Eckert: RB Stats Study 1.0 Using Advanced EPA, Success Metrics

James Conner

Today I am starting my “Running Back Study Series”, similar to the “Quarterback Study Series” I just completed.  With more time this offseason, I wanted to dive deeper into some questions I usually don’t have time for. One of these questions revolved around rushing success.  Two data points came to mind that I wanted to learn from in this study:

  • Expected Points Added (EPA)/Rush= expected points after rush play-expected points before
  • Success/Rush= binary indicator (1=success, 0=unsuccessful) whether the rush was successful

Before the graph comes up, this is important!  The chart below is viewing rushing only, so this is a simple starting view.  Another goal of this study was to start with the primary skills every running back requires, adding the most value to team in the running game (example: not every running back is a receiving threat, but which running backs succeeded and added points in the run game).  With that in mind, here’s what 2020 looked like:

I wanted to take the top 35 running backs by rushing attempts to get as much info as possible from the backs that had the most opportunities.  Now, we have all heard the common narrative this offseason, the Steelers had the worst rank rushing attack in 2020.  When most people hear that, it’s generalizes all of the running backs for the team together.  This study focused on which individual players had the most usage/success league wide, to see if we can give some credit or criticism compared to previous beliefs. Looking at James Conner from last year, he played in 13 games and when he was playing, was slightly above average compared to the higher usage backs from 2020.  This makes me think other than the games he missed, were Steelers fans a bit hard on him (lumping him into the worst rushing attack narrative?).  I believe so, but most people would agree that since the Steelers made the investment and drafted Najee Harris in the first round this year, we have optimism for 2021 regardless.

Also, I was curious to see which teams had the most usage/success with “rushing by committee” (two or more rushers used regularly) or “bell cow” (one running back used primarily).  I know there will be some names that we don’t see here, but it’s important to remember that these are the top 35 running backs in rushing attempts.  The table below is an illustration of that.  (quarterbacks that matched rushing attempts included in parenthesis):

Team Rushers Success/Rush Rank (by players)
BAL Gus Edwards, J.K. Dobbins, (Lamar Jackson) 1, 18
TEN Derrick Henry 2
SEA Chris Carson 3
MIN Dalvin Cook 4
PHI Miles Sanders 5
SF Jeff Wilson 6
NE Damien Harris, (Cam Newton) 7
NO Alvin Kamara, Latavius Murray 8, 11
GB Aaron Jones 9
IND Jonathan Taylor 10
TB Ronald Jones 12
PIT James Conner 13
LAR Darrell Henderson, Cam Akers 14, 35
NYG Wayne Gallman 15
WAS Antonio Gibson 16
MIA Myles Gaskin 17
DAL Ezekiel Elliott 19
DEN Melvin Gordon 20
KC Clyde Edwards-Helaire 21
CLE Kareem Hunt, Nick Chubb 22, 24
LV Josh Jacobs 23
CAR Mike Davis 25
CHI David Montgomery 26
ARI Kenyan Drake, (Kyler Murray) 27
HOU David Johnson 28
JAX James Robinson 29
ATL Todd Gurley 30
NYJ Frank Gore 31
CIN Giovanni Bernard 32
BUF Devin Singletary 33
DET Adrian Peterson 34

I really enjoyed the outcome of this table, especially the context of team approach to the running game, then the quality of play that the player gave with the most opportunity in the run game.  So, the teams that got the best value from multiple high attempt rushers in a “committee approach” were the Ravens, Patriots, Saints, Rams, Browns, and Cardinals (James Conner’s new home, and in my opinion a better role for success and perhaps less injuries without a “bellcow” role).

Let’s take a look from this angle for a moment, who did the best with the most attempts (more for “bellcow” value):

Rushing Attempts Rank Rushing Yards Rank Success/Rush Rank
1. Derrick Henry 1 2
2. Dalvin Cook 2 4
3. Josh Jacobs 8 23
4. David Montgomery 6 26
5. Ezekiel Elliott 10 19
6. James Robinson 5 29
7. Kenyan Drake 12 27
8. Jonathan Taylor 3 10
9. Melvin Gordon 9 20
10. Aaron Jones 4 9
11. Kareem Hunt 15 22
12. Todd Gurley 26 30
13. Ronald Jones 11 12
14. Nick Chubb 7 24
15. Alvin Kamara 13 8
16. Frank Gore 28 31
17. Clyde Edwards-Helaire 17 21
18. Antonio Gibson 18 16
19. James Conner 20 13
20. Mike Davis 29 25
21. Miles Sanders 14 5
22. Devin Singletary 23 33
23. Adrian Peterson 32 34
24. David Johnson 21 28
25. Wayne Gallman 24 15
26. Latavius Murray 27 11
27. Cam Akers 30 35
28. Gus Edwards 19 1
29. Myles Gaskin 34 17
30. Chris Carson 25 3
31. Darrell Henderson 31 14
32. Damien Harris 22 7
33. J.K. Dobbins 16 18
34. Jeff Wilson 33 6
35. Giovanni Bernard 35 32

Taking the above information, I think this is a fun way to look at, running back rushing rankings by 2020 rush yards for totality, success/rush for quality, and rush attempts(desc) for usage:

Top 5 in all Rushing Ranks

  1. Derrick Henry
  2. Dalvin Cook

Top 10 in all Rushing Ranks

  1. Jonathan Taylor
  2. Aaron Jones

Top 15 in all Rushing Ranks

  1. Ronald Jones
  2. Alvin Kamara

Top 20 in all Rushing Ranks

  1. Ezekiel Elliott
  2. Melvin Gordon
  3. Antonio Gibson
  4. James Conner

Top 25 in all Rushing Ranks

  1. Josh Jacobs
  2. Kareem Hunt
  3. Nick Chubb
  4. Clyde Edwards-Helaire
  5. Miles Sanders
  6. Wayne Gallman

Top 30 in all Rushing Ranks

  1. David Montgomery
  2. James Robinson
  3. Kenyan Drake
  4. Todd Gurley
  5. Mike Davis
  6. David Johnson
  7. Latavius Murray
  8. Gus Edwards
  9. Chris Carson

Now for a fun ending question, how could Najee Harris stack up in 2021?  The one possible low-level concern I have as far as team value is the long rush attempts (20+ yard runs), but I believe he will make up for it in the passing game and successful runs.  Without any other glaring holes in his game, this will be so fun to watch the new look 2021 Steelers offense and see if Harris is the “bellcow” runner most of us think he can be (barring injury).

Thanks for reading and let me know your thoughts in the comments!

To Top
error: Alert: Content is protected !!