Stealing Signals, Week 2, Part 1
Football's not dying, plus the Signal and Noise from early games
While Week 1 is all about confirmation bias, Week 2 gets even more dangerous. It feels like only two things can be true: The takeaways from Week 1 were correct or incorrect. If the takeaways from Week 1 were correct, that’s just how it is. We have two data points now!
What’s important to understand is things that have happened for two weeks will still be non-representative of the larger sample. At the same time, of course trends are absolutely starting to develop. So many of the things this week are going to be about me saying, “I got that call right last week,” or “I got it wrong.” It’s all going to anchor to what we discussed in Week 1.
I feel pretty dumb for my Calvin Ridley take, for example, that his volume was somehow a bad thing. I’m still very concerned about Will Levis, but they did a good job of getting him some better looks early in Week 2, and the broader point is being the focal point of a pass game is obviously a good thing. Things will get drawn up for you. Gameplans tend to work in your favor. Sometimes I make fantasy a little more complicated than it is, perhaps in search of something insightful when it’s a simple analysis. I’ve been doing this too long; repeated too many of the same concepts. I’m glad I was quicker to acknowledge I might be very wrong on Alvin Kamara, because Week 2 confirmed that preseason fade isn’t going to go well.
But there’s other stuff I tried to preach caution on where I could reasonably claim a victory. Anthony Richardson was in a boring football game, which has more or less been the point of all my words of caution, that while he’ll definitely score, his preseason price was tricky when considering that these games would definitely occur, too. That kind of thing is gonna happen some this year, in addition to the ways he will accelerate things.
I’m proud of the commentary that the rookies were perhaps the biggest Week 1 storyline. Brock Bowers, Malik Nabers, and Marvin Harrison showed why that trio was viewed by some as a generational top of a draft class (look at that, I expanded “generational” to refer to a whole group of players, and I think it actually might work).
Week 2 was quite fun, and the start of some recent seasons have been frustrating at times, but I’m feeling pretty darn good about where things sit, even on teams that aren’t necessarily smashing. I guess the one concern I keep seeing is about a lack of scoring, but I guess it doesn’t bug me as much as some of the commentary I’m seeing, perhaps because it’s something we’ve discussed ad nauseum around here and it isn’t exactly a shocking development.
If you’ve been reading here, you know that during the 2021 season, I wrote an intro that emphasized the league-wide dropping pass aDOTs in real time, that more or less predicted the scoring decline in 2022, then spent last offseason cautioning that the expected offensive counterpunch wasn’t obligatory, because what defenses had done was focus more on what their goals were, and what they wanted to accomplish, rather than being reactionary, as so well-documented in the podcast series The Playcallers from The Athletic.
The particularly notable thing about those defensive goals was the focus on removing explosive plays — something offenses had emphasized as they took a more analytical approach, but then suddenly defenses were taking that same analytical approach to what they should try to stop (and sacrificing rushing efficiency to do it). Defense became less about showing your manhood and physicality and stopping the run and they joined the chess game offenses have always been playing. The sport changed, as it has countless times before, like when the West Coast Offense ushered in a decades-long revolution in shorter and higher-efficiency passing with increasing completion rates and declining interception rates, or subtler things like the 2014 illegal contact and defensive holding stuff I’m always talking about that made smaller WRs more viable.
If you thought about the “why” for any of the recent changes in defensive dominance, like we talked about around here, none of that toothpaste was ever going back into the tube. I honestly have a hard time understanding why people expected anything different, but I’ve been arguing with people for three years now that this was the sport evolving, and being called a fearmonger, and told I don’t know ball countless times from people in the comfort of the consensus appealing to the authority of football coaches who got where they were through years of experience in a field that is about evolution. But after years of this being the clear direction, a couple down weeks have people thinking the sport is dying.
I really didn’t want to get into a long intro today because it’s my daughter’s birthday and I have to get this out before she’s out of school, but no, we’re not fucked. We’re just in a bad place right now because, to reiterate, experience is not the best coaching trait when compared to thinking dynamically about new problems, which obviously isn’t unique to football. I hate business-school and tech-world jargon, but football is in a period of disruption on the defensive side. How offenses disrupt back is of course what I’ve always meant when talking about innovative coaches, but we got to watch Shane Waldron fail again last night because those in power in football don’t understand the sport, and it’s way outside the self interest of most of the people inside the sport to admit new ideas are needed. If you defer coaching decisions to people with experience, they will tell you to hire coaches with experience. Meanwhile, Mike Macdonald is this young defensive wunderkind and he went down to the collegiate ranks to get my alma mater’s offensive coordinator, Ryan Grubb from UW, which whether it works or not is absolutely good process. I don’t think an endless supply of these guys exist, but you absolutely have to try to find them.
