Editor's Note: Get an edge on draft day with our 2022 Football Draft Guide powered by Rotoworld that's packed with hundreds of player profiles, expert fantasy rankings, projections, mock drafts, sleepers, busts and much more. And don't forget to use promo code DRAFTGUIDE to gain access to it all for just $5 for the first month. Click here to learn more!
Greetings and salutations to you, a fine person of the Internet! I would like to start this article with a definition from the beloved Dictionary.com.
Statistics. a procedure for determining a relationship between a dependent variable, as predicted success in college, and an independent variable, as a score on a scholastic aptitude test, for a given population. The relationship is expressed as an equation for a line (regres·sion·line) or curve (regres·sion·curve) in which any coefficient (regression coefficient) of the independent variable in the equation has been determined from a sample population.
Often a source of contention in the fantasy football Twitter space, regression can be both positive and negative. It's hard to believe but it's true.
When considering how a player could perform from one season to the next, looking at regression models can help us determine whether or not a player is worth our precious draft capital. At the very least it can serve as a tool for making an educated decision.
It's easy as fantasy managers to remember the best about a player who led us to a championship or the worst about a player who burned us. But memories triggered by emotion can often distort reality — which only adds further value to these exercises.
In hopes of getting a better understanding of players who overperformed and underperformed expectations in 2021, I used the RotoViz Screener tool to look at some linear regressions. Now fully woke to every good player to draft this upcoming season, I thought the next best thing to do was share it with the world.
Positive Regression Candidates
Cole Kmet (CHI)
Perhaps two of the most unlucky seasons in 2021 came at the tight end position. Bears tight end Cole Kmet had arguably the worst luck of the two players, as the second-year pro took a major leap in receptions (60) and receiving yards (612) but fell flat in touchdowns with zero.
I touched on Kmet back in a June ADP article where I noted that his scoreless 2021 on 93 targets was a dud for the ages. Looking back to the 2000 season, here is how touchdown production has played out for tight ends to see 70+ targets in any individual season.
My quick look at this historical data revealed that tight ends who drew 70+ targets in a season averaged 5.1 touchdowns. According to the RotoViz linear regression model, Kmet's 93 targets should have resulted in 4.95 touchdowns — lining up almost perfectly with my look at the historical average.
The regression model also suggests that Kmet fell well short of his fantasy point expectations. He finished as the TE21 with 121.2 fantasy points, while the model suggests he should've scored closer to 161.5 points — which ranks him as the TE9 in the expected points model.
When it comes to this season, not much has changed for Kmet's situation. The Bears moved on from Matt Nagy at the end of 2021, hiring Matt Eberflus as their new head coach. A coaching change should bode well for Justin Fields and the entire offense as a whole, while Kmet's target competition outside of wide receiver Darnell Mooney remains relatively thin.
Kmet is currently going as the TE12 in full-PPR leagues, which may be close to his ceiling. But last year's numbers suggest that ceiling can be tapped into with a similar volume and some positive touchdown regression.
Kyle Pitts (ATL)
Close to Kmet on the above is Falcons tight end Kyle Pitts, who erupted for 68 receptions for 1,026 yards as a rookie but managed just one touchdown.
An athletic maven at the position, Pitts lived up to his pre-draft billing, finishing fifth in yards per route run (2.02) and ninth in YAC (318). Viewed by some as a wide receiver in a tight end's body, Pitts led all tight ends last season by lining up out wide on 34.2% of his snaps — his 43.8% snap share from the slot was tied for 15th.
Much like Kmet, Pitts also should've seen better touchdown numbers on his 109 targets — the fifth-most among tight ends. He and Kmet saw the worst touchdown luck of any tight ends in 2021, narrowly edging out the likes of Mike Gesicki, Darren Waller and Dan Arnold.
Currently sporting a 39.3 ADP in Yahoo drafts, which are .5 PPR, Pitts won't come at nearly the discount Kmet is. But his elite athleticism and impending target volume make him well worth the pick. The only downside to Pitts is he may be playing in a bottom-10 offense in the middle of a rebuild.
DJ Moore (CAR)
Death, taxes and THIS being the year DJ Moore finally reaches his full fantasy potential. It's been one of the few certain things in life throughout much of Moore's career, as the uber-talented receiver always seems to fall a few touchdowns short of being the elite fantasy producer we know he can be.
Looking again at the differentials between actual touchdowns and predicted touchdowns based on volume, we can see Moore has fallen well short of the end zone on multiple occasions.
Since entering the league in 2018, Moore has seen truly elite volume at the wide receiver position. His 834.2 expected fantasy points rank seventh behind Cooper Kupp, DeAndre Hopkins, Tyreek Hill, Stefon Diggs, Keenan Allen and Davante Adams. I assume you have heard of them. His career 12.3 fantasy points over expectation rank him among the likes of Gunner Olszewski, Jake Kumerow, Isaiah McKenzie and Chris Conley — guys I'm sure you've also heard of, but haven't been dying to start over the last four years.
Being force-fed targets from a broken-down Cam Newton along with Teddy Bridgewater and Sam Darnold sounds like a bad recipe for fantasy success, but several No. 1 receivers have turned high volume into touchdowns throughout their careers. Andre Johnson spent 14 years in the NFL and never met a good quarterback.
For all the criticism Baker Mayfield receives, there's a good possibility he's the best quarterback Moore has played with in his professional career. His overall targets from last season may take a slight hit, assuming Christian McCaffrey can pull off a mostly healthy campaign in 2022, but even 130 targets should put Moore within range for six or seven scores.
