Stealing Signals, Week 10, Part 1
The chaos of NFL seasons, plus TNF and early Sunday games
I write about the “chaos” of NFL seasons — how things evolve in unpredictable ways — a lot in the offseason. But it’s always hard to discuss how things evolve week to week, in August.
It’s quite fun, then, to hit on those examples in real time here in the season. I did this a little bit after Week 3, when I emphasized how much crazy stuff had happened through three weeks in 2022 that wasn’t at all indicative of our final understandings of that season. There are many more ways we could discuss this here over the past few weeks.
Let’s start in Arizona. Kyler Murray was back, and looked great. If you made that gamble, it looks to have paid off. And that was a gamble — there were risks about the team being so inept they’d shut him down or mismanage him (instead Jonathan Gannon has looked great for a coach who had lost 8 of his first 9 games), or his healing not going great, or even (as I’ve discussed recently) whether he’d be comfortable from a mobility standpoint once he did return. But Kyler looked completely healthy, was clearly comfortable moving, and the whole idea of “once he’s back you have a legit upside QB” looks totally valid. Watching him, it felt inevitable, like that was the only possible outcome all along.
That’s a major Signal for me this week. But then there are echoes with these types of things. As a result of this Kyler outcome, Trey McBride’s situational range of outcomes shifts meaningfully. Separate from that, McBride the player needed to develop after showing a little bit at the end of his rookie year, but finishing with a 0.84 YPRR on 317 routes where he mostly didn’t look like a high-end TE prospect.
He’s looked better all year in 2023, and two weeks ago he had a 14-10-95-1 receiving line in the first game Zach Ertz missed. But as I wrote in Signals that week, 6 of his 10 catches came in the fourth quarter, in comeback mode, mostly underneath. There were still huge positives to take from that game, which I did in that writeup, but Week 10 was — for me — the true breakout. McBride made a fantastic downfield catch on an underthrown ball, basically tossing a defender out of the way to come back for it. The underneath throws he caught, he did so with a clear smoothness to his game and a quick transition to being a load in the after-the-catch element. If you hadn’t seen anything but this game, you’d think he was a Travis Kelce-level TE, so dominant as to be the clear best receiving option in his passing game, with the perfect combination of comfortability/smoothness and aggression/power that is maybe better explained in a George Kittle comparison, and no I don’t care about how ridiculous these comps are because yes he looked that good.
Kyler’s play elevates what this all means. Suddenly, McBride’s passing game has an upside path that he as a player looks very ready to swerve into as a Year 2 TE that looks to have made the leap. Identifying that as possible in McBride’s range of outcomes wasn’t all that hard, honestly, but it did require faith in Kyler’s rehab, faith even in Gannon and the new regime to be competent in Year 1, and the faith in McBride’s individual development as well. Again, in hindsight, a player with McBride’s profile hitting like this feels obvious. This feels like it was the only possible outcome, now that it’s happened.
That’s one team, and one example. And I’ll keep emphasizing it was all known in the range of outcomes. Sometimes, these situations that create echoes on others’ ranges of outcomes are a little harder to foresee. Keaton Mitchell is maybe a clear one where he was a fun prospect to a lot of people, but the JK Dobbins injury was the first major element, and even still you had/have Gus Edwards and Justice Hill involved, and also had underperformances by some veterans like Kenyan Drake and Melvin Gordon who got some reps. And betting against those types isn’t exactly hard, but it’s worth noting that if Drake or Gordon had hit on some high-end outcome for whatever their potential ranges of outcomes were at this stage, that might have complicated the path for Mitchell.
The point is you needed several things to happen, and most notably probably a major injury, but also that those things are very hard to predict, and even now it’s hard to predict how this plays out because the Ravens for some very difficult-to-understand reason only gave Mitchell four total touches yesterday despite his first two going for gains of more than 30 yards, one on each of their first two drives in the first quarter.
Mitchell’s not the best example. It might be better to talk about things not going great, like with Tony Pollard. The concern for Pollard was whether the green zone work would actually manifest. It has, in spades. If you told me he’d have that role, and Dak Prescott would be healthy and this Dallas defense would be dominant and they’d be winning games and among the four clear favorites in the NFC by futures odds, I would have probably ranked Pollard RB2 ahead of Austin Ekeler, around fifth overall I think? The biggest Pollard stans would have probably been taking him top three, perhaps ahead of Christian McCaffrey.
