change about Green Bay from 2019 to 2022. They lost Alexander for most of last year, but in doing so, they added Rasul Douglas. In addition, they of course also drafted Eric Stokes. Back in 2020, that big ugly #2 WR rank (24th) was largely thanks to Kevin King. That 2019 number against primary receivers (15th) is also King.
Opposing offenses prefer to throw to receivers if possible instead of TEs or RBs, as wideouts are generally the most efficient offensive weapons on the field. But they like to target bad defenders for obvious reasons. Combine the two, and your defense will suffer. Last season the Packers may have been worse in the aggregate, but at the very least, they made opponents do something they didn’t want to do. They checked down and they found their tight ends. It worked, and that’s a problem, but it’s a problem you can adjust to, and for the Packers adjusting to it is as simple as having all three corners healthy.
We should also take a moment to mention the related issue of the Packers’ run defense (28th last season). While run defense matters far less than pass defense, it still does matter, and even more so if you’re atrocious. If you’re part of the big blob of average run defenses, other teams won’t hurt you too badly, but the Packers are frequently so bad that opponents can do real EPA-positive damage against them.
The functional result of this issue is that they don’t put opponents in as many long down-and-distance situations as their peers, which was one of the primary drivers of the efficiency Packer opponents had with their tight ends and backs last season. Efficiency stats, more than anything, are driven by converting first downs. Backs and TEs tend to average between 8-10 yards per reception, and the more 3rd-and-4 (or 2nd-and-4) opportunities you give them because you couldn’t stop the run, the more efficient they will be. More stuffs on first down equals less efficient “short yardage receivers” on 2nd and third down.
And so, as to the question of whether Alexander will make a huge difference in 2022, the answer is almost certainly yes. To understand why, you merely have to look at snap counts, and understand the Packer weaknesses properly. In 2020 Alexander played 900 snaps, the 2nd most on the defense, and of course, the Packers were not a top 5 defense as Aaron Schatz pointed out. But that’s primarily because Chandon Sullivan played 729 snaps (6th on the defense), King played 664 (7th), and Josh Jackson played 331. Alexander was a workhorse, but defenses had other places to go. In 2021, Alexander missed most of the season, playing just 219 snaps. Had that happened in 2020, the defense probably would have collapsed entirely. In 2021 though, Eric Stokes was there to play 935 snaps (4th on the defense), while Rasul chipped in 680 after being acquired. Douglas and Alexander didn’t overlap until the playoffs, and so practically speaking, they basically count as one corner, which is why Chandon Sullivan, clearly having lost a step and playing some his worst football, got 827 snaps. Kevin King still managed 303 as a 4th corner.
In 2022, if everyone is healthy, we won’t have a situation where Alexander is accompanied by King and Sullivan. We will instead have a situation where Alexander pushes Sullivan completely off the roster (which indeed has already happened) and King is looking for work. Instead of Alexander and two giant weak spots, we will have Alexander eliminating one of the biggest remaining weak spots.
When defenses are missing players, they are forced to compensate, and the Packers did this last season by focusing on receivers and de-emphasizing the middle of the field. They counted on De’Vondre Campbell, and the general inefficiency of running the ball to make it all work. Having to cover for Sullivan and occasionally for Henry Black means that something will be open somewhere. The Packers picked their poison, and it mostly worked, but a vulnerability is still a vulnerability.
Enter Quay Walker, and perhaps more importantly, Devonte Wyatt. The Packers have lacked a quality run-stuffer since the likes of Ryan Pickett and B.J. Raji, but Wyatt should, in theory, fill that role immediately. Just like assuming that regression to the mean won’t hit your team, it’s a mistake to overrate, and count on the presence of rookies to have an immediate impact. That said,
Devonte Wyatt Will Have an Immediate Impact
Wyatt isn’t an enormous tackle, and he may provide more production than serving as a simple run stuffer in the future, but the good thing about defensive linemen is that run-stopping doesn’t take much time to learn. When lining up next to Kenny Clark, hunting backs should be a simple task for Wyatt. With Dean Lowry and Kingsley Keke at defensive end, the Packers were extremely light next to Clark, and the occasional heavies they did bring in lacked the versatility to stay in on passing downs. When the Packers would go nickel or dime, which was always, they would often be far too light to stop the run. Wyatt and Clark should essentially end that problem, as a pair of two-way defensive linemen capable of at least providing average production against run and pass.
