An Objective Measure of Talent in RBs (2 Viewers)

The article with the core solution for WRs is appearing in this year's Football Prospectus. Once it's published (7/21 - buy a copy) I'll have a lot more to say about it.

This article is to present my findings on RBs. I'm going start with the method and will use the discussion and future posts to further explain each criteria and why I think it works. If I'm right here, running backs can be quantitatively evaluated almost perfectly once you've got a general idea where the scouts are projecting them. I'm going to use actual draft position in this article, but you'll see that it works well enough as long as you've got a guy slotted within about a round of where he'll actually go.

Big backs are fairly simple. The initial pool is all RBs drafted between 1998 and 2008 who weighed 220 pounds or more at the time of the combine. Each step below contains one criteria for starter quality running backs in the NFL. Only those who satisfy every criteria can be expected to be elite talents or even starters for more than a season or two.

1) Eliminate RBs who attended BCS conference schools and were drafted in the second half of the draft (roughly after the first 110 picks) as well as backs that attended smaller schools who were drafted in the 7th round:

Cedric Cobbs

Correll Buckhalter

Dwayne Wright

Edwin Watson

Ken Oxendine

Kolby Smith

LaBrandon Toefield

Luke Staley

Rock Cartwright

Ryan Torrain

Sean Bennett

2) Eliminate those over 240 pounds:

Brandon Jacobs

Curtis Enis

Eric Shelton

Greg Jones

Jermaine Fazande

Jonathan Wells

Michael Bush

Ricky Williams*

Ron Dayne

T.J. Duckett

* note that after the draft he dropped enough weight to play under 240

3) Sort the remaining pool by age - and eliminate everyone over 23.0 years old on 9/1 of their rookie season, unless they were a first round pick. These are backup/role-player quality RBs:

Anthony Thomas

Artose Pinner

Brian Leonard

Frank Moreau

J.J. Johnson

Joe Montgomery

Ladell Betts

Michael Robinson

Skip Hicks

Travis Prentice

4) Eliminate the players with known and major pre-draft character problems.

Maurice Clarett

Onterrio Smith

William Green

5) Create a power/speed metric using the following formula: (BMI * 1429)/(40yd time^4). Group all players with an index score below 100.0. I used combine measurements and 40-yard dash data as reported by www.nfldraftscout.com. Note that this metric is a variation of Bill Barnwell's index score that uses weight instead of BMI, and that it's dynamic. It should be updated as current weights are reported. Again, these are borderline starters and quality backs ups for the most part:

Cedric Benson

DeShaun Foster

Ryan Torrain

Tony Hunt

William Green

6) Take the remaining players and sort them according to draft position. The first rounders are elite and will generally be starters in the NFL until age 30+ barring injury:

Ronnie Brown

Jamal Lewis

LaDainian Tomlinson

Fred Taylor

Jonathan Stewart

Deuce McAllister

Rashard Mendenhall

Willis McGahee

Steven Jackson

Chris Perry

Larry Johnson

Kevin Jones

The 2nd+ rounders are players who have the talent to be solid starters, but are prone to failures of opportunity, less leeway following injury and poor situations:

LenDale White

LaMont Jordan

Chris Henry

Travis Henry

Musa Smith

Kevan Barlow

Chris Brown

Rudi Johnson

Sedrick Irvin

Marion Barber

Tim Hightower

Michael Turner

Jalen Parmale

The smaller backs are somewhat more complicated, and will require more explanation as we go. The initial pool of small RBs are those under 219 pounds drafted between 1998 and 2008. Again, each criteria is used to whittle away a chunk of the pool - leaving only RBs perfect on all measures behind.

1) Eliminate RBs who attended BCS conference schools and were drafted in the second half of the draft (roughly after the first 110 picks) as well as backs that attended smaller schools who were drafted in the 7th round:

Ahmad Bradshaw

Alvin Pearman

Autry Denson

Brian Allen

Ciatrick Fason

Cory Boyd

Dante Hall

Darren Sproles

De'Mond Parker

Josh Scobey

Justin Forsett

Lee Suggs

Leon Washington

Michael Wiley

Mike Hart

Olandis Gary

Patrick Pass

Quinton Ganther

Shyrone Stith

Tashard Choice

Thomas Brown

Travis Stephens

Wali Lundy

2) Eliminate all RBs under 195 pounds at the time of the combine:

