If I never hear the term "regression" again it'll be too soon

You're looking for outliers

I not quite sure why, but I recently dug into my oldest archive of fantasy football posts over at RotoViz, which began in 2015, even linking to a couple. And man, are there some terrible posts over there (as well as some kinda cool ones I’d forgotten about).

Among those archives was a series I called, “FF Accounting,” because I was a certified bean-counting nerd when I started doing this stuff. It’s probably the biggest thing I could point to if you asked me how my playing style has evolved over the years that I’ve done this more seriously, as a job. Back then, I thought everything could be answered with projections, and understanding past data better than the market. Today, I think it’s the single most exploitable element to all of this.

The things people take for granted in analyses, like last year’s team volume being at all predictive of this year’s, or that the league itself isn’t morphing at all times. But there’s one specific mistake that needs to be held high above all else. And it’s with a heavy dose of irony, given where I started in this industry, that I say this:

Every single number-crunching data nerd who boils fantasy football analysis down to the available opportunity because they’ve over-regressed efficiency out of every player’s profile — presuming all players are essentially the same, since efficiency metrics aren’t particularly stable — needs to get stuffed in a locker.

As the subtitle says, we’re looking for outliers. The way you win in fantasy football, ultimately, is to find the players who post unprojectable efficiency. The vast majority of fantasy content revolves around roles, and opportunity, and what is there for players. And that stuff does matter, but not to the degree it’s focused on. It’s just easier to predict, so we focus on it.

At the end of the day, the league-winners win leagues, and the league-winners are the superstars who do stuff we couldn’t predict. And of all the worst fantasy analysis, the biggest mistake made — bar none — is the stuff that says, “This past efficient player was too good so we can’t pick him this year.” If what we should actually be looking for is the hyperefficiency — and again, I get it, it’s harder to find — then we should probably start with the really good players who have already proven capable of posting it.

I had my Joker moment on this topic with A.J. Brown, after a rookie season that saw him placed in a ridiculous cohort. I don’t remember the specifics clearly, but I know the cutoffs were simple ones: target volume (I want to say 75), a high aDOT (I think this was especially high at like 14.0 or something), and total yards after the catch (something like 400 total YAC). These were superstar profiles who earned volume downfield and also generated yards after the reception, which captured Calvin Johnson’s best seasons, and Tyreek Hill and Julio Jones and Randy Moss and AJB was there with them, in his rookie season.

That cohort wasn’t the only relevant data point for Brown, but the next offseason it felt like the only arguments against him were team- and volume-related (i.e. he couldn’t get enough targets). And I always go back to this same point because those people weren’t even wrong, necessarily. He didn’t really earn elite targets in Tennessee’s offense, in Year 2. He earned more, certainly, but he was a smash hit because he was efficient again, posting a YPT over 10 and scoring 11 touchdowns. Because he’s awesome at football.

Fast forward three years, and the market has finally caught up. And that’s the thing about efficiency — people love to hate it, until they’ve seen it so much it’s undeniable, but then by that point the reality of the NFL is that they are probably about to lose it. I said my Joker moment on this topic came with A.J. Brown, but probably more accurately it came from his former teammate, Derrick Henry, where I just refused to believe he could keep doing what he just kept doing, because it didn’t compete for me that he could post the ceilings he did, even in PPR, with a complete lack of receiving.

But even for King Henry, that had its limit. The past two seasons, he’s been more dependent on volume than ever, seeing his rushing efficiency fall off considerably in a way that is concerning for his age-29 season here in 2023 and beyond. He’s been a tough fade on these lines of betting on historically efficient players, but we do have a long history that suggests — even for elite backs — that when it goes, it goes, and it doesn’t come back.

One of the players that prompted this post here in 2023 was Breece Hall, and some discussion I saw on him a week or so ago about how you had to regress his per-touch efficiency from last year (it was elite, so sure), but then once you do that the volume becomes a concern, because of Dalvin Cook’s presence and his rehab and all of those things. And this is exactly what I would point to as the biggest issue that gets me to want to stuff people in lockers. Wiping away all the historical efficiency, and then saying, “and the opportunity isn’t good at that point.”