NFL

Fantasy Football: The Top Touchdown Regression Candidates for 2022 (Wide Receivers)

Which wide receivers should score more often or less often in 2022? Here's what the data has to say.

Touchdowns are one of the major components of fantasy football, and I don't mean just that they are one of only a few ways to generate fantasy points. I mean that they matter a lot.

Sort a fantasy football leaderboard by yardage, and you'll see a pretty tight dispersal overall. It's touchdowns that really shake up the fantasy standings (especially in standard and half-PPR leagues).

If a player finds paydirt at a position-high rate, he's going to have a great fantasy season even if the other stats are lagging behind. And, of course, there are always productive players from a yardage standpoint who just don't score much in a 17-game season (whether it's randomness or their offense or their role).

But one truth remains: players generally regress year to year, especially in the touchdown column, and that should help us as we build our fantasy football and best ball lineups.

Here are the biggest touchdown regression candidates among wide receivers for 2022. If you don't care about the process or its accuracy, just skip ahead to the results.

But First: The Process

I'll keep this brief, but here's how I'm looking at things.

If you know a player's receiving yardage output, you -- generally -- know a good deal about his touchdown output. After all, it's not like we see 10-touchdown seasons from guys with 300 yards and 1,400 yards from guys with no red zone role.

Mathematically speaking, the R^2 between yardage and touchdowns among receivers since 2016 is 70.5%. That just means receiving yardage explains more than 70% of touchdown output.

That R^2 climbs to 76.2% if we compare receiving touchdowns to our Reception Net Expected Points (NEP) metric, which accounts for things like field position.

We can then use a model from there to find expected touchdowns (one based on yards and one based on Reception NEP and then average them out) and compare them to actual touchdowns scored.

Does It Work, or Is It Just Fancy?

It works, and it's fancy (sorta).

If you (or me -- I'll do it for us) were to average out expected touchdown totals from yardage and Reception NEP (which are quite close to one another but not exact) and compare it to a player's actual output, you can find touchdowns over expectation (or touchdowns below expectation if the model says they should have scored more often than they actually did in a given season).

If we look at players with at least 35 targets in a season and at least 35 targets in a follow-up season since 2016, we get 389 receiver seasons to examine.

Among those 389 seasons, we see 65 instances in which a receiver scored at least 2.0 more touchdowns than the combined model suggested he should've. This group had an average touchdown rate (receiving touchdowns per target) of 9.0% (nearly double the position average of 4.8%).

Of those 65 overachievers, only 3 (or 4.6%) increased their touchdown rate the following season. Those that did -- Adam Thielen in 2020, Mike Evans in 2021, and Stefon Diggs in 2019 -- did so by 0.5 percentage points or fewer. So, they effectively just stayed the same and certainly didn't build on it much at all.

The full 65-receiver sample fell by an average of 4.1 percentage points (back down to a sample average of 5.4%). These are (or were for one season, at least) elite touchdown scorers, and while their touchdown rate fell back toward the NFL average (again, 4.8%), they still were a tick above. They just weren't on another planet. That's the takeaway.

On the flip side, we have 53 players fall short by 2.0 expected scores. Of them, 75.5% increased their touchdown rate the following season, and the sample did so by an average of 1.9 points back to the position average level (4.5%).

That's a little less than we saw with the overachievers (5.4%), but this also tracks. Littered at the bottom of this list are slot and smaller receivers such as Cole Beasley, Pierre Garcon, Danny Amendola, Jarvis Landry, and T.Y. Hilton.

This suggests the underachievers should be back around the position average this year, and overachievers should still be a tick above that mark.

All right, then, who are they?

Positive Touchdown Regression Candidates

Or, you know, progression candidates. It's not worth getting into here. We just want the goods.

Name xTD
(Rec NEP)
xTD
(Rec Yards)
xTD
(Average)
TDs
Scored
TD
Over
Expected
Laviska Shenault 3.68 3.83 3.76 0 -3.76
Kenny Golladay 3.19 3.22 3.20 0 -3.20
D.J. Moore 7.03 7.22 7.13 4 -3.13
Chase Claypool 4.88 5.35 5.12 2 -3.12
Jakobi Meyers 4.66 5.39 5.03 2 -3.03
Cole Beasley 3.58 4.30 3.94 1 -2.94
Quez Watkins 3.75 4.01 3.88 1 -2.88
Courtland Sutton 4.75 4.82 4.79 2 -2.79
Jerry Jeudy 2.65 2.88 2.76 0 -2.76
Kadarius Toney 2.72 2.58 2.65 0 -2.65
A.J. Green 5.60 5.28 5.44 3 -2.44
Deebo Samuel 7.89 8.79 8.34 6 -2.34
Darnell Mooney 6.02 6.58 6.30 4 -2.30
Zay Jones 3.07 3.37 3.22 1 -2.22
Adam Humphries 1.89 2.35 2.12 0 -2.12
Tyler Johnson 1.93 2.20 2.07 0 -2.07
Rashod Bateman 2.92 3.18 3.05 1 -2.05


The top of the list features two players with no scores in 2021: Laviska Shenault and Kenny Golladay.

