Biggest lessons learned about all 32 teams from the projections process
Cliff's Notes of my personal risers and fallers from a month-plus of research
I finished up my full 32-team projections yesterday morning, before recording the eighth and final division podcast with Mike Leone over at Establish The Edge. That podcast series is always such a pleasure to do with Mike, and there’s so much great info in it as we both went over all of the information we learned doing projections this year.
I’ll be releasing my full projections in the next day or two, after a final pass to make sure I’ve adjusted for all the latest news. As always, I won’t be keeping the projections updated after the initial release; they’ll stay locked in at that time as a reflection of how I thought in late July. It was fascinating for me to look at last year’s projections, and see how things evolved through August with my rankings (which do update essentially daily after initial release) and the positional tiers articles and all those things.
For me, the projections process is about the process. Last year, I released my full projections with a bit longer of a writeup than I’d sent in a couple years, re-covering a lot of the concerns and important considerations around the outputs of that process. I started by talking about the second time in my fantasy career where I did a full analysis of my prior season play volume, pass volume, and rush volume projections for each of the 32 offenses, and found the same thing I had years prior while at CBS — for roughly half the league’s teams, I was way off on one of those key metrics. That’s just the team volume; the degrees to which individual target shares and rushing shares can be off only spiral off of that.
This wasn’t because I’m terrible at projections; I’m actually tethering what I’m doing to market expectations in multiple ways. What I wrote in these two sentences captures the real issue better than I can write today.
That’s because NFL seasons are chaos. Trying to project a full season of football is trying to project a moving target, where we know major injuries and other unforeseeable stressors to individual team environments will change the parameters of the assumptions we thought we were making.
This same concept will be true just through training camp and the preseason, and if I was entering these projections into some kind of contest, I’d need to be making almost daily adjustments, and sometimes they’d be pretty substantial ones to reflect the latest news of that day and how it might have changed the range of outcomes for a given player pretty meaningfully. Sometimes, that news gets contradicted by something else meaningful a week later, and the whole projection would need to shift back. Such is the story of fantasy football value through August, as we try to learn as much about the various teams and their plans, while they often try to keep that all a closely-guarded secret.
I took the time to write out all the key things that dramatically changed the landscape for projecting 2023 teams between what we thought we knew in the summer, and what the season brought us. Again, these are the notes that impacted the 2023 season in a way we couldn’t really foresee (not last year).
The obvious example last year is what happened with Aaron Rodgers and the Jets. But consider that this time last year Miles Sanders and Alexander Mattison were projecting for strong new lead roles, Austin Ekeler was a first-round pick who had led the NFL in touchdowns two years in a row, and people were certain Najee Harris was a workhorse because Mike Tomlin always has workhorses.
There’s plenty more. Jordan Love was raw the first half of the year, then a star the second half, changing the way we’d view upside in the Packers’ receiving corps — this time last year the Packers had declined his fifth-year option; this time this year he’s the highest paid player in the league.
Justin Fields, Kenny Pickett, Sam Howell, and Desmond Ridder were young starters who here in 2024 are on different rosters. Anthony Richardson’s injury dramatically changed the upside scenarios for the Colts, something I discussed in my Offseason Stealing Signals piece for them.
I could easily write a similar section about 2024, just perusing last year’s final projections. So, fuck it, let’s do that.
For starters, Jordan Love suffered multiple injuries and during the season after the Packers made him the highest-paid player in the league, as referenced above, the Packers wound up throwing the third-fewest passes in the NFL, ahead of only dominant offenses with dual-threat QBs and those veteran RB adds who had the huge rush rates in Baltimore and Philadelphia.
Offensive line and perhaps schematic issues in Chicago and Houston led to two promising young QBs to lead the NFL in sacks. Perhaps both of those players were worse than expected, as well, and there is often a hint of “miss” tinged with the results; I was too high on C.J. Stroud for sure, relative to his 2024 season, but you also can’t comment on his second season without acknowledging that two of his top three WRs — that were a big part of the bull case — suffered season-ending injuries, in addition to the offensive line woes. The assumptions I was making at projection time about his supporting cast were not accurate, as so often is the lesson.
There were players that were clear “hits” for me, as an analyst, and yet they hit in a way that was still not capturable in a projection, and further reflects the limitations of the final projection number. I had Jayden Daniels projected as the 10th QB, which was above his ADP, but I had him ranked even higher as the biggest target for so many of you in your leagues based on the rankings and the QB Tiers breakdown — the stuff that is coming, and which will give you better insight into how to play things than projections. But as far as Daniels’ projection, and the projection for the Commanders as a whole, Vegas lookahead lines had them scoring the seventh-fewest points, something I don’t just ignore in this process. They would instead score 54 offensive TDs, tying for fifth most.
Bucky Irving displaced Rachaad White without injury, despite White’s far superior ADP. White was coming off a year in 2023 where workload carried him to the RB4 overall season in PPR, in part due to 64 receptions; again, there were limitations to the decline that could be projected if taking into account past data in any meaningful way. I did project a substantial decline in both rush attempts and receptions, and had Bucky’s projection among some RBs that were going well above him in drafts, and yet the projections still didn’t reflect how extreme my positions were on those two players in draft environments, which the rankings and tiers discussions did.
Brock Bowers caught 100 passes, something I didn’t get close to projecting, in part because Davante Adams was still on the roster and coming off a 30%+ TPRR season with a lofty ADP this time last year. I had a pretty decent projection for Jonathon Brooks with an expectation he’d play a lot down the stretch; Chuba Hubbard thus was a major miss as a guy who just consolidated backfield work all year.