I’m particularly frustrated by all this today because my Huskies lost to their rival Cougars this weekend in a way that made it clear the new coach — while somewhat in touch with a few interesting offensive concepts — has no modern game-management skills (tempo, poor fourth-down decisions, run/pass splits in a game where UW averaged more than 10 yards per pass, we lost a game we never should have lost almost purely because of not pushing the easy buttons). I mention that because I wrote over the past two years how fun it was to cheer on the old coaching staff, and so now my fandom has brought me very much back into my “Appeals to authority are lame and I know better” era.
But we’re not fucked, to be clear. I do think the panic has gone overboard. There aren’t even enough offenses doing the simple things they should be doing for us to know, but if you saw Chris Olave’s big catch early in the New Orleans rout, and noticed the motion and then play action and then how it got him open for a “splash zone” target on a first read in the middle of the field, you might have thought, “Hm, more teams should be trying that kind of thing.” I’ve called those types of things “easy buttons” — I wrote an intro about how motion is the future either last year or the one before — because defenses are trying to force QBs to “work,” as Cody Alexander so perfectly put it in a thread this morning.
I did want to make sure to link that thread, because for those of you who have been following my commentary on this over the past few years, I’ve been more or less begging for someone with way more Xs and Os comprehension to share details like that exact thread, since I’ve been in front of these changes but mostly haven’t understood the details (as I’ve argued, I’m not sure many people understand the details, because it’s all so new). To that point, I’ll also pick a minor nit with Cody’s first implication that pundits calling it “shell coverage” somehow don’t understand the nuances, because many like myself, who do understand the broad strokes but just not the specifics in as detailed of a way as he does, just use that to shorten the things he described. This is no longer directed at Cody, but I’ve seen this a lot, and I love this semantical point people who want to pretend they know ball keep making that “it’s not really cover 2,” or “it’s not really a shell, it’s more than that,” because the whole point of that term from the very beginning has been to simplify the complex. Some people do get it confused and think it’s really all about Cover 2 defenses on every snap, but anyone talking about it at any depth is using that term differently, and it’s annoying to get all the stuff about, “Cover 2 wasn’t invented yesterday, these guys don’t even know about Tony Dungy’s Tampa 2, har har.” (I have heard so much smugness on all these points, so pardon me as I mock all of it.)
I realize it’s not true for everyone, but when I talk about “Cover 2 shells,” what I’m saying is, “defenses are rotating more post-snap, and being way more multiple on their end, and QBs can’t trust what they are seeing, and it’s wreaking havoc,” and the reason that all works is they are often showing this vanilla look pre-snap that allows them to get to a ton of different looks post-snap. That’s what the Cover-2 shell is, at least as I understand it. Sorry, this is all a huge aside to the broader point, and again definitely not an attack on Cody, but it’s just such a great microcosm of the broader issues with football that people can correctly identify all issues at hand and get shamed over the semantics of terminology as if someone who doesn’t understand the Xs and Os can’t understand the broader goals of the schemes and how those things interact, and have something useful to say about it. Thinking dynamically versus experience. This is a game, and the broad tactics aren’t actually that complicated to understand. You don’t need 20 years of experience within the sport to comment on those tactics; that’s bullshit, and it’s bullshit in service of the people who again need it to be true that experience is way more important than it is.
My grievances about years of gatekeepers telling me I don’t know ball because I was constantly saying this was a more meaningful change than consensus aside, we are already shifting back. The extreme run rates are some of the fallout as teams really understand what they can and can’t do. But running the ball a ton won’t be the eventual answer; it’s just the short-term solution when you need an answer now.
If you need the optimism, it’s as simple as looking at the Saints and the Dolphins and the Vikings. The sport isn’t dying; the people in power are just idiots hiring bad coaches who have been shown to be bad, and there are only 32 teams, and the inevitable move to an NFL where motion and more creative play designs are table stakes is just taking way longer than it should. Defenses found something, and as Cody put it, they are making QBs work, and get through progressions, and are messing with what they are seeing pre-snap and post-snap. Some of the QBs just do not handle that very well. Part of the answer is we do need quarterbacks who can make those post-snap reads, and some of that is development.
