Last week I analyzed what fantasy production we can expect from A.J. Green this year in light of how many games he’s missed the past three years. This week, I want to expand that analysis to all veteran WRs.
You might want to read that linked article about Green first, but in expanding the analysis I found a few places to improve on the tables in that article. As a reminder, I’m using point-per-reception (PPR) scoring here, with 10 yards = 1 fantasy point (FP), TDs = 6 FP, and receptions = 1 FP. Also, this study only deals with WRs who will be 24 or older in 2019 (age = 2019 minus birth year) who already have at least three years of NFL experience.
The next table is the first of three similar tables I’ll use in building today’s analysis:
All WR FPG, 1988-2018
|
Age Group |
Games Played Previous 3 Seasons |
|||
|
36 or less |
37-43 |
44-47 |
All 48 |
|
|
24-26 |
6.7 |
8.8 |
10.1 |
10.9 |
|
27-29 |
5.8 |
7.6 |
9.2 |
10.6 |
|
30-32 |
6.2 |
7.8 |
9.0 |
10.5 |
|
33-35 |
6.8 |
7.0 |
9.2 |
12.1 |
|
36+ |
10.1 |
9.9 |
8.0 |
7.8 |
I’m using the same Age Groups as last week, but I’ve cut down on the Games Played categories. In looking over the data, I found break points in FP/G performance between 36 and 37 GP (over the previous three years), between 43 and 44 GP, and 47 and 48 GP. So these divisions are still subjective, but conform better to the data than in my previous study.
As I pointed out last week, if you go from left to right across all columns, the FP/G numbers almost invariably get bigger. Meaning that it appears that previous injuries (as measured by total games played in a three-year period) affect future performance. Sample size issues in the 36+ Age Group and explain the exception to the rule there. Going down each column, the numbers generally get smaller, but samples sizes and survivor bias create more exceptions to the rule.
While the first table is for all WRs in the period for 1988-2018 who had played in all three previous seasons, the next table narrows the WR population down to just those who had at least one Top 60 finish in total FP in the previous thee seasons:
WR With Top-60 Finish, Previous 3 Years, FPG 1988-2018
|
Age Group |
Games Played Previous 3 Seasons |
|||
|
36 or less |
37-43 |
44-47 |
All 48 |
|
|
24-26 |
8.9 |
10.1 |
11.1 |
11.6 |
|
27-29 |
8.1 |
9.4 |
10.4 |
11.1 |
|
30-32 |
8.6 |
9.3 |
10.3 |
11.6 |
|
33-35 |
8.1 |
8.6 |
10.1 |
12.1 |
|
36+ |
10.4 |
9.9 |
8.6 |
8.1 |
The same patterns we saw in the first table still apply, although note the highlighted 8.9 FP/G in the 24-26 Age Group Pattern. I have subjectively “smoothed” this data to account for the small sample size: relatively few young WRs miss substantial time and I chose to make this number conform better with the overall patterns.
The third table further narrows the WRs down to just those with at least one Top 20 ranking in total FP in the preceding three years:
WR With Top-20 Finish, Previous 3 Years, FPG 1988-2018
|
Age Group |
Games Played Previous 3 Seasons |
|||
|
36 or less |
37-43 |
44-47 |
All 48 |
|
|
24-26 |
12.5 |
13.1 |
14.7 |
14.5 |
|
27-29 |
10.8 |
12.2 |
13.6 |
13.6 |
|
30-32 |
10.8 |
11.8 |
13.0 |
12.8 |
|
33-35 |
9.8 |
10.5 |
10.5 |
13.0 |
|
36+ |
11.7 |
9.6 |
9.1 |
8.5 |
Again, I have chosen to “smooth” the data, in two cases this time.