I was really high on Chase Brown, projecting him for about as many touches as Zack Moss and more PPR points overall, in part because he had workhorse credentials from him time at Illinois that I thought were being missed by a market obsessed with his size, and Moss being the bigger body. But Moss’ injury really paved the way for Brown to realize that potential, and contingent upside cases are always difficult to capture in a single projection that is mostly going to assume health.
Like the ADP market, I had CeeDee Lamb projected as the overall WR1, but Dak Prescott’s injury didn’t help that cause. I had Tyreek Hill as WR2, but his own injury (and Tua Tagovailoa’s concussion) didn’t help that cause. My 49ers projection had a whole bunch of work for Christian McCaffrey, as well as Brandon Aiyuk, and those things obviously didn’t materialize. I had an optimistic projection for the Chiefs’ passing game that assumed Rashee Rice and Hollywood Brown would play more than a combined six games during the regular season.
I had Derek Carr, Chris Olave, and Rashid Shaheed healthy for a Saints team expected to potentially contend in the NFC South. I had the Bucs projected for about 38 offensive TDs after their 36 in 2023; they scored 57.
You get the picture. I’m just clicking through last year’s projections sheet and writing some quick notes, but the point is clear that dialing in the specifics of a projection is more about seeing what’s possible, rather than trying to be exact.
Actually, let’s back up. For me, the whole process is about the research. This is my annual time to dig deep into each offense, from coaching trends, to more time spent this year on offensive line environments, to the players deeper on the depth chart and how they impact the range of outcomes for the starters, and then of course to the key players themselves, the ones with ADPs, and where they are in their careers and where I expect them to go in 2025 and beyond. I spend a ton of time digging into each offense to make these projections happen, and it’s about the process of dialing in all my research more than the final output. A lot of the time there’s so much uncertainty in the projection that the final output feels almost meaningless compared to the info I’ve taken in from the process about various things that are possible.
Because of all that, I tried something new this year, where over the past month of deep dives into each of these 32 offenses, I made quick notes at the end about which players I wound up higher on or lower on while doing the projection. It was sort of that “perfect storm” concept I did last year that led to this, and obviously these notes were meant to help with the Offseason Stealing Signals posts, but those have evolved to be very in-depth, and essentially this was something where I wanted to just have quick bullets about the players or opportunities that I found interesting.
So before I release the full projections, and all the numbers and outputs I came to, I want to go team by team with this more qualitative and less quantitative list of all the things I learned going through the process. This isn’t going to be an exhaustive list, and broadly the key point about my projections I want you to take away is that the rankings and tiers-based discussion that will come — as well as what I’m writing in the Offseason Stealing Signals series — will tell you far more about how to play the different picks than their projection.
What I mean by that is as a first order of business, you should mostly be making picks based on who the players are, in terms of their actual profile and production record, and where they are priced. Good players tend to overcome bad circumstances, and vice versa, and the market tends to undervalue the actual profiles themselves, and how the elite upside we’re chasing mostly comes from the very good players, or very good prospects.
Most of what I’m going to talk about below is thus a second order of business beyond that. But it’s meaningful because the profiles themselves are never certain, and we’re so frequently targeting young players where we’re actually attacking that uncertainty, intentionally, because that’s where so many of the big wins can come from. And so within that equation, the things that go into (and come out of) a projection are meaningful, in that they help inform the probabilities and the ways various players might hit in different areas of their range of outcomes.
Alright, that’s enough introduction and groundwork being laid for this concept. It’s a new piece this year, but it’s one I’ll strive to do every year going forward (because it’s so obvious and I should have done it in the past). As I said, this won’t be exhaustive, and I’m going to move quick through all 32 teams. At the same time, I could have taken these notes and made them into like 10 different listicles with catchy names like “best handcuff RBs,” “deep sleepers,” “busts,” etc., because I’m going to go through and basically say a handful of the clearest names for all of those concepts.
Let’s get into it.
Arizona Cardinals
The team made defensive picks with each of Arizona’s five draft selections in Round 1 through Round 5, for defensive head coach Jonathan Gannon on that side of the ball. The only offensive player they drafted was a sixth-round offensive lineman, and there aren’t major free agent additions, either.
I wasn’t super into Trey Benson after a limited rookie year but the depth chart competition and situation is all very positive. Almost has to get a lot of work at some point.
James Conner has still been very good and I also came out higher on him.
Obviously Trey McBride is a favorite, and then Marvin Harrison sort of has to get force fed work as well.
Atlanta Falcons
Drake London should feast.
Bijan Robinson’s workload looks massive, so if the skill ceiling is there, he’s obviously elite.
Tyler Allgeier is a high-end handcuff with no RB3 competition.
Baltimore Ravens
I did see the Derrick Henry upside, and I’m usually not into this type of back. (I still probably won’t draft much of him, let’s be honest.) Similarly, got excited about Keaton Mitchell again as the potential second rushing back (Justice Hill would seemingly still be the receiving back).
Broadly, the “only one ball” notes from the team are tough. The top four guys all deserve more volume than last year, based on efficient profiles, and the addition of DeAndre Hopkins probably isn’t going to meaningfully impact Zay Flowers, Rashod Bateman, Mark Andrews, and Isaiah Likely, but he’s probably more of an issue than Nelson Agholor was last year. You’d like to think there’s contingent upside for some of these pass-catchers, but I’m not 100% sure how that would go. I love this team, but Lamar Jackson was insanely efficient last year and there still weren’t major ceilings for pass catchers, which made it one of the projections where I was most bummed out by the research.