Derrick Blaylock

Garrett Wolfe

John Avery

Lorenzo Booker

Trung Canidate

3) Sort the remaining pool by age - and group everyone over 23.0 years old on 9/1 of their rookie season, unless they were a first round pick. Generally speaking these are backup and role player quality RBs. If they get a starting gig they are very unlikely to hold it:

Curtis Keaton

Doug Chapman

James Jackson

Jerious Norwood

Julius Jones

Justin Fargas

Kenny Irons

Kevin Faulk

Michael Cloud

Michael Pittman

Rashaan Shehee

Tatum Bell

Tavian Banks

Vernand Morency

4) Create a power/speed metric using the following formula: (BMI * 1429)/(40yd time^4). Group any players with an index score below 100.0. These are also backups and role players. I used combine measurements and 40-yard dash data as reported by www.nfldraftscout.com. Note that this metric is a variation of Bill Barnwell's index score that uses weight instead of BMI, and that it's dynamic. It should be updated as current weights are reported. Again, these are backup and role player quality backs, at best:

Chester Taylor

Lamar Gordon

Kevin Smith

Mwelde Moore

Ruben Droughns

5) Eliminate the players with known and major pre-draft character problems.

Cecil Collins

6) Take the remaining backs and group those drafted in the 2nd+ rounds who went to BCS schools or had more than 300 carries their last two years of college. If the NFL scouts get a good look at you as a small back and don't draft you in the first round you're almost certainly not going to be better than a backup or role player. They're very accurate judges of a back's ability to make defenders miss:

Amos Zereoue

Antonio Pittman

Brian Calhoun

J.J. Arrington

J.R. Redmond

Jamaal Charles

Maurice Jones-Drew*

Maurice Morris

Quentin Griffin

Ray Rice

Steve Slaton

Travis Minor

*I believe Maurice-Jones Drew fell to the 2nd round due to his height. Similar to qualifying backs with fewer than 300 carries (see below), the scouts didn’t trust what their eyes were telling them.

7) Group any first round picks that fail only on the age metric. It's very hard to judge this group given the small size and the unusual circumstances of Robert Edwards and Caddy (career ending injuries?) and Joseph Addai (100% perfect situation), but I believe you're unlikely to get full value by using a first round pick on them. Compare them to ‘elite’ list below to see why.

Carnell Williams

DeAngelo Williams

Joseph Addai

Michael Bennett

Robert Edwards

8) Take the remaining backs and create a simple size metric using the formula Weight * BMI. Group those with an index score under ~6100. I’m not really sure what to do with these backs right now, but I believe they are high risk in terms of their ability to be a #1RB in the NFL. I've got a graph that goes with this that's pretty interesting. Note that this metric is dynamic – and should be updated using current weights as they become available:

Chris Johnson

Darren McFadden

Felix Jones??

Kevin Smith??

Reggie Bush

9) The remaining backs are elite and will generally be starters in the NFL until age 30+ barring injury:

Adrian Peterson

Ahman Green*

Brandon Jackson*

Brian Westbrook*

Clinton Portis*

Domanick Williams*

Edgerrin James

Felix Jones??*

Frank Gore*

Laurence Maroney

Marshawn Lynch

Matt Forte*

Ryan Moats*

Shaun Alexander

Thomas Jones

*attended a non-BCS school, or had a very low carry total their last two years of college. I'll talk more about this in the discussion below, but again the idea is that scouts are very accurate in judging a small back's ability to make a defender miss. If they've got a lot of film and the back went to a BCS school they're comfortable slotting them into the first round. If they see an elite talent, but the back went to a small school (Westbrook, Moats, Forte), or doesn't have a large body of collegiate work to judge (A Green, B Jackson, Portis, D Williams, F Jones, Gore) they ding them a round or two because they're risk averse. I believe this is a market inefficiency. If I forget, remind me to talk about the Packers [kudos to them] in the discussion.

very very interesting. Good work wdcrob.

:(

my head is spinning. There are so many rules and counter rules, and endpoints that are different depending on other variables, and best fitting of data. http://en.wikipedia.org/wiki/Selection_bias.

What would you say about Priest Holmes and Willie Parker? What about Terrell Davis? Warrick Dunn? Lawrence Phillips? Tim Biakabatuka? Napolean Kaufman? Rashaan Salaam? Charlie Garner? Ki-Jana Carter?

How many false positives or false negatives do we get by applying your rules to the period right before your cutoff, or to free agents who would be excluded by your BCS/weight/other formulas rules if they would have happened to have been drafted in the 6th round rather than free agents?

Cool stuff. Have been eagerly awaiting these results.

