2024 WR/TE Positive Regression Candidates
Welcome to the second piece of the Fantasy Football Cookbook! Before we get into the thick of it here I first wanted to say thank you for checking out my work and that I will be releasing at least 2 articles a week, much like this one, from now until the start of the season.
Writing these pieces is just as, if not more, helpful for me as it is for you. It’s a great thought exercise trying to articulate trends you notice and takes that you have. I enjoy making this content for you all and engaging in discussions from it, if you like what you see, I would greatly appreciate you subscribing (totally free!) to the newsletter and even dropping me a follow on Twitter!
Now let’s talk Fantasy Football Positive Regression Candidates.
What is Positive Regression?
Instead of leading off with giving a textbook definition of Positive Regression, let’s give a simple example.
Player X has averaged 1.02 Fantasy Points per Target over their last 48 games played with Player Y at QB, but over their next 6 games, Player X only averages 0.54 Fantasy Points per Target with Player Y as their QB, a difference of 0.48 Fantasy Points.
You may see this and come to the conclusion ‘Oh Player X sucks now, look how bad he is doing compared to his career average’ which could be the case, but the way we want to attack it is ‘Oh wow look at Player X, based on his career average he has been underperforming and looks due for a breakout game once he finally regresses to the mean.’
So essentially, a Positive Regression candidate is someone who has performed below their expected outcome based on a certain metric, who we believe should regress to the mean over time. Today, we are using volume-based metrics to identify these candidates for Positive Regression.
What Metrics Are We Using?
The two main metrics we will be using in this article to identify Positive Regression candidates are Targets per Game and Air Yards per Game. Some of you may not be familiar with Air Yards, so here is exactly what these two metrics mean.
Targets per Game: The total amount of targets a player gets divided by the total number of games they have played. Ex; Player X has gotten 69 targets over his last 10 games, giving Player X an average of 6.9 Targets per Game.
What are Air Yards?
To give a quick rundown for those unfamiliar, Air Yards are the total yards a ball travels in the air before a completion or incompletion, if Tyreek Hill gets a target where the ball traveled 35 yards downfield in the air, he gets credited with 35 Air Yards, regardless of if he caught the ball and then ran for 16 yards or didn’t catch the pass at all. We use Air Yards to show us how deep a player is getting targeted downfield, and the deeper a ball is thrown downfield, the higher the upside of the target.
Air Yards per Game: The total amount of Air Yards a players gets divided by the total number of games they have played. Ex; Player Y has totaled 694 Air Yards over his last 10 games, giving Player Y an average of 69.4 Air Yards per Game.
The idea behind using these two metrics to identify Positive Regression candidates is that volume is highly correlated to Fantasy Points Scored, and players with a high amount of targets and Air Yards are players who are getting a lot of volume, and volume equals more opportunity, and with opportunity comes fantasy points.
Using These Metrics to Identify Positive Regression Candidates
Target and Air Yards data is provided by FantasyLife and can be found here
First, let’s take a look at whether these metrics even correlate to fantasy points, what I did here is I took every WR/TE with at least 35 targets last year and then calculated their Targets per Game (Tar/G) and Air Yards per Game (AirYds/G) and tabulated them to give each player their adjusted Targets per Game & Air Yards Per Game (aTar + AY/G) score.
Formula for aTar + AY/G Score : AirYds/G*0.10 + Tar/G
This data set had a total of 100 WRs and 40 TEs, so I broke down the WRs into 5 groups of 20 and the TEs into 4 groups of 10. I then sorted by their aTar + AY/G and added up each player grouping’s *Fantasy Points per Game (FPPG) to see if where they finished in FPPG had any correlation to their aTar + AY/G. To be clear, the Top WR group consists of the total fantasy points scored by the Top 20 WRs in aTar + AY/G, the second group would be the Top 40, and so on. The same thing applies to TEs but for groups of 10. After doing that, we then average those totals to get an average score per grouping.
*Since we are trying to find volume-based positive regression candidates, I did not use TDs in totaling a player’s FPPG, so what you see for a player’s FPPG is taken solely from their receptions + receiving yards in PPR format, for every TD a player scored I subtracted 6 from their total fantasy points that included TDs.
