Share this article:Tweet about this on TwitterShare on FacebookShare on Google+Email this to someone

ahmadbradshaw

In my last article, I looked at mid and late round small RBs to enter the NFL in the last 20 years and produce at least one top-30 season. These guys on average weigh about 200 lbs. and were acquired with fourth-round picks, and in general they look very similar to the first-round small RB cohort of the same time frame. They’re not quite as athletically gifted, but they’re still good athletes with superior college production, particularly outperforming the first-round cohort as undergrad rushers. Given that they compare to first-round players but are available late in fantasy drafts, these guys have immense value.

For most of these RBs, a component of this rushing excellence is the capture of a large portion of their respective teams’ total rushing production. In other words, these RBs have a high market share of their teams’ rushing yards and TDs. While the RotoViz guys have been applying the concept of market share to WRs for a while, Shawn Siegele was the first (whom I know of) to apply his concept of Dominator Rating to RBs. Ryan Rouillard quickly followed suit with his running back model, in which he uses a combination of draft position, speed score, and dominator rating to project NFL productivity. He has a whole series of articles in which he examines the Giants backfield, the Packers backfield, and the Rams backfield in the context of his RB model. Check out his latest article on the Dolphins backfield, particularly Lamar Miller.

To the work started by these guys, I wish to add my own contribution: the non-QB Dominator Rating (nQBDR). As you might sense, nQBDR is the rushing DR for all players who aren’t QBs (or punters and kickers). Basically, nQBDR seeks to answer this question: When the QB doesn’t run the ball, when a team decides to have the QB hand the ball off to somebody, who is he handing it off to and what exactly is that production? In short, nQBDR factors out the QB position as much as possible and seeks to know only how one RB did in comparison to his teammates who also took handoffs.

The reasons for eliminating the QB from this metric are several:

  1. The rushing production of QBs differs in nature from that of RBs, and so when one evaluates the latter any reference to the former results in an inexact comparison.
  2. In college, sacks count as rushes, and that fact really muddies the waters of analysis.
  3. Some RB shouldn’t be punished just because a mobile QB like Cam Newton can turn into Superman on any broken pass play and scramble for a long TD. Essentially, nQBDR evaluates comparative production on designed running plays in which the QB delivers the ball to a runner.

I don’t have a full database of nQBDRs yet, but from the substantial work I’ve done so far I’ve noticed these trends:

  1. The implications of a weak nQBDR correlate inversely with draft position. If a first-round player has a low nQBDR, that doesn’t necessarily mean anything. If a late-round RB has a low number, he seems more unlikely to have success than if his nQBDR were high.
  2. The implications of a weak nQBDR correlate negatively with size. If a big runner has a low nQBDR, he could still have NFL success, although he might be less likely to do so than a similarly sized runner with a higher nQBDR. If a small runner has a low number, he seems more unlikely to have success than if his nQBDR were high.
  3. In players who are small and/or not highly drafted, I have found a high nQBDR to be highly predictive of future top-30 positional success. Of all these types of players for whom I have done calculations, over 50% of those with high nQBDRs had top-30 seasons at some point. For perspective, if you look at this list of second-round RBs to enter the NFL since 2000, you’ll see that only 16 of the 31 runners have had a top-30 season (51.6%). What I’m saying is that nQBDR alone returns a group of disregarded prospects who beat that number.

What is a high nQBDR? I prefer a luda-crazy high of 90%, but 80% is certainly still high enough to prove predictive in many cases. And one more thing. When I’m looking at these low-risk cost-nothing prospects, I want to gauge upside, and so I use the DR from their best seasons, not necessarily their last. And I don’t count any missed games against them. I want to know what they do when they play, not the production they miss when they don’t.

I could go on with more details about the calculations, but let’s just get to the data.

Who are some of these disregarded prospects to emerge as nQBDR studs? According to PFR, last year 9 of the top-30 RBs (30%) were guys who weren’t drafted in the first three rounds—guys no NFL teams were in a hurry to acquire. Here they all are (the years provided are the seasons of their highest nQBDRs).