But the other part is the offensive counterpunches are already known, and we’re just waiting for more good schemes. In a world where the QBs have to work like this, plays like the one to Olave I described above are vital. Offenses need to manipulate defenders, rather than be reactionary to the post-snap rotations. Who is forcing the other side to react?
If your coordinator doesn’t have a scheme that gets his QB 2-3 easy first-read completions for decent gains basically every game, then he probably isn’t it. That’s the whole point of the easy button stuff, and it’s what I wrote about UW’s offense last year, that they more or less spammed all that stuff into all their plays. People will say, “You can’t do that every play,” except the teams that do it as much as fucking possible are doing well. Again, this isn’t even complicated, people are just deferring to nonsense conventional wisdom, and half the league’s offenses are run by dinosaurs.
But is the sport dying? Or in 10 years do we finally have 32 offensive systems that take full advantage of all the things that they should. I think the latter. We’re already moving that way, and more and more of these (younger) coaches are being hired because they think dynamically about the future of the sport rather than being hired based on their history. I don’t think football is solved, and defenses won. I do think offense will rebound. But it’s more about coaching and scheme than ever before, at least right now. (And because I got this reply on Twitter, no it’s not as simple as just motioning, and yes, you have to do it with some intent. The point is that when done well, you start to force the defense to react pre-snap again, and you’re always looking for breaks. Or put differently, the point is that lining up and doing boring shit is not stressing them at all. That doesn’t mean all motion is good enough.)
I also suspect that on a longer timeline, since defense is just a multiple thing now, that QB play will emphasize those post-snap read traits at the lower levels, and we’ll eventually have another wave of QBs who are just dynamic thinkers on the field, an offensive coordinator under center who understands what the defense is trying to do as well as the guys being coached on that side of the ball. That kind of thing isn’t easy, but I have come around to this idea that the average NFL QB’s “Football IQ” is lower today than it was 10 or 20 years ago.
But no, the sport is not dying. It’s just lame right now.
Today’s writeups will need to be quicker, because it’s my daughter’s ninth birthday, and her happiness is more important than yours. I’ve also changed the graphics to include Missed Tackles Forced and Yards After Contact for RBs, from NFL Pro.
You can always find an audio version of the posts in the Substack app, and people seem to really like that. You can also find easier-to-see versions of the visuals at the main site, bengretch.substack.com.
Data is typically courtesy of NFL fastR via the awesome Sam Hoppen, but I will also pull from RotoViz apps, Pro Football Reference, PFF, the Fantasy Points Data Suite, and NFL Pro. Part 1 of Week 1 had a glossary of key terms to know.
Bills 31, Dolphins 10
Key Stat: De’Von Achane — 10 HVTs (led Week 2), 4 green zone touches
I wrote about De’Von Achane (22-96, 7-7-69-1) Friday, talking about how good he looked and how he clearly wasn’t injured, playing just shy of his career high in snap rate. His TD came on a sick “escort block” play where fullback Alec Ingold motioned out and was the lead blocker on a swing pass where Achane followed him that direction. He looks like an absolute massive hit, and the reincarnation of the slender efficiency superstars of the past, e.g. Jamaal Charles and Chris Johnson. This is what it felt like to watch those guys, and Achane’s 10 HVTs mean the usage supports the efficiency in a way that creates a massive ceiling. Health permitting, he’s a top-five kind of fantasy asset. I frankly thought the Dolphins used him too much in a blowout loss when he was questionable entering the game, but Jeff Wilson (2-2-12) did leave early, and Jaylen Wright (5-4) maybe wasn’t ready to handle certain things. We did get our first look at Wright, who didn’t look as good as the preseason, but is still a hold as every other Dolphins’ RB has had some kind of injury already this year.
On the other side, James Cook (11-78-2, 1-1-17-1) looked fantastic, but more in a “spike week” kind of way. He was very efficient, and found the end zone three times, but only had 12 touches overall. A completion down to the 1-yard line set up one of the three TDs, though he did sub in for that. He ran hot here more than this being some huge change, though it’s great to see the idea that he’d be more of a focal point this year paying off early. Ray Davis (9-29, 1-1-(-1)) looked good alongside Cook, and Ty Johnson (1-(-1), 2-1-33) made the catch down the 1, and ran a good chunk of routes (30%).