So as you can see above, the Top 20 Ranked WRs in aTar + AY/G averaged 13.63 FPPG, while the Top 10 Ranked TEs averaged 10.31 FPPG. You should also notice that with each tier we go down, both WRs and TEs begin to average less FPPG. Proving to us that in a vacuum, the more volume/opportunity a player gets, the more FPPG they average.
Now that we have that established, let’s take a look at how we are using this data to identify our Positive Regression Candidates.
To find players due for Positive Regression, we are taking a player’s aTar + AY Rank (where the player ranks out of the whole data set in aTar + AY/G and subtracting it from their PPR Rank (where the player ranks out of the whole data set in FPPG.)
We then assigned the difference in those rankings to each player and ranked them based on how big that difference is. The lower a player’s PPR Rank compared to their aTar + AY Rank, the better a candidate they are for positive regression.
Since we already established that aTar + AY/G is correlated to a player’s FPPG, we can assume that a player who finished the season noticeably lower in PPR Rank compared to their aTar + AY Rank is a candidate for Positive Regression.
Now that we have ALL of that out of the way, it is time for the fun part, identifying players due for Positive Regression based on these metrics!
Positive Regression Candidates
Before we dive into these candidates, I want to make a note that a situation changing does alter me including a player on this list, for example, Jonathan Mingo has the 8th best Positive Regression Rank (PR Rank) out of all WRs, but Carolina drafted Xavier Legette and traded for Diontae Johnson, so projecting Mingo for the similar volume in 2024 compared to 2023 is not logical. This does work both ways though, even with Zay Jones now being on the Cardinals, his role in Arizona looks to be very similar to his role in Jacksonville, so I have no qualms with including him despite him being on a different team as his situation is somewhat similar.
2024 Position Rank is a player’s ranking at their respective position. Ex; CeeDee Lamb is currently the first WR drafted on Underdog so his Position Rank is WR1. Position Rank is based on current Underdog Best Ball ADP.
Passing Attempts per Game + Plays per Game Data Provided by TeamRankings.com
Kyle Pitts (ATL):
2023 aTar + AY/G Finish: #5 - 2023 FPPG Finish: #17 - Difference: -12
2024 Position Rank: TE6
Another year, another off-season of us trying to talk ourselves into Kyle Pitts, trust me, I get it. The difference this time is the Falcons have a new HC in Raheem Morris, an OC coming from the McVay coaching tree in Zac Robinson, and arguably most importantly, a new QB in Kirk Cousins.
Everything is pointing up for Pitts and the Falcons’ offense as a whole this year, partner that with the fact Pitts only finished as the TE17 in adjusted FPPG, the reasons he is a candidate for Positive Regression become clear.
With the additions of Zac Robinson and Kirk Cousins, the Falcons should finish higher in Passing Attempts per Game, a stat in which they ranked 25th in 2023, and with no major additions to the pass-catching core, Pitts looks like he’ll be getting even more volume than he did last year.
The only downside here is you are not getting a discount on his prior lack of performance, but Pitts has the upside to finish as the TE1 overall this year when you take everything into consideration, making him my favorite candidate for Positive Regression for the 2024 Fantasy Football season.
Zay Jones (ARI):
2023 aTar + AY/G Finish: #21 - 2023 FPPG Finish: #59 - Difference: -38
2024 Position Rank: WR81
As I mentioned previously, looking for Positive Regression candidates becomes a little murky when the player at hand is going to a new team and or projects to command less volume due to the addition of other players at his position, but if the situation still looks close to the same despite moving teams, I am still willing to call a shot on a player having an improved season. Zay Jones is the epitome of this.
Jones’ only competition for commanding the WR2 spot on the Cardinals is rookie Michael Wilson, a player Arizona selected in the 3rd round BEFORE the team signed Jones, and Greg Dortch, a guy who has only commanded 108 targets through his first three seasons in the NFL, so I am willing to bet on Zay carving out a role in this offense.