Name Ht Wt Year School Season Games Played nQBDR 2012 PR
Arian Foster

73

226

2007

Tennessee JR

14

61.27

3

Alfred Morris

70

219

2011

Florida Atlantic SR

12

90.76

5

Michael Turner

71

237

2003

Northern Illinois SR

12

84.38

17

Ahmad Bradshaw

70

198

2005

Marshall SO

11

91.96

18

BenJarvus Green-Ellis

71

219

2007

Mississippi SR

12

89.15

19

Darren Sproles

66

187

2004

Kansas State SR

11

96.64

22

Danny Woodhead

68

197

2006

Chadron State JR

13

91.4

24

Vick Ballard

70

219

2010

Mississippi State JR

12

71.86

25

Joique Bell

71

220

2006

Wayne State FR

11

92.72

29

Avg

70

213.5556

NA NA NA

12

85.57111

18

Median

70

219

NA NA NA

12

90.76

19

Note that the two RBs with the two worst nQBDRs are big guys. As I said earlier, for large RBs a less-than-great nQBDR isn’t necessarily damning. Also, I find this intriguing: My previously mentioned system of finding big RBs with lots of collegiate production identifies all of the FBS RBs on this list over 200 lbs. as players to target coming out of college. In other words—two stones, five birds, one throw.

The average of 85.6% is insanely high. The median of 90.8% is unbelievable. And, yes, this average nQBDR is higher than that of the other 21 remaining top-30 runners—the guys drafted in the first three rounds. For low-round and undrafted players, a high nQBDR is especially significant. It signifies that the players under consideration could very well outperform their draft positions.

And a high nQBDR also means a lot for small guys. Woodhead, Bradshaw, and Sproles? All of them have nQBDRs above 90%. Let’s return to the cohort of mid- and late-round small RBs from the last twenty years.

 

Name Ht Wt Year School Season Games Played nQBDR Highest PR
Brian Westbrook

68

200

2001

Villanova SR

11

89.35

2

Jamaal Charles

71

200

2007

Texas JR

13

71.67

4

Steve Slaton

69

197

2005

West Virginia FR

9

77.94

6

Karim Abdul-Jabbar

70

194

1995

UCLA SR

11

63.08

6

Darren Sproles

66

187

2004

Kansas State SR

11

96.64

10

Ahmad Bradshaw

70

198

2005

Marshall SO

11

91.96

13

Jerome Harrison

69

201

2005

Washington State SR

11

86.74

24

Derek D. Brown

69

201

1991

Nebraska JR

11

41.62

28

Avg

69

197.25

NA NA NA

11

77.375

11.625

Median

69

199

NA NA NA

11

82.34

8

(Abdul-Jabar’s actual nQBDR is higher than 63.08, because he missed one game that I am unable to factor out of his score. It potentially could be as high as 70.)

Note that 1) the four guys with the best positional rankings are the four third-round draft picks and that 2) their average nQBDR (75.51) is worse than the others’ (79.24). In general, a high nQBDR is more important for later-round players. Just look at Brown—he has the worst nQBDR, and his best positional finish is also the worst.

Also notice that the guys with the two lowest nQBDRs played college football almost twenty years ago. In context, their low numbers make sense. In the 90s college football was much more run-centric than it is today. Back then it would’ve been almost impossible for a guy to achieve a really high DR. Now, it’s possible—if he’s a stud. For the six post-2000 RBs, the average nQBDR is 85.72—almost identical to the average for the nine 2012 top-30 afterthoughts. For small RBs (especially those not drafted in the first three rounds), a high nQBDR is especially significant.

With this information, I have some particular recommendations regarding 2013 rookie RBs: players to avoid and players to target. Who are they? Find out in the next article.

Subscribe for a constant stream of league-beating articles available only with a Premium Pass.

Share this article:Tweet about this on TwitterShare on FacebookShare on Google+Email this to someone
0 comments