Josh Allen only threw 19 passes, which wasn’t even that low of a PROE at -2.2%, because they were just so comfortably in control. They did come out throwing early, and specifically got designed first-read targets to Dalton Kincaid (4-4-33) on the first and third snaps from scrimmage. Unfortunately, Kincaid got kneed in the head after the play on that second reception, and then missed the next couple drives while presumably being checked out for a concussion. The Bills didn’t run many late second-quarter plays because Cook hit for a 49-yard TD run on their first snap of their penultimate drive of the half, with the final one just being a kneeldown. To start the second half, Allen hit Kincaid of the first snap for 9 yards, but the Bills wound up punting four plays and just two pass attempts later, then got an interception touchdown, and didn’t get the ball back until there was 2:40 remaining in the third period. In other words, Kincaid’s concussion check and then weird play volume during the period after he returned and then ultimately a blowout script late led to a really weird game, and yet he had another catch very late in the third quarter, and it was frankly clear to me they Bills wanted to get him rolling as their No. 1 option, as evidenced by what ended up being a 27% TPRR, with his four targets totaling -8 air yards because they were basically all scripted. In other games where Allen throws more, Kincaid will have a shot to work downfield some, and while the start to the season hasn’t looked great, it’s not over for Kincaid.
That said, the Bills’ pass game is not one you want a ton of exposure to. It’s going to be unconcentrated, and probably pretty run heavy, if these first two games are any indication. The hope with Kincaid was he might rise above that, and I do think there’s still potential for that after watching Week 2. Other than him, Khalil Shakir (5-5-54) was the only other player who caught multiple passes, and we did see Shakir’s routes rise to 85%. Keon Coleman’s (1-0-0) routes stayed steady at 90% but the rookie unfortunately did not record a catch.
The Tua Tagovailoa concussion was the low point of Week 2 around the league, and everyone obviously hopes for the best there. My expectation is he’ll miss some time, and the Dolphins acquisition of Tyler Huntley perhaps signals that. Mike McDaniel is a good coach, but the Dolphins will be worse without Tua.
Tyreek Hill (6-3-24, 1-12) and Jaylen Waddle (4-4-41, 1-4) both had quiet nights, as Achane was really the Dolphins’ only offense. Jonnu Smith (7-6-53) was decent, but the Bills had big linebacker issues and it was a clear matchup exploitation. At least one of his catches came from Skylar Thompson after Tua left, and after the other starters left in the early fourth quarter. I know this seems like just anti-Jonnu commentary, but note that fellow TE Julian Hill (2-0-0) had two nearly identical-looking near misses on seam throws, one from each QB, where he was out-stretched and just couldn’t pull in a fingertip catch. If you want to really buy Jonnu, you’d want to have seen him getting all the TE work in this extreme plus matchup.
Signal: De’Von Achane — 10 HVTs, efficiency to boot, top-five overall kind of fantasy asset; Dalton Kincaid — four schemed targets, 27% TPRR (missed some time to a concussion check and Bills only threw 19 passes, but some minor usage optimism here); Khalil Shakir — routes rose from 70% in Week 1 85% in Week 2, team-high 5 targets
Noise: James Cook — 3 TDs (just 1 green zone touch, stacked explosives plus ran hot with a completion getting down to the 1-yard line where he subbed in); Jonnu Smith — 7 targets (clear matchup advantage with Bills LBs, but didn’t even get all the Miami TE work, as Julian Hill got two end zone targets); Dolphins — passing offense without Tua Tagovailoa is going to look a little different
Raiders 26, Ravens 23
Key Stat: Brock Bowers — 32% TPRR, 0.70 wTPRR, 9 targets
Brock Bowers (9-9-98) was the story, but so was the Raiders already recognizing their strength and posting a +10.2% PROE, second-highest in Week 2, en route to a come-from-behind victory. Davante Adams (12-9-110-1) had a couple near misses that could have led to an even bigger game, but the Raiders threw a ton, recording a 72.8% actual pass rate that was also second-highest on the week, and they concentrated those looks onto their two stars, as Jakobi Meyers (5-4-29) chipped in some, and the rest of the WRs and TEs totaled just 4 targets. Bowers was only out there for a route on 65% of dropbacks, which is the kind of stat you should look at and salivate over, because he drew a target on 32% of routes, caught every one, made everything look way easier than it has any right to look, and was so clearly one of the most dominant players on the field that there’s no world where his routes are at serious risk of staying that low all season. We do often see this early in the year with young players, but it’s not a long-term issue when it’s this clear; coaches might make mistakes but they aren’t flat-out stupid, in most cases. They want to win. When everyone on the team can see how good someone is, they will use him. Bowers’ ability to draw 17 targets in his first two games, and for the setup to look so good for him to be able to continue to do that, and for everything to look so easy for him right out of the gate, with the prospect profile he has and what that likely means in terms of him just being that good — he’s absolutely in the conversation for the redraft TE1 rest of season, and he’s a buy high, and he’s probably your 2025 TE1 in ADP, and all of those things. To be clear, if I were drafting a team for rest of season right now, I’d draft him ahead of all other TEs. We seek upside plays and the possibility exists that this is just the beginning. It’s not a certainty, at all, and if he doesn’t go on to a 1,200-yard season it’s going to be easy to point at this and say, “You were overzealous,” but these first two games create a trajectory where the floor is also so high, and it’s just more or less impossible to hyperbolize what this start means, but it’s part of why we draft rookies and target these types of players. My notes initially said, “Bowers def made some more fun plays in this one, clear hit through two games.” Then added, “Bowers catch down to the 1 to set up short TD.” And, “Was even better late. Massive focal point.” You don’t back into a game like this, and the fact that he didn’t score more fantasy points is why I’m trying to emphasize so clearly what it means, because there is potentially a buying window.