No player tracked with this data had a bigger difference in where they finished in FPPG compared to the amount of volume they received, and even if he didn’t positively regress, he’ll be finishing higher than the WR81 where he is being drafted if he stays healthy. You’re not taking on any risk in drafting a guy as late in a draft as Jones is going and we pounce on the opportunities to take players with upside with little to no risk.
Davante Adams (LV):
2023 aTar + AY/G Finish: #4 - 2023 FPPG Finish: #16 - Difference: -12
2024 Position Rank: WR13
Thinking a player who finished as a borderline WR1 in fantasy last year is a candidate for Positive Regression may seem counterintuitive but it’s not, you can have a great season that could’ve been even better and that is the boat the Davante Adams fall into.
Adams is the WR13 in my Positive Regression rankings based on his -12 difference in where he finished in aTar + AY/G compared to FPPG, which is the highest discrepancy for any player who averaged at least 10 targets a game in 2023. To put it into perspective, only 4 WRs averaged at least 10 targets a game last year and none of them ranked higher than the WR52 in Positive Regression rank. Not only that, but without including Adams, the three other WRs who averaged 10 targets a game (Tyreek Hill, Keenan Allen, and CeeDee Lamb) averaged 18.20 FPPG, and don’t forget TDs are not included in our FPPG average, compared to Adams’ 12.32 FPPG, a difference of 5.88 FPPG.
There is no reason to think Adams won’t command similar volume again this year as the Raiders didn’t add anyone who should eat into his targets besides Brock Bowers, who is a rookie TE. Being able to get one of four WRs who commanded at least 10 targets a game last year at WR13 is a good problem to have, especially when the two other players who did that and didn’t change situations in Tyreek Hill and CeeDee Lamb are going as the WR1 and WR2.
Chris Olave (NO):
2023 aTar + AY/G Finish: #6 - 2023 FPPG Finish: #18 - Difference: -12
2024 Position Rank: WR12
Olave is in a very similar situation to Davante Adams, he’s tied with Adams with a -12 difference in where he finished in volume ranking compared to FPPG, and also like Adams, he remains the clear #1 pass catcher in his offense.
While Olave only averaged 8.63 targets a game compared to Adams’ 10.06, they were both two of only six players to finish with an aTar + AY/G of over 20.00, with the other four players being Justin Jefferson, Keenan Allen, Tyreek Hill, and CeeDee Lamb, which is pretty good company to be in.
Players who scored at least 20.00 in aTar + AY/G last year averaged 16.01 FPPG while Olave averaged only 12.14 FPPG, a difference of 3.87 FPPG, making him an amazing candidate for Positive Regression.
I wouldn’t say you are getting Olave at the biggest discount at his current ADP, but I think you can confidently draft him as the WR12 with the upside of someone who could finish as a Top 5 fantasy WR.
Terry McLaurin (WAS):
2023 aTar + AY/G Finish: #20 - 2023 FPPG Finish: #30 - Difference: -10
2024 Position Rank: WR32
Terry is a great example of someone whose situation changed but in a positive way. He now has Jayden Daniels instead of Sam Howell throwing him the ball and Washington didn’t add anybody that should cut into Terry’s volume, so he looks to project for similar usage as last year.
It is somewhat concerning that Washington tied Dallas for most passing attempts per game at 37.4, meaning that the Commanders passing the ball even more than last year, by a meaningful amount, is off the table. However, Washington only ranked 18th in plays per game last season and ranked almost dead last in the final four games of the season, so we are hoping to see more sustained drives where the defense can’t simply hone in on stopping the pass because they know it’s coming.
I won’t argue that the situation on its own might not be enough to make him a good Positive Regression candidate, so here is a stat for you. Out of the 26 WRs to average at least 7.00 targets a game last year, McLaurin finished 24th in FPPG, trailed by only DeAndre Hopkins and Calvin Ridley. Now if we take a look at the WRs who averaged at least 7.00 targets a game but less than 10.00 targets a game, we find that those players on average scored 12.56 FPPG, while McLaurin only scored 10.31 FPPG, a difference of 2.25 FPPG, to really put it in perspective, that means that Terry scored almost 20% less fantasy points than players with similar volume.