All of that has happened, and yet Pollard has been a bitter disappointment. Friend of the newsletter Davis Mattek tweeted comparing his career path to Ben Tate’s this morning, which was a great call. A couple quote tweets compared him to Lamar Miller, which is another great comp. I mentioned Tevin Coleman, Miles Sanders, and Kareem Hunt in last week’s Stealing Signals, as consistently efficient players in limited roles where the efficiency just dried up instantly around the age-26 or age-27 mark. All of these names were good-not-great prospects, drafted between the second round and the fourth round.
But do you know who else comes up as a very obvious comp when we go back to preseason expectations? Ekeler. For his first 250-touch season in 2021, Ekeler was 26, and had been a consistently efficient player where the biggest question was whether he’d get goal-line work. He did, and after three TDs the year prior, and just 25 in four seasons up to that point, he led the NFL with 20 that year, and then led the NFL again last year with 18 more as a still-efficient 27-year old. We’re starting to see that rushing efficiency slip a bit this year, but he’s still been very good as a receiver, and he’s 28 now. For him, it’s been more gradual.
But the point of the Ekeler comparison is to say that liking Pollard as a guy with multi-year displayed efficiency where the green zone work might materialize and make him a fantasy superstar at age 26 — that obviously had upside for Pollard. That was the bet, and in the chaos of an NFL season, that bet has hit in exactly the way you wanted to see it, but the part very few were questioning — the efficiency — has been the anchor weighing down his production. He became 2022 Joe Mixon.
And it’s interesting to think about how that’s happened. For the first few weeks, the stuff we wanted to see was there, and that was awesome. I remember in Ekeler’s breakout 2021 season, he caught zero passes in Week 1, but got goal-line work, and I wrote about how huge that was from a “Signal” perspective, and how the lack of receiving was obviously “Noise.” He went on to catch 70 balls and have the 20 TDs.
That’s what the start of Pollard’s season felt a little like, and then as the inefficiencies have stacked up, it’s still felt like there were solid cases for regression. This isn’t a “chaos” point in terms of Pollard’s season having ebbs and flows as much as it’s a “chaos” point in terms of the ways we might analyze Pollard’s range of outcomes vacillating all over. Particularly with the weird circumstances in many of Dallas’ early games (i.e. blowouts), acting on the underlying green zone usage — which was the known indicator of his potential upside coming into the year — felt sharp. And then the thing Pollard was known for — the efficiency — is what has let that down.
This doesn’t mean it was never within the realm of possibility that Pollard might lose some efficiency. But for his efficiency to “over-regress” with the added volume to such a degree that he’s now inefficient is a pretty tough pill to swallow for many, and rightfully so. It didn’t feel like a significant enough part of the range of outcomes for this player to dramatically impact a buy/sell (read: draft/don’t draft) decision at his cost. For someone like my Stealing Signals podcast host Shawn Siegele, who was pretty out on Pollard and for reasons I could never really pin down, I think this was probably one of those concerns (he mentioned a couple times not wanting to chase up these types of efficiency plays once they got into the second round discussion, and instead looking for the next version of them, which at different times was a way to speak positively about guys like Travis Etienne and James Cook).
But it’s funny, because even though Shawn more or less nailed this — from an actionability standpoint, he did, which is absolutely what matters most — it was a situation where the specifics were hard to articulate in advance. It became something about not wanting to make that type of bet at that type of cost — more of a Tacit Knowledge thing — because it was frankly hard to talk about the Pollard of 2022 going down this kind of path. It’s hard to explain it now! He was so dynamic last year.
Man, how about Garrett Wilson as another example of the chaos? He’s played so well since losing such a huge part of his equation with the Week 1 Aaron Rodgers injury, and you’re getting a certain level of production that in my opinion just validates how much better he would have been if the bet would have been what we thought it was (i.e. Rodgers was healthy).
The point of the “chaos” conversations in the offseasons, year after year, is to remember that past is prologue, but that football data changes. What’s rare is the players that are dominant for long stretches, especially so when it’s the same type of dominance (compared to some kind of evolution of skill sets). I talk about this with season-long projections all the time, while I’m doing them and when I’m talking through them with the awesome Michael Leone in our summer podcast series — it’s just so hard to pin down how a season might go. For a player, for a team, for the whole league even. We’ll definitely be talking more — and soon — about how No. 1 WRs continue to dominate their passing games in a way that has not happened for years.