Thus, the corners should allow Joe Barry greater flexibility in covering up the few remaining holes that do exist, and Wyatt should improve the run defense enough that it’s not an immediate fallback for offenses.
The tight end problem is still real, but it’s not as if there is no solution there. Maybe Quay Walker actually excels in coverage as a linebacker. Maybe Darnell Savage takes a step forward. Maybe with fewer holes in the secondary, Adrian Amos has more 1-on-1 opportunities there. Maybe corners like Douglas or Alexander are more effective covering those players when they line up in the slot. The point though, is that the Packers are down to a single weakness, and defenses can move and adjust to a single weakness. Especially when the tight end position is still a disaster across the league.
Finally, the single biggest driver of defensive success in a vacuum is the quality of the offenses they face. It is, of course, complete folly to attempt to predict how good NFL teams will be in the following season.
The Packers Will Face a Ton of Bad Offenses in 2022
Maybe a few teams will surprise us, and it’s not like the Packers face no good quarterbacks with Tom Brady and Josh Allen on the docket, but it’s a simple fact that they mostly face terrible quarterbacks in 2022. They have Daniel Jones (or perhaps Tyrod Taylor). They have Zach Wilson, or perhaps one of his backups. They face Tua Tagavailoa and Mac Jones and Carson Wentz and Justin Fields twice and Jared Goff twice and Jalen Hurts or one of his backups.
Do Dak Prescott or Matthew Stafford or Kirk Cousins scare you? Because they are among the best quarterbacks in the NFC and they on the schedule. What about Ryan Tannehill, who finished 27th in DVOA last season, now throwing to Robert Woods and rookie Treylon Burks and having been figured out by the league?
The Packers avoid the AFC West and their outstanding quarterbacks. There will be no rematch with Davante Adams. There’s no Joe Burrow or Lamar Jackson or, if he’s allowed to play, Deshaun Watson. There’s no Matt Ryan or Kyler Murray. There is a LOT of trash on the books that should help the defense shoot up the charts. DVOA adjusts for quality of competition, but there’s only so much adjusting that can take place. And EPA doesn’t care.
Finally, looking back at that weakness against tight ends, last season, George Kittle went for 97 yards against them in their first meeting. Travis Kelce had 68 yards and a touchdown. And Baltimore’s Mark Andrews tore them up for 136 yards and two scores. They don’t totally escape good tight ends this year and will face Rob Gronkowski if he rejoins the Bucs, but not facing the Kelce/Kittle/Andrews trio should be a big boost to their TE numbers.
Aaron could still be right
Plenty could still go wrong for the Packers. Jaire’s shoulder could still be a problem. Rasul Douglas, while grading out well on an every down basis, did create a lot of value on splash plays (the end zone pick against Arizona and some pick-sixes) and might regress a bit. Eric Stokes was great as a rookie, but he still can look like a sideways giraffe at the catch point. De’Vondre Campbell may not be able to repeat his brilliant 2021 season. Joe Barry may yet be a terrible coordinator. And the ever-present threat of injuries is always lurking, ready to expose their depth at safety or edge. DVOA’s view of the Packers’ defense in 2022 may be wiser than we thought.
And some of that likely will happen, which is why I would temper expectations a bit myself. Still, I suspect the ceiling here is closer to Mina’s take, given the quality of the top 3 corners along with Rashan Gary and Kenny Clark’s brilliance, Campbell’s ascension, Preston Smith’s return, and Adrian Amos’ steady hand in the back. Moreover, even if they experience the “normal NFL hiccups,” given their schedule, I think a top-10 finish is pretty likely for this unit. Injecting some first round talent, even if it was arguably slightly over-drafted, should help to shore things up at some of their weak points.
Going from 22nd to 2nd would be an amazing feat, but it’s a staple concept of analytics that defense is highly volatile — that seemingly minor additions or subtractions can have huge cascade effects, both positive and negative. The Packers have made some impressive additions. Absent injuries, it’s hard to see how they don’t take a significant leap forward.