At WR32, McLaurin is already going later than the WR30 he finished at last year, and a WR30 finish on the volume of the WR20 to boot. Him turning 29 early this season may be a valid cause of concern, but it feels like that is already baked into his ADP.
DeAndre Hopkins (TEN):
2023 aTar + AY/G Finish: #11 - 2023 FPPG Finish: #31 - Difference: -20
2024 Position Rank: WR45
Hopkins averaged 7.82 targets per game last year, and with the off-season addition of Calvin Ridley, I can see how people would expect his volume to dip, however, I disagree. No team attempted fewer passes a game than the Titans 29.1 last year and not only that, no team averaged less than the Titans 58.9 plays per game besides the Seahawks. Will Levis locked into Hopkins last year once he took over the reins too, DHop averaged 8.44 targets per game in the 9 games Levis played last year, a difference of +0.62 targets per game.
The only player to average less FPPG last year while getting at least 7.00 targets a game than Hopkins was his new teammate Calvin Ridley, with Hopkins averaging 10.27 FPPG and Ridley averaging 10.11 FPPG. Even though both of them are valid candidates for Positive Regression, I am giving the edge to Hopkins.
Out of the 11 players who got at least a 19.00 in aTar + AY/G score, Hopkins was the only WR not to rank in the Top 20 in FPPG, and not only was Hopkins not in the Top 20, he wasn’t even Top 30 as he finished 31st in FPPG. Taking all of this into consideration alongside the fact he finished 11th in aTar + AY/G last year and is currently being drafted as the WR45(!!!), everything is pointing towards him being a great Positive Regression candidate who is also undervalued.
Christian Watson (GB):
2023 aTar + AY/G Finish: #24 - 2023 FPPG Finish: #54 - Difference: -30
2024 Position Rank: WR47
This one isn’t as cut and dry as the other ones as Green Bay has four WRs who are all good at football in their own right, but there is no way around the fact that when Watson was on the field last year, he was the #1 WR on the Packers. Watson’s 5.89 targets per game lead the team and his team-leading 98.11 Air Yards per game was almost 30 yards more than the next highest WR which was Romeo Doubs who had 68.59.
Out of the players who were Top 20 in Air Yards per game, only Zay Jones’ 7.12 FPPG ranked lower than Watson’s 7.37 FPPG. That may not sound like something that should drive a needle, but once you figure out that no other player in the Top 20 averaged less than 10.00 FPPG, him being a Positive Regression candidate really starts to make sense. Here is a stat to really drive the point home, players ranked in the Top 20 in Air Yards per game scored an average of 12.57 FPPG, so Watson scored 5.20 FPPG less than his fellow Air Yards earning constituents.
Although the path to Watson busting or underperforming again is apparent. Injury risk aside, getting the WR24 in aTar + AY/G who scored around 40% less FPPG than other players with similar volume at WR47 feels like a great value.
Diontae Johnson (CAR):
2023 aTar + AY/G Finish: #30 - 2023 FPPG Finish: #40 - Difference: -10
2024 Position Rank: WR44
Diontae went from competing against George Pickens for targets (and winning) to competing against rookie 1st round pick Xavier Legette and Adam Thielen who will be 34 years old by the time the Week 1 games kickoff, so I am not too concerned about his new situation being vastly worse for his potential volume.
Johnson still managed to finish 30th in aTar + AY/G last year despite being on the Steelers who ranked 27th in pass attempts per game and even though new Panthers HC Dave Canales has let it be known he wants to run the ball, with how bad the Carolina defense projects to be, the game scripts shouldn’t allow for the Panthers to have the luxury of not being forced to pass.
Of the 30 WRs to score at least 15.00 in aTar + AY/G last year, Diontae is one of only 5 WRs who didn’t score at least 10.00 FPPG, with the other four being Zay Jones, Christian Watson, Marquise Brown, and Tyler Lockett. I would argue that Diontae Johnson is the most talented WR out of that group as well, by a decent margin too.
Getting him as the WR44 in drafts this year may not feel like a great value, but Diontae could average even less volume than the year before and still outperform that ranking, Pittsburgh’s QB situation last year was abysmal and I am willing to bet that Bryce Young is a better QB than Kenny Pickett and Mason Rudolph, and so should you.