As I said, this was also a great game for Adams, and the Raiders showing a willingness to get very pass heavy and also be concentrated — getting the ball to their best players — plays into both the excitement for Bowers and also a bump in what to expect from Adams going forward. Gardner Minshew showed us that the competency we were hoping to see at QB is probably there, and he was able to support both of these guys, on the road, against a very good defense. Will he be able to do that every week? That’s tough to say, but the expectations were low here, and this type of game being possible is a strong positive symbol in the current landscape.
Zamir White (9-24, 4-3-14) got back into a clear lead role, dominating the snaps. Alexander Mattison (4-1-1) got the short-yardage TD, but totaled 1 yard and only played 22% of the snaps. White’s routes shooting up to 54% were great to see, but it’s hard to know what to expect going forward, as Mattison was at 60% in Week 1 before falling to 19% here, and Ameer Abdullah (2-1, 2-2-1) also ran 12% routes in this one.
The Ravens were in control of this game throughout, but just struggled to put the Raiders away. In their defense, Vegas benefitted from a phantom facemask call on Minshew that helped extend a fourth-quarter drive, and then it was just one of those things where they were up 10 in the fourth quarter, but the Raiders were able to score on three straight possessions and force two punts in between, and leave basically no time on the clock at the end of that, hitting the go-ahead field goal with 27 seconds remaining to take their first lead of the game.
Zay Flowers (11-7-91-1) was the focal point of the passing game, and had a big day. Mark Andrews (5-4-51) was sort of the second option, along with Rashod Bateman (4-3-40), but Andrews ran just 75% routes for the second straight week, which is tough. I noted last week his average last year was 85%. Isaiah Likely (3-2-26) also lost routes, falling to 50%, and we unfortunately saw that these guys are probably going to be more similar to most offenses that show a big game of two-TE sets and then move off it some the next week, rather than redefining things. For Andrews to stick at 75% routes and Likely to drop 19 percentage points was a big bummer for them combined, after the Week 1 usage. If the two-TE rates are going to be in this range, and Likely is still going to impact Andrews’ routes and keep them around 75%, Likely is probably a spike-week TE2, while Andrews is still solid but a bit less valuable than where his initial ADP was. You’d definitely prefer Kittle and especially Bowers over him right now, but you could argue names like Kincaid (that debate would feel very similar to the preseason for me, where Andrews would be higher floor but Kincaid would still feel higher ceiling; Kincaid was at 75% routes even while missing multiple drives, for reference).
Derrick Henry (18-84-1, 1-1-12) had a strong day on the ground, and got 3 green zone touches. Justice Hill (4-22, 2-2-10) dominated the passing-down snaps.
Signal: Brock Bowers — the ceiling is the roof, the floor looks better than it has any right to be, TE1 rest of season, buy high candidate; Raiders — +10.2% PROE (willing to get very pass heavy, which was a great sign), concentrated targets; Mark Andrews, Isaiah Likely — Andrews stuck at 75% routes, still down from last year’s 85% average, while Likely fell from 69% to 50%, which suggests Week 1’s very high rate of two-TE usage probably won’t be the norm, a downgrade for both
Noise: Raiders RB routes — Alexander Mattison fell from 60% to 19% this week, while Zamir White jumped to 54% from 25% (not clear what the split might look like long term, as this seems like they were trying to emphasize White a little)
Chargers 26, Panthers 3
Key Stat: JK Dobbins — 5 missed tackles forced, +41 RYOE (fourth most in Week 2, through Sunday)