(This is one of those trends I have not dug into the numbers on, but I’m sure exists. I imagine the way to show it would be to look at percentage of receiving fantasy points the top-scoring WR/TE scored among all WR/TEs for each offense league-wide, and show how that’s evolved over the years, and my guess is that it has risen in 2023 relative to past years. I’m not sure why, necessarily, but have some theories, which I’ve talked about on some recent podcasts.)
Alright, let’s get to the games. It was a fun week with plenty of points and several great finishes!
Data is typically courtesy of NFL fastR via the awesome Sam Hoppen, but I will also pull from RotoViz apps, Pro Football Reference, PFF, Next Gen Stats, Fantasy Life, the Fantasy Points Data Suite, and I get my PROE numbers from the great Michael Leone of Establish The Run. Part 1 of Week 1 included a glossary of important statistics to know for Stealing Signals.
You can always find an audio version of the posts in the Substack app, as well as easier-to-see versions of the visuals at the main site, bengretch.substack.com.
Bears 16, Panthers 13
Key Stat: Panthers — 57 plays, 213 total yards
This wasn’t a game really worth spending much time on. Bryce Young didn’t look great, but the whole team doesn’t, and we know he’s not getting much help. The Panthers’ only TD came on a punt return early, and it just led to a really conservative and slow game. Former size/speed sleeper Mike Strachan (2-1-45) got out there for a few snaps and Young hit him for a 45-yard gain, which just went to emphasize the rest of the WR group isn’t very good, because we haven’t seen anything like that from Young all year really. Jonathan Mingo (7-3-20) continued to struggle, while Terrace Marshall (3-1-11) didn’t do much in place of DJ Chark. The TEs split work with Hayden Hurst (4-2-14) leading in routes and targets but Tommy Tremble (3-3-16) in yards. Adam Thielen (10-6-42) got his 10 targets at a 3.6-yard aDOT for 42 yards.
Chuba Hubbard (9-23, 2-2-16) continued to lead the backfield, but in a surprise twist, Miles Sanders (2-(-5), 2-2-15) ran routes on 43% of dropbacks to easily lead the backfield. Even when he was the early-down back, Hubbard would run quite a few routes; Hubbard seemed to be preferred there. Sanders has notoriously been a bad receiving back for several years, so taking away early-down duties but then having him lead in RB routes is just odd. For Hubbard, it was a season-low 30% routes, though he’d hit 31% a couple times early in the year while working as the No. 2. But that’s the point — he’s now the lead and yet he’s running fewer routes than he ever did as the No. 2? I’m not sure the Panthers’ coaching staff is very engaged at this point. Frank Reich gave play-calling duties to his OC a few weeks back, but wouldn’t rule out taking them back this week. We’ve talked about how the owner David Tepper likes to meddle, and whether that’s had a big impact, it certainly feels like the team has no direction right now.
On the Bears’ side, we got a lot of D’Onta Foreman (21-80-1, 2-2-12) and a lot of D.J. Moore (9-5-58, 1-1) early, though that faded as the Bears went run heavy later (-7.3% PROE for the game, and 37 total rush attempts, second most in Week 10 through Sunday, behind only Arthur Smith’s Falcons). You can’t really blame the Bears for just running it at a team that couldn’t really do much offensively; Roschon Johnson (5-18, 6-4-14) got some nice involvement with nine touches, while Darrynton Evans (2-5, 2-0-0) lost some playing time in this script.
Cole Kmet (7-5-45, 1-1) had another nice day, and he continues to be the second option when it’s not DJM, though DJM is another one of the league’s clear No. 1s. Darnell Mooney (4-2-14) was the only other non-RB to see multiple targets.