Using these tables, I generated a prediction of “Injury-Adjusted FP/G” for all WRs who have played the past three seasons. In the last table, I used those numbers to then create an Expected FP/G for those players in 2019:
|
Player |
Best |
AGE |
GP |
FP/G |
Inj Adj |
Exp |
|
Davante Adams |
Top 20 |
27-29 |
44-47 |
21.8 |
13.6 |
19.1 |
|
Antonio Brown |
Top 20 |
30-32 |
44-47 |
21.6 |
13.0 |
18.7 |
|
DeAndre Hopkins |
Top 20 |
27-29 |
44-47 |
21.1 |
13.6 |
18.6 |
|
Tyreek Hill |
Top 20 |
24-26 |
44-47 |
20.5 |
14.7 |
18.6 |
|
Michael Thomas |
Top 20 |
24-26 |
44-47 |
20.0 |
14.7 |
18.2 |
|
Julio Jones |
Top 20 |
30-32 |
44-47 |
20.6 |
13.0 |
18.1 |
|
Adam Thielen |
Top 20 |
27-29 |
All 48 |
19.2 |
13.6 |
17.3 |
|
Mike Evans |
Top 20 |
24-26 |
44-47 |
17.9 |
14.7 |
16.8 |
|
Odell Beckham |
Top 20 |
27-29 |
36 or less |
19.4 |
10.8 |
16.5 |
|
Stefon Diggs |
Top 20 |
24-26 |
37-43 |
17.6 |
13.1 |
16.1 |
|
T.Y. Hilton |
Top 20 |
30-32 |
44-47 |
17.1 |
13.0 |
15.7 |
|
Robert Woods |
Top 20 |
27-29 |
37-43 |
16.6 |
12.2 |
15.1 |
|
Brandin Cooks |
Top 20 |
24-26 |
All 48 |
15.2 |
14.5 |
15.0 |
|
Emmanuel Sanders |
Top 20 |
30-32 |
37-43 |
16.5 |
11.8 |
14.9 |
|
Tyler Boyd |
Top 20 |
24-26 |
37-43 |
15.8 |
13.1 |
14.9 |
|
A.J. Green |
Top 20 |
30-32 |
36 or less |
16.8 |
10.8 |
14.8 |
|
Julian Edelman |
Top 20 |
33-35 |
36 or less |
17.3 |
9.8 |
14.8 |
|
Amari Cooper |
Top 20 |
24-26 |
44-47 |
14.6 |
14.7 |
14.7 |
|
Keenan Allen |
Top 20 |
27-29 |
36 or less |
16.3 |
10.8 |
14.4 |
|
Tyler Lockett |
Top 20 |
27-29 |
44-47 |
13.8 |
13.6 |
13.7 |
|
Jarvis Landry |
Top 20 |
27-29 |
All 48 |
13.6 |
13.6 |
13.6 |
|
Will Fuller |
Top 60 |
24-26 |
36 or less |
15.2 |
8.9 |
13.1 |
|
Alshon Jeffery |
Top 60 |
27-29 |
37-43 |
14.3 |
9.4 |
12.6 |
|
Marvin Jones |
Top 20 |
27-29 |
37-43 |
12.9 |
12.2 |
12.6 |
|
Robby Anderson |
Top 20 |
24-26 |
44-47 |
11.5 |
14.7 |
12.5 |
|
Golden Tate |
Top 20 |
30-32 |
44-47 |
12.1 |
13.0 |
12.4 |
|
Albert Wilson |
Top 60 |
27-29 |
36 or less |
13.8 |
8.1 |
11.9 |
|
Doug Baldwin |
Top 20 |
30-32 |
44-47 |
10.9 |
13.0 |
11.6 |
|
Adam Humphries |
Top 60 |
24-26 |
44-47 |
11.8 |
11.1 |
11.6 |
|
Demaryius Thomas |
Top 20 |
30-32 |
44-47 |
10.4 |
13.0 |
11.3 |
|
DeSean Jackson |