My big fear, which others (EBF?) have discussed in your previous thread is that this is a classic example of data mining. What we really need to see is if the model has any predictive power. That said, you've done great work and I look forward to testing the model prospectively.

So I just read the OP again, and while I'm still impressed with the depth of thought and amount of time spent, as I summarized results for the '08 class I became more skeptical.

Elite:

Jonathan Stewart

Rashard Mendenhall

Backup/borderline starter:

Ryan Torain

Backup/role player:

Jamaal Charles

Ray Rice

Steve Slaton

Bust:

Tashard Choice

Mike Hart

No guidance offered because wdcrob "is not sure what to do with them":

Darren McFadden

Felix Jones

Kevin Smith

Chris Johnson

Ok, so great: 4 of the top 7 in most drafts don't easily fit into the model. So how useful is this model then? And why are there arbitrary cutoffs like "RBs who weigh less than 219 lbs"? Why not choose 220 lbs? Or 216, if we're picking a random number?

Don't get me wrong, I think wdcrob has done some great stuff, and he's pushed everyone's thinking on this topic....as much as any other poster on the board. But I really question the results - even if this year's prospects are "weird", I'd still like to see a model that accounts for these oddities. Additionally, I'm concerned about the predictive power....and the common pitfalls of data mining.

Keep up the great work though wdcrob. Will definitely give food for thought on rookie draft day.

JKL said:

my head is spinning. There are so many rules and counter rules, and endpoints that are different depending on other variables, and best fitting of data. http://en.wikipedia.org/wiki/Selection_bias.

What would you say about Priest Holmes and Willie Parker? What about Terrell Davis? Warrick Dunn? Lawrence Phillips? Tim Biakabatuka? Napolean Kaufman? Rashaan Salaam? Charlie Garner? Ki-Jana Carter?

How many false positives or false negatives do we get by applying your rules to the period right before your cutoff, or to free agents who would be excluded by your BCS/weight/other formulas rules if they would have happened to have been drafted in the 6th round rather than free agents?

Click to expand...

I think wdcrob's work is very interesting, and it's outside the box, but I ultimately agree that it's more data mining than anything (not that data mining can't be impressive or noteworthy, it's just simply not a source of predictive power).I've done quite a bit of studying the NFL drafts. Just about any variation that would indicate inefficient markets (i.e., BCS school or not, height, weight, BMI) is the product of small sample size and not the ineptitude of NFL drafters.

In other words, as far as predictive powers would go, I'm very skeptical when anyone would claim that A, who fits X, Y and Z patterns is going to be a better pro than B, who was drafted earlier. I haven't ruled out the chance that there are some inefficiencies in the draft market, but I haven't seen one yet.

at the complaints about "predictive powers". He only just came up with this system of classification, and he's already applied it to the 2008 rookies, so what else is he supposed to predict here? There's insufficient data as of yet to predict even 2009, so what's the problem here?

Time will tell. Good work.

JKL said:

my head is spinning. There are so many rules and counter rules, and endpoints that are different depending on other variables, and best fitting of data. http://en.wikipedia.org/wiki/Selection_bias.

Click to expand...

Yes, it would have been preferable to form the theory using only players whose names start with A-K and then to test its accuracy using only players whose names start with L-Z (or to otherwise divide the universe into two or more parts without mixing the data used to form the idea and the data used to test it).It's still interesting, but you can't really trust data-mined info to have any predictive value. To quote myself:

It is a fundamental rule of hypothesis-testing that, whenever possible, you should not use the same data to both formulate and test your hypothesis. A short example will illustrate why this is so.

Suppose I roll a six-sided die 100 times and analyze the results. I will be able to find many patterns in the results of those 100 rolls. I may find, for example, that a three was followed by a six 40% of the time, or that a one was never followed by a six.

Would you trust any such patterns to hold true over the next hundred rolls? You shouldn't. If they do, it would just be coincidence. It is easy to find patterns by looking for them in a given set of data; but the test of whether those patterns are meaningful is whether they hold true in data that have not yet been examined.

Still, the OP has some interesting ideas in it that should serve as the basis for further investigation.

Time will tell indeed. And there's plenty of sample withheld since I've included the young backs for whom the results aren't yet known.

It's not random data mining though - I've got a pretty good theory for why each criteria is important. Which I'll try to write up when I've got time.

It's also hard to absorb it the way I've written it. So I'll come at it another way sometime soon as well.

