Going Deep

How Predictive Is Late-Season Production?

by Hayden Winks
Updated On: June 2, 2020, 11:22 am ET

A few weeks ago, I wrote two columns about how each player did at the end of the 2019 season thinking that late-season fantasy production and volume would be helpful for creating my 2020 fantasy football rankings. But how predictive is that late-season production for fantasy football?

The short answer: Late-season fantasy production is useful but because the sample size is smaller, it’s not as predictive as season-long fantasy production.

The long answer: I gathered data from 2010 to 2019 in an attempt to see how correlated the previous season was to the following season and if there’s any value in looking at end-of-season data. In the tables you’ll see below, I have “Entire Season” and “Late Season” labels. “Entire Season” is the entire previous year’s season, while “Late Season” is only the previous season’s data during Weeks 13 through the Super Bowl. The higher the R-Squared, the more useful the stat is at predicting how players will finish in fantasy the following season at their position, specifically using PPR points per game. Let’s dive in.

 

All Positions

Stat

R-Squared with Next Season’s Rank

Entire Season’s Position Rank

0.46

Late Season’s Position Rank

0.38

 

The “Entire Season” data has more predictive power compared to “Late Season” data. In fact, the previous season’s PPR per game positional ranking explains 46% of the variation in the following season’s PPR per game positional ranking. Simply using last year’s leaderboards for this year’s projections gets us half way there.

For most players, there’s no reason to look into how well a player did at the end of the previous season, but I do think there’s some value in those stats when analyzing younger players and if a team had a change in scheme. A.J. Brown (rookie) and Tyler Higbee (Rams using more two-TE sets) are examples of players whose early/late season splits are probably valuable to look at. In general, late-season production overrated.

 

QB

Stat

R-Squared for Next Season’s Rank

Entire Season’s FPPG

0.23

Late Season’s FPPG

0.21

Entire Season’s Position Rank

0.20

Late Season’s Position Rank

0.19

 

RB

Stat

R-Squared for Next Season’s Rank

Entire Season’s FPPG

0.35

Late Season’s FPPG

0.31

Entire Season’s Position Rank

0.32

Late Season’s Position Rank

0.30

Entire Season’s ToYD Per Game

0.33

Late Season’s ToYD Per Game

0.29

 

WR

Stat

R-Squared for Next Season’s Rank

Entire Season’s FPPG

0.37

Late Season’s FPPG

0.28

Entire Season’s Position Rank

0.36

Late Season’s Position Rank

0.26

Entire Season’s ToYD Per Game

0.36

Late Season’s ToYD Per Game

0.26

 

TE

Stat

R-Squared for Next Season’s Rank

Entire Season’s FPPG

0.37

Late Season’s FPPG

0.24

Entire Season’s Position Rank

0.28

Late Season’s Position Rank

0.22

Entire Season’s ToYD Per Game

0.37

Late Season’s ToYD Per Game

0.24

 

For each position and every metric I looked at, the entire season data was more useful than the late season data. 

 

My Fantasy Football Content

1. 2019 Yards Per Carry Plus/Minus Rankings

2. 2019 Rushing Efficiency Rankings

3. 2019 Receiving Efficiency Rankings

4. 2019 Expected Passing TD Rankings

5. 2019 Expected Rushing TD Rankings

6. 2019 Expected Receiving TD Rankings

7. 2019 Big-Play Rushing Rankings

8. 2019 Expected Yards After The Catch Rankings

9. 2019 Deep Target Efficiency Rankings

10. 2019 Expected YPA Rankings

11. 2019 Expected Fantasy Points (WR)

12. 2019 Expected Fantasy Points (TE)

13. Reviewing Late-Season Production - QBs & TEs

14. Reviewing Late-Season Production - RBs & WRs

15. NFL Depth Charts - QB, RB, WR, TE

16. Free Agency Winners and Losers

17. Rookie RB Pass-Blocking Report

 

My Real Football Content

1. Stop Running It Up The Middle

2. Run From Shotgun, Not Under Center

3. Throw Beyond The First Down Marker

4. How Often Do Running Backs Block?

5. Should the NFL Run More in the End Zone?

6. How Predictive Is Late-Season Production?