Signal: D’Onta Foreman — another game of heavy usage, both green zone touches, scoring one (continues to solidify at least a share of the early-down backfield even after Khalil Herbert returns)
Noise: Bears — 37 rush attempts (second most in Week 10, through Sunday; influenced by Panthers’ general lack of competitiveness); Chuba Hubbard — 30% routes were a season low, fewer than games where he was the clear No. 2, despite still being the No. 1 here with the early work and a team-high 52% snap share (it’s hard to understand why Miles Sanders saw an uptick in routes in the No. 2 role)
Colts 10, Patriots 6
Key Stat: Jonathan Taylor — 89% snaps, 75% routes, 24 touches (Zack Moss — 16%, 7%, 1)
With both teams dealing with pass-catching injuries, they leaned into run-heavy gameplans. The Patriots were especially run heavy, finishing at a -14.6% PROE, while the Colts were at -6.8%. When New England did pass, it was underneath, as they finished with a 4.3-yard aDOT that was lowest in Week 10, through Sunday, totaling just 115 air yards.
Jonathan Taylor (23-69-1, 1-1-6) dominated the backfield, playing an 89% snap share with Zack Moss (1-2) getting just one touch. We’ve reached the point with a clear lead back and backup dynamic, as expected.
Michael Pittman (12-8-84) is one of those No. 1s dominating volume, and that was especially the case with No. 2 Josh Downs (3-2-40) limited to just 43% routes by that knee issue that made him a DNP on the practice report all week. It’s hard to understand why Downs even suited up for this game instead of the Colts just letting their star rookie get healthy, although he made a sick diving catch on a crucial third down late, reading the scramble drill and breaking up field to give Gardner Minshew somewhere to go with the ball, and then bringing it in for a new set of downs that forced the Patriots to use all of their timeouts. Even when not fully healthy, Downs continued to earn the trust of his coaching staff and come through when called upon, which is big.
Alec Pierce (1-1-21) also entered this game questionable, and he made a nice, aggressive catch on his one reception. The difference between him and Downs is a worse profile and worse target-earning so far in their early careers; Pierce is a good example of a guy who is probably good at the catch point but doesn’t earn enough volume for that to matter. Isaiah McKenzie (4-4-31) was the only other Colt with multiple receptions, mostly playing the other half of Downs’ limited role as their routes added up to 89% of dropbacks. The TE rotation continues to render that position not relevant for fantasy.
Rhamondre Stevenson (20-88, 5-3-14) looked better than he has in some time, routinely making the first guy miss and picking up 6 or 8 yards. Ezekiel Elliott (13-54, 2-2-34) was solid as well, and I’ve realized is a pretty good example of one of those “no wasted movement” RBs (that I was recently joking Isiah Pacheco is the opposite of). Elliott has that classic RB aesthetic, where he reads holes well and hits them hard, and it is aging well. It sort of wouldn’t surprise me if he did this for another few years, lingering not for as long but sort in the way Frank Gore did.
Demario Douglas (9-6-84) continued to lead the way for the Patriots, though he had a 30-yard catch on the final play of the first half, where in a Hail Mary situation from midfield, the Patriots opted to throw a screen and compile some stats while not really threatening to score. Without that catch, Douglas wouldn’t have had a particularly notable day despite his nine targets amounting to one-third of the team’s 27 pass attempts.
Hunter Henry (5-3-21) was the only other WR or TE to see more than three targets in the low-volume game. Mac Jones was bad again, and eventually got benched for Bailey Zappe with the game on the line. Zappe threw a pick to end it. This remains a poor passing game for fantasy.
Signal: Jonathan Taylor — season-high 89% snap share, Zack Moss got just one touch; Demario Douglas — 33% target share
Noise: Josh Downs — 3 targets, 2 catches (DNPs all week, only 43% routes, and one of his catches was a fantastic diving play on a huge late third down); Demario Douglas — 30-yard catch before half (WR screen against prevent defense in Hail Mary situation); Patriots — (-14.6%) PROE, 4.3 aDOT, 115 air yards
Browns 33, Ravens 31
Key Stat: Keaton Mitchell — 13 snaps, 4 touches, 66 yards, 1 TD
I mentioned on Stealing Bananas last night how the Ravens seem to lose some games they should really win, at least once per year it seems like. Doesn’t necessarily make them a terrible team or anything, but it’s obviously a bad trait, especially for a team that should be so dynamic running the ball. Well, here’s a bad stat that encapsulates that!
It’s also kind of funny in the sense that in Lamar Jackson’s early career, the complaint was they wouldn’t be able to come from behind. But clearly that hasn’t been much of an issue if every regular season loss has come after being in great position to win at some point in the fourth quarter.