Shenault, whose best-ball average draft position (ADP) is just 237.4, is a pretty solid example of regression: he had 5 scores on 600 yards in 2020 (overperforming expectation by around 3.5 scores) and then had 619 yards with no touchdowns in 2021 despite 21 more targets (79 to 100). His 2020 scoring rate of 6.3% was too high, and he fell back. Now, the data says he should have better luck after it would've anticipated a decrease for him in 2021.

Golladay has seen touchdown totals of 3, 5, 11, 2, and 0 since 2016 and ultimately is 1.92 touchdowns under expectation in that span. Specifically from last season, he was a big outlier in the touchdown department, but he's not the only New York Giant on this list. Kadarius Toney made the unfortunate cut, as well.

D.J. Moore is a perennial touchdown regression candidate and has yet to come close to expectation. He has finished -2.28, -3.38, -3.37, and -3.13 touchdowns below expected through four seasons. His best-ball ADP ranks him as the WR24.

Chase Claypool balanced out almost exactly in 2021. Claypool had 9 receiving scores on 5.84 expected touchdowns in 2020 (+3.16) and then 2 on 5.12 expected scores (-3.12) this past season. That number should tick back up toward the position average in 2022.

We were all in on Jakobi Meyers touchdown watch last season. Meyers had not scored in his first two seasons and ultimately hit paydirt twice in 2021. Overall for his career, he's down 9.36 touchdowns from where he should be.

Negative Touchdown Regression Candidates

Or, you know, regression candidates.

Name xTD
(Rec NEP)
xTD
(Rec Yards)
xTD
(Average)
TDs
Scored
TD
Over
Expected
Mike Evans 8.25 6.46 7.35 14 6.65
D.K. Metcalf 6.66 6.03 6.35 12 5.65
Adam Thielen 5.11 4.51 4.81 10 5.19
Allen Lazard 4.07 3.17 3.62 8 4.38
DeAndre Hopkins 4.55 3.54 4.04 8 3.96
Ja'Marr Chase 9.40 9.10 9.25 13 3.75
Cooper Kupp 13.00 12.20 12.60 16 3.40
Jauan Jennings 2.14 1.71 1.93 5 3.07
K.J. Osborn 4.36 4.06 4.21 7 2.79
Amari Cooper 5.43 5.38 5.41 8 2.59
Hunter Renfrow 6.62 6.47 6.55 9 2.45
Randall Cobb 2.82 2.30 2.56 5 2.44
Mack Hollins 1.87 1.34 1.61 4 2.39
Greg Ward 0.68 0.53 0.61 3 2.39
Gabriel Davis 4.08 3.39 3.74 6 2.26


As a reminder, Mike Evans (WR8 by FanDuel best-ball ADP) did already in his career beat regression to maintain an elevated scoring rate this past season (his touchdown rate was 11.9% in 2020 and then 12.3% in 2021), but that's a huge outlier rate. He has now scored 12.89 more times the past two years than the underlying data would assume, including 7.35 more this past season.

But what does this mean, practically speaking? Let's think through it. Let's give him some props for playing in a good offense and being a scoring threat (his 12 targets from inside the 10, tied for fifth in football) and say that, instead of 7.35 scores, he scored 10 times instead of 14 times (that's quite generous).

That's still 24 fantasy points he'd be missing out on if he scored at a more reasonable rate. That optimistic outlook alone would have taken him from WR8 to WR15 in half-PPR formats last season. If he scored 7 times, we're looking at a 42-point decrease, which would've made him the WR21. This stuff matters, and all we're doing is considering a league-average scoring rate for Evans.

With a quarterback change and an unsustainable touchdown rate (9.3%), we should think long and hard about drafting D.K. Metcalf this season. He's currently the WR17 in FanDuel best-ball formats. A perennial overachiever in the touchdown column, Metcalf will have his work cut out for him while playing in a less efficient offense without Russell Wilson under center.

Did someone say touchdown regression? Adam Thielen's never heard of it. This season, Thielen (WR19 in best-ball ADP) converted 6 of 8 targets from inside the 10 and 7 of 13 targets from in the red zone overall into scores and once again notched double-digit touchdowns (14 in 2020 and 10 in 2021).

He's up to 16.12 more scores than anticipated over the past three seasons while posting a double-digit touchdown rate in all three seasons. Without an outlier scoring rate, Thielen likely lacks the yardage upside (418, 925, and 726 the past three years) to make a huge difference in fantasy formats, so this is definitely a situation to monitor.

Allen Lazard's season-by-season yardage outputs: 477, 451, 513. His touchdown tallies: 3, 3, 8. How? His red zone targets climbed: 3, 6, 14. That helps. Still, the underlying data doesn't lie very often, and that data says he had too many. Of course, a substantial role exists with Davante Adams out of the picture. This is a good reminder not to cross off all these names on this list -- just maybe adjust expectations a bit.

We, as a consensus, aren't going to be low Ja'Marr Chase (WR2) or Cooper Kupp (WR1). I get that. Elite talents can perform at outlier levels. Just don't be surprised if some touchdowns wind up going to other players based on how good they were in that department in 2021.