Top 60 |
33-35 |
37-43 |
12.6 |
8.6 |
11.3 |
|
Mohamed Sanu |
Top 60 |
30-32 |
44-47 |
11.4 |
10.3 |
11.0 |
|
Allen Robinson |
Top 60 |
24-26 |
36 or less |
11.9 |
8.9 |
10.9 |
|
Sterling Shepard |
Top 60 |
24-26 |
37-43 |
11.3 |
10.1 |
10.9 |
|
Devin Funchess |
Top 20 |
24-26 |
44-47 |
8.8 |
14.7 |
10.7 |
|
Sammy Watkins |
Top 60 |
24-26 |
36 or less |
11.5 |
8.9 |
10.6 |
|
Nelson Agholor |
Top 60 |
24-26 |
44-47 |
10.3 |
11.1 |
10.6 |
|
Geronimo Allison |
Other |
24-26 |
36 or less |
12.5 |
6.7 |
10.5 |
|
Larry Fitzgerald |
Top 20 |
36+ |
All 48 |
11.5 |
8.5 |
10.5 |
|
Josh Gordon |
Top 60 |
27-29 |
36 or less |
11.6 |
8.1 |
10.4 |
|
Jordy Nelson |
Top 20 |
33-35 |
44-47 |
10.3 |
10.5 |
10.4 |
|
Tyrell Williams |
Top 20 |
27-29 |
All 48 |
8.6 |
13.6 |
10.3 |
|
Taylor Gabriel |
Top 60 |
27-29 |
44-47 |
9.6 |
10.4 |
9.9 |
|
Michael Crabtree |
Top 20 |
30-32 |
44-47 |
8.3 |
13.0 |
9.9 |
|
Rashard Higgins |
Top 60 |
24-26 |
44-47 |
9.2 |
11.1 |
9.9 |
|
Cole Beasley |
Top 60 |
30-32 |
44-47 |
9.4 |
10.3 |
9.7 |
|
Ted Ginn |
Top 60 |
33-35 |
36 or less |
10.5 |
8.1 |
9.7 |
|
Randall Cobb |
Top 60 |
27-29 |
37-43 |
9.8 |
9.4 |
9.7 |
|
Jamison Crowder |
Top 60 |
24-26 |
37-43 |
9.2 |
10.1 |
9.5 |
|
Kenny Stills |
Top 60 |
27-29 |
44-47 |
8.8 |
10.4 |
9.3 |
|
John Brown |
Top 60 |
27-29 |
37-43 |
9.0 |
9.4 |
9.1 |
|
Donte Moncrief |
Top 60 |
24-26 |
37-43 |
8.3 |
10.1 |
8.9 |
|
Paul Richardson |
Top 60 |
27-29 |
37-43 |
8.4 |
9.4 |
8.8 |
|
Dontrelle Inman |
Top 60 |
30-32 |
37-43 |
8.5 |
9.3 |
8.8 |
|
Willie Snead |
Top 60 |
27-29 |
37-43 |
8.4 |
9.4 |
8.7 |
|
Danny Amendola |
Top 60 |
33-35 |
37-43 |
8.5 |
8.6 |
8.5 |
|
Marquise Goodwin |
Top 60 |
27-29 |
37-43 |
7.9 |
9.4 |
8.4 |
|
Quincy Enunwa |
Top 60 |
27-29 |
36 or less |
8.1 |
8.1 |
8.1 |
|
Terrelle Pryor |
Top 20 |
30-32 |
36 or less |
6.6 |
10.8 |
8.0 |
|
Seth Roberts |
Other |
27-29 |
44-47 |
7.1 |
9.2 |
7.8 |
|
Chester Rogers |
Other |
24-26 |
37-43 |
7.1 |
8.8 |
7.6 |
|
Bruce Ellington |
Other |
27-29 |
36 or less |
8.5 |
5.8 |
7.6 |
|
Pierre Garcon |
Top 60 |
33-35 |
36 or less |
7.3 |
8.1 |
7.6 |
|
Josh Doctson |
Top 60 |
27-29 |
36 or less |
7.3 |
8.1 |
7.5 |
|
Chris Hogan |
Top 60 |
30-32 |
37-43 |
6.6 |
9.3 |
7.5 |
|
Jermaine Kearse |