Negative Regression Candidates
Dawson Knox (BUF)
Keeping the tight end theme alive for this one, let's talk about Dawson Knox.
The veteran tight end enjoyed his most productive season to date in 2021 when he went for 49-587-9 on 70 targets — career highs across the board. It seems good, right?
Knox plays with one of the best quarterbacks in the league in Josh Allen and is on a team that ranks second in neutral pass rate since 2020 (62%).
What's not to love?
Well for starters, outside of his fantasy production numbers, Knox's raw volume numbers all ranked 15th among tight ends or worse.
Knox's nine receiving touchdowns were good for second-most in the league, as were his 10 targets in goal-to-go situations. However, his impressive touchdown volume resulted in him finishing as the overall TE9 in points per game last season (10.9) — a finish that was fine for a player who was going as the TE29 in PPR leagues last season.
Now, Knox is being drafted as the overall TE10, which is likely his ceiling for 2022.
Despite playing in a high-volume passing offense, Buffalo still has Stefon Diggs lining up on the outside and emerging wide receivers Gabriel Davis and Isaiah McKenzie — who have both received rave reviews during camp. Buffalo also drafted James Cook to serve as yet another option in the passing game, which could put Knox in a position to be third or fourth on the target tree.
I have no doubt Knox will still have his big games — he went for 18+ fantasy points in four games last season. But he still only managed a top-12 finish in 47% of his games, making the decision to start or sit him a weekly headache in redraft leagues.
The RotoViz linear regression model suggests Knox outscored his predicted touchdown total by 4.99 touchdowns — the most of any tight end last season. He's been an easy fade for me all offseason and will continue to be at this point.
DK Metcalf (SEA)
Last season, DK Metcalf finished with a touchdown differential of +5.27 — finishing as the fourth-most efficient touchdown scorer behind Ja'Marr Chase, Cooper Kupp and Mike Evans. His WR23 finish in points per game (14.4) was a steep drop from the WR9 finish he enjoyed in 2020.
The difference between Metcalf and those three players heading into 2022 is Seattle downgraded from Russell Wilson to Drew Lock/Geno Smith this offseason while the quarterback situations of the other receivers remain unchanged.
Whether it's Lock or Smith under center to start the season remains unknown, but it probably doesn't matter much either way. The careers of both quarterbacks appear to be circling the drain, with a fourth-place finish in the NFC West likely signaling the final shot for at least one of the signal-callers.
Lock, who was acquired in the offseason trade that sent Wilson to Denver, is a 59.3% career passer and has a career EPA/play of 0.017, which ranks 35th among 52 qualified quarterbacks. His completion percentage over expected of -4.1% ranks 46th.
Metcalf has gotten by on years of efficient touchdown production. Living in D.J. Moore's opposite world, you can see his experience under Wilson has yielded results Moore could only dream of.
The public has been consciously fading Metcalf this offseason. Going as the WR6 in PPR leagues just a year ago, he's currently being drafted as the WR19 in that format. Despite the efforts to move off the hyper-efficient touchdown producer, Metcalf's WR19 price in an offense led by Lock/Smith could still be too steep.
My expectations for Seattle are low in 2022. If I were interested in any part of this receiving game, I'd be drafting tight end Noah Fant (TE17) or wide receiver Tyler Lockett (WR42), who are much better values.
2020 version of J.K. Dobbins (BAL)
The fantasy community lost J.K. Dobbins to a torn ACL before the 2021 season even got underway. He's been doing individual drills and isn't a lock to be ready for Week 1, but drafters have been eager to select the talented back playing in one of the league's most run-heavy offenses.
Currently going as the RB22 in drafts (a price that will likely go up in the coming weeks), Dobbins' value in 2022 is one I've wrestled with.
As I mentioned earlier, it's easy to be emotional about players who led us to glory or subsequently burned us in any given season, and in 2020 Dobbins was a near league-winner.
Through the first 10 weeks of his rookie season, Dobbins battled with incumbents Mark Ingram and Gus Edwards for playing time. He averaged an underwhelming 7.4 points per game as a result, finishing twice as a top-24 back.
In Week 11, however, something changed. Dobbins received double-digit touches in each of the final six games of the season, finishing as a top-24 back in every contest including two RB1 finishes. His 16.9 points per game over that span was good for 11th-best.
Despite seeing double-digit touches in each of his final six games, Dobbins totaled 100+ yards from scrimmage just twice and caught a grand total of three passes on four targets. Baltimore hates throwing to its running backs. Since Lamar Jackson took over as the full-time starter in 2019, the Ravens rank dead last in running back targets per game at 3.6.
The story behind Dobbins' strong 2020 finish lies in his touchdown production.
In Weeks 11-17, Dobbins found the end zone seven times, more than doubling the linear regression prediction of 3.37 touchdowns on 77 carries. Even in a run-heavy offense, it seems unlikely Dobbins would average a touchdown per game in 2022.
Ignoring the fact he'll share the backfield with at least one running back for much of the season, we also can't ignore Jackson's rushing abilities.
Since 2018, Jackson has accounted for 31% of Baltimore's rushing attempts, 50.8% of its rushing yards, and 34.4% of its touchdowns. This is by no means a death blow to Dobbins, but it could certainly cap his ceiling this upcoming season.
Dobbins also benefitted from his efficient running — his 6.0 YPC as a rookie led the league. That's likely to take a dip as well.
Assuming he makes a mostly healthy return, I'd expect Dobbins to offer a decent return on overall points. But on a per-game basis — the only measurement of fantasy points that actually matters — I could see him finishing closer to his current RB22 ADP in PPR leagues.