Jump to content
Sign in to follow this  
AlNFL19

Running Back Statistical Projections

Recommended Posts

Here's another draft projection model, this time for running backs.

Basically, this model projects a player's average Approximate Value in years 3-4 based on the 2006-2015 drafts. It is correlated to NFL AV at 0.51, which is mediocre but slightly better than draft position (-0.46). It also includes a Bust Chance (<5.0 AV) and Pro Bowl Chance (10.0+). It works for running backs drafted in Rounds 1-3.

It includes:

  • Draft Position
  • Volume Statistics like Rushing Yards, Receptions
  • Efficiency Statistics like Rushing Yards Per Team Attempt and Adjusted Rushing Yards Per Attempt
  • Combine Performance in the 40-Yard Dash and Broad Jump
  • Speed Score

Here's this year's projections, which reflect the consensus that this class wasn't very strong.

1. DARRELL HENDERSON, MEMPHIS

D. Henderson, RB, Memphis
Statistic Figure
Projected AV 4.89
Bust Chance 50.95%
PB Chance 16.25%

Henderson coming in as the frontrunner with a 4.89 projection means nobody in the class even hits 5.0 on average. Henderson overcame a relatively low draft position with a group-best 1,909 rushing yards, at a clip of 8.92 per rush. His adjusted yards per attempt is knocked down a little to 8.32, however, because he was just below the average threshold for carries. After being hailed as a speedy rusher, Henderson's 4.49 40 time was a bit of a surprise, with scouts saying he could run in the 4.3's.

2. MILES SANDERS, PENN STATE

M. Sanders, RB, Penn State
Statistic Figure
Projected AV 4.74
Bust Chance 55.81%
PB Chance 13.66%

Sanders is no Saquon Barkley - Barkley's projection is the second-best since 2006 - but he had some success of his own at Penn State. Sanders worked out pretty well at the Combine, and was helped in projection by his 24 receptions and 124" broad jump. Neither sounds very impressive, but they both ranked second in the class to Boise State's Mattison (among Rounds 1-3 picks). 

3. JOSH JACOBS, ALABAMA

J. Jacobs, RB, Alabama
Statistic Figure
Projected AV 4.62
Bust Chance 59.69%
PB Chance 11.58%

My model questions whether Jacobs deserved to be a Day One pick out of Alabama. Scouts raved about him, but he didn't put up spectacular stats. His 640 yards (and a modest 4.66 adjusted YPA) were below-average, and despite being praised for his third-down abilities, Jacobs caught just 20 passes in his final season with the Crimson Tide. His 4.60 40-yard dash time was also not great, and his 9'4" broad jump was five inches worse than any other RB in the list.

4. DAVID MONTGOMERY, IOWA STATE

D. Montgomery, RB, Iowa State
Statistic Figure
Projected AV 3.89
Bust Chance 83.33%
PB Chance <1.00%

Talk about not a great RB class. From 2006 to 2015, five players had a projection below 4.0 AV (zero were hits in the NFL - though some recent players appear to be on the right track in James Conner and Kareem Hunt). This class has four by itself. According to the model, Montgomery's Pro Bowl chances are almost nill. Despite playing at a small school, Montgomery turned in a pedestrian 4.73 yards per carry, beating out only Alexander Mattison. A 4.63 40 and 121" broad jump are nothing special, which is basically what the model thinks Montgomery will be at the next level.

5. DAMIEN HARRIS, ALABAMA

D. Harris, RB, Alabama
Statistic Figure
Projected AV 3.76
Bust Chance 87.54%
PB Chance <1.00%

Harris may have shed some weight in his senior season at Alabama, but it didn't seem to help him much on the field. Harris' production dropped from 1000 yards on 135 carries to 876 on 150.  The model likes that he caught more passes than fellow Alabama product Josh Jacobs, but it doesn't like his mediocre production and low draft position. Harris wasn't really a feature player in Alabama's offense, and there's reason to believe he won't be one in New England either.

6. DEVIN SINGLETARY, FLORIDA ATLANTIC

D. Singletary, RB, FAU
Statistic Figure
Projected AV 3.57
Bust Chance 93.69%
PB Chance <1.00%

Singletary couldn't repeat his 32-touchdown 2017 campaign, and his projection suffered. Singletary's Speed Score is below 90, the threshold which nearly every recent successful back has hit. At the end of the day, a 4.66 40 isn't great, and it's even worse if you weigh just 203 pounds. Singletary also caught a laughable six passes in 2018. Six!

7. ALEXANDER MATTISON, BOISE STATE

A. Mattison, RB, Boise State
Statistic Figure
Projected AV 3.37
Bust Chance >99.90%
PB Chance <1.00%

Mattison's prospects for success are poor according to the model. From 2006 to 2015, only Matt Jones had a lower projection. However, two recent lowly-projected players, Kareem Hunt and James Conner, are providing a precedent for success with a bad projection. This comes as a bit of a surprise, since Mattison's statistics weren't awful. His 1,417 rushing yards and 27 receptions both look pretty good at first glance, although his 4.69 rushing yards per attempt does taint the merit of his rushing totals. Ultimately, two things mattered quite a bit: Mattison was drafted low, and he was slow. At 4.67 in the 40, Mattison finishes last in the draft class, and isn't at a weight that would make the time impressive for a prospect. And at pick 102, Mattison only qualified for the model because of compensatory selections.

 

Other statistical models (this marks five of them): 

 

 

 

 

 

Share this post


Link to post
Share on other sites

How much weighting in your model goes towards college production vs measurables? How do you account for a guy playing behind a stud and/or not getting as many opportunities due to scheme? Teams like the Tide tend to use a platoon method while a guy like Sanders was playing behind a future GOAT candidate in Barkley. What about quality of opposition?

Share this post


Link to post
Share on other sites
5 hours ago, Karnage84 said:

How much weighting in your model goes towards college production vs measurables?

Taking into account all the statistics used, college production in total is weighted at roughly 125% of that of measurables. 

 

How do you account for a guy playing behind a stud and/or not getting as many opportunities due to scheme? Teams like the Tide tend to use a platoon method while a guy like Sanders was playing behind a future GOAT candidate in Barkley.

A lot of the statistics are rate statistics, so for example if the team doesn't run the ball that much, they can still do well if they have a good share of total team production. But if they don't have a really high share, like if they were an Alabama running back, there's probably a good reason why. I'm sure Nick Saban would have no problems giving one guy a hell of a lot of carries if he earns them (which we've seen before - he gave Derrick Henry 400 carries despite also having Kenyan Drake on the roster). It's not perfect, but if a guy struggled to handle a lot of carries in college, there's a good chance he won't flip some magic switch in the pros, and it is a probability model after all in some regards.

 

What about quality of opposition?

Quality of opposition isn't a major included factor at this point, but draft position does to some degree filter that out (ex. guys that play in terrible conferences only get drafted highly if they were really dominant).

 

Share this post


Link to post
Share on other sites

If there’s one position where measurables won’t be as helpful out of all the skill positions it’s RB IMO.   40-yard dash times are probably nowhere as important as change of direction and suddenness (so 3-cone and shuttle may be even more helpful).   Not to mention intangibles like vision and balance.   

The variables that affect workload also really cloud the picture.  I like the concept I just see a lot more methodological flaws for this position.   

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  



×