Top 60 |
27-29 |
44-47 |
5.7 |
10.4 |
7.3 |
|
Cordarrelle Patterson |
Other |
27-29 |
44-47 |
6.2 |
9.2 |
7.2 |
|
Breshad Perriman |
Other |
24-26 |
37-43 |
6.2 |
8.8 |
7.1 |
|
DeVante Parker |
Top 60 |
24-26 |
37-43 |
5.5 |
10.1 |
7.1 |
|
Brandon LaFell |
Top 60 |
33-35 |
37-43 |
6.3 |
8.6 |
7.0 |
|
Ryan Grant |
Top 60 |
27-29 |
44-47 |
5.3 |
10.4 |
7.0 |
|
Tavon Austin |
Top 60 |
27-29 |
37-43 |
5.6 |
9.4 |
6.9 |
|
Phillip Dorsett |
Other |
24-26 |
44-47 |
5.1 |
10.1 |
6.8 |
|
Demarcus Robinson |
Other |
24-26 |
All 48 |
4.7 |
10.9 |
6.7 |
|
Martavis Bryant |
Top 60 |
27-29 |
36 or less |
6.0 |
8.1 |
6.7 |
|
Jarius Wright |
Other |
30-32 |
37-43 |
6.1 |
7.8 |
6.7 |
|
Jakeem Grant |
Other |
27-29 |
37-43 |
6.0 |
7.6 |
6.6 |
|
Kelvin Benjamin |
Top 60 |
27-29 |
44-47 |
4.6 |
10.4 |
6.5 |
|
Chris Conley |
Other |
27-29 |
37-43 |
6.0 |
7.6 |
6.5 |
|
Aldrick Robinson |
Other |
30-32 |
44-47 |
5.0 |
9.0 |
6.3 |
|
Eli Rogers |
Other |
27-29 |
36 or less |
6.5 |
5.8 |
6.3 |
|
Laquon Treadwell |
Other |
24-26 |
37-43 |
4.7 |
8.8 |
6.1 |
|
Jordan Matthews |
Top 60 |
27-29 |
37-43 |
4.4 |
9.4 |
6.1 |
|
Cody Latimer |
Other |
27-29 |
36 or less |
6.0 |
5.8 |
5.9 |
|
Jaron Brown |
Top 60 |
27-29 |
37-43 |
3.8 |
9.4 |
5.7 |
|
Cameron Meredith |
Top 60 |
27-29 |
36 or less |
4.4 |
8.1 |
5.6 |
|
Brandon Marshall |
Top 60 |
33-35 |
36 or less |
4.4 |
8.1 |
5.6 |
|
Torrey Smith |
Other |
30-32 |
37-43 |
4.4 |
7.8 |
5.5 |
|
Travis Benjamin |
Top 60 |
30-32 |
37-43 |
3.4 |
9.3 |
5.4 |
|
Bennie Fowler |
Other |
27-29 |
37-43 |
4.2 |
7.6 |
5.3 |
|
Maurice Harris |
Other |
27-29 |
36 or less |
5.0 |
5.8 |
5.2 |
|
Tajae Sharpe |
Other |
24-26 |
36 or less |
4.5 |
6.7 |
5.2 |
|
Allen Hurns |
Other |
27-29 |
37-43 |
3.8 |
7.6 |
5.1 |
|
Russell Shepard |
Other |
27-29 |
37-43 |
3.4 |
7.6 |
4.8 |
|
T.J. Jones |
Other |
27-29 |
36 or less |
3.3 |
5.8 |
4.1 |
A brief explanation of the first five columns: “Player” is obvious; “Best Rank” is the best finish of the Player in the previous three seasons – either Top 20, Top 60, or Other; “AGE GRP” is based on the WR’s age in 2019; “GP CAT” is the appropriated games played category for each player based on his total GP in 2016-2018; “FP/G 2018” is self-explanatory.
“Inj Adj FP/G” or Injury-Adjusted FP/G is found by taking the player’s Best Rank, going to the appropriate one of the first three tables and then finding his AGE GRP and GP CAT in that table. For example, Davante Adams has at least one Top 20 finish in 2016-2018, will be 27-years old in 2019 so he falls in the 27-29 AGE GP, and has played in 45 games the last three seasons so he’s in in 44-47 GP CAT. With that, we enter the third (Top 20) table on the 27-29 Age Group row, go to the 44-47 GP column and read 13.6 FP/G. That is the number of FP/G he can be expected to score for his age and injury history.
That bring us to the “Exp FP/G” column of the final table: Expected FP/G for 2019. It turns out that how many FP/G a WR scored last year is a pretty good prediction of how many he’ll score next year. I found that combining last year’s FP/G with the injury-adjusted FP/G produced a (slightly) better prediction, which I’m calling “Expected FP/G.” This number is twice last year’s FP/G plus the Inj Adjusted FP/G divided by three. That is, last year’s FP/G is weighted twice what the injury adjustment counts. Those two numbers combined explain roughly half of actual FP/G scored by WRs since 1988. Which is pretty good since we’re ignoring QBs, other personnel, coaches, offensive identity, etc., etc. (For the numerically-inclined the actual linear regression equation is Actual FP/G = -0.62 + 0.62*Previous Year’s FP/G + 0.35*Inj Adj FP/G, with an R^2 = 0.48; I only used WRs with one Top 100 finish in the previous three seasons in running the regression as I’ve found using too many bottom-feeders in these type of regressions inflates the R^2 with players we don’t care about).
The last table is sorted from most to fewest Expected FP/G. Using this method, Davante Adams is the top-rated VETERAN WR for 2019. I stress “veteran” because this study ignores potential studs like JuJu Smith-Schuster and Calvin Ridley who haven’t been in the league long enough to qualify. And of course rookies are excluded. This analysis is useful for comparing the included WRs to each other, it’s not meant to be a definitive ranking of WRs for 2019.
This method will favor younger WRs over older ones, and recent durability over players who’ve missed a lot games the last three years.
Some comments on specific players I’d like to highlight:
- Antonio Brown: yeah, if he were acting “normal” and returning to the Steelers, he’d clearly be near the top of any list. But right now, who knows what’s up worth him and where he’ll be playing.
- Odell Beckham: he takes a big hit due to his injury history and for moving into the 27-29 age group; together those factors drop him below Adam Thielen and Mike Evans in this ranking.
- AJ Green: between age and injury, this method likes him a hair less than teammate Tyler Boyd, and less than some other WRs you might like less than Green. Personally, I can see Green being a Top 5 guy or not even a Top 25 WR in 2019 and for me this ranking reflects that risk.
- Julian Edelman: the 33-35 Age group is very volatile in the tables. He gets hammered for having missed time (not all due to injury) and his age in the Inj Adj column. But I watched the playoffs too so this feels low.
- Keenan Allen: like Green but with a better (but older) QB. Allen is younger than Green but the table penalizes them both equally for their injury history.
- Robby Anderson: the first WR in the table to really get a boost from his Inj Adj FP/G, I’ll be watching his ADP to see if he makes a nice mid-round pick with upside.
- Devin Funchess: I’ve never liked or owned this guy. But he does have a Top 20 year under his belt and maybe a healthy QB in 2019. He’s got sneaky upside if you believe this method but I get any subjective rejection of him.
You can go through the table and find other players who seem too high or too low: Emmanuel Sanders and Demaryius Thomas are too high based on their late season Achilles injuries which didn’t cost them enough games to affect this ranking. Larry Fitzgerald‘s numbers are based on the least-reliable age group. And on and on. The point of this or any limited-factor model is to identify risks (Beckham, Green) or players you might otherwise overlook (Anderson, Funchess) and get you to think harder about all the factors that go into the other half of their scoring that this model doesn’t account for.