But the gist of it is that elite backs are elite because they're qualified on EVERY criteria:

Age: players develop AFTER they get to the NFL, so drafting younger makes sense

Speed/Weight metric: a measure of how much force is generated by the back

BMI: basically a measure of how much force can be brought to bear at the point of impact

Character: the NFL is grueling - you'd better be solid mentally

Small vs Large backs: big backs can be effective through use of power and only marginal ability to make defenders miss; small backs better be damn shifty

Market Inefficiencies: there's a bias against small school backs, backs who show up on the radar late or for whom there's not much track record

As for the five backs I'm unsure about... it's really unusual that four of the five guys in this category were drafted in a single year. No idea what to make of that fact, or why the NFL switched gears so suddenly. But here's what I think...

If McFadden plays at 210 he'll struggle as a classic bell cow RB

If Felix Jones plays above 207/8 he'll be a Tiki Barber clone

Kevin Smith is marginal in terms of size, and he's not fast enough given his build. On the Lions he's almost certainly a flop unless he bulks up.

Chris Johnson will be strictly a 3rd down/COP/gimmick player (though maybe a great one) - I don't think bulking up is an even an option for him

Reggie Bush is far more likely to follow the career path of Eric Metcalf than Tiki Barber

As for the players drafted prior to 1998 - not sure what you want me to do. I don't have enough data to work with them.

220 as the cutoff for big vs small backs - it's a best guess. You could set it anywhere between 220 and 225 and I wouldn't argue. But at some point around 220 it's more than enough to hit a hole hard - you don't need to be very elusive. c.f. Marion Barber

I don't expect everyone to agree. It classifies all drafted backs and it's a hypothesis that can be tested. I guess I'll also say in my defense that it was this same combo of 'data-mining' and intuition that cracked the WRs - and there's not much doubt about that one if you read the article.

JKL said:

How many false positives or false negatives do we get by applying your rules to the period right before your cutoff, or to free agents who would be excluded by your BCS/weight/other formulas rules if they would have happened to have been drafted in the 6th round rather than free agents?

Click to expand...

Again... I don't have combine data for pre-1998, so there's not a lot I can do there.BUT... the FAs who've come out of nowhere by and large do fit the mold above with the exception that they fell in or through the draft. Usually because they didn't get playing time behind another back, or were injured. Ryan Grant, Priest Holmes, T Davis for example. Even Willie Parker fits - since my model would say he's a marginally talented guy who won't be able to hold his starting job for more than a couple years before a genuinely elite back eats his lunch.

wdcrob said:

Market Inefficiencies: there's a bias against small school backs, backs who show up on the radar late or for whom there's not much track record

Click to expand...

Once again, I really like your work, so I don't mean to be overly critical.But I'm not sure this is true, and it goes against a lot of the (unpublished) work I've done.Tomlinson and Faulk were small school backs that went in the top five. Antowain and Tony Smith were first rounders, too.

Chase Stuart said:

wdcrob said:

Market Inefficiencies: there's a bias against small school backs, backs who show up on the radar late or for whom there's not much track record

Click to expand...

Once again, I really like your work, so I don't mean to be overly critical.But I'm not sure this is true, and it goes against a lot of the (unpublished) work I've done.Tomlinson and Faulk were small school backs that went in the top five. Antowain and Tony Smith were first rounders, too.

Click to expand...

Thanks Chase. Again... I'm willing to be wrong here. And the BCS sample is so small that it's entirely possible I am.ETA... I do think it helps alot if you isolate each of the variables by looking at only the RBs who failed on only a single metric.

Again... to push back just a bit... don't you find it a little odd that the five smaller RBs who made it big after being drafted outside the first round were a Div III player, and four players with fewer than 300 carries in the previous two years? Fewer than 300 carries in two years of college for a high draft pick is RARE.

Ahman Green*

Brian Westbrook*

Clinton Portis*

Domanick Williams*

Frank Gore*

And doesn't it at least make you curious that these three highly touted guys passed all the tests except the one related to force?

Cedric Benson

DeShaun Foster

William Green

Those groupings should 'feel' similar to you (that's the best I can explain it - there's literally a recognizable feeling when I see similar data points, patterns, etc).

And where are the false negatives? Where are the drafted players this system EXCLUDES that turned out to be great? Shouldn't there be at least one?

wdcrob said:

Again... to push back just a bit... don't you find it a little odd that the five smaller RBs who made it big after being drafted outside the first round were a Div III player, and four players with fewer than 300 carries in the previous two years? Fewer than 300 carries in two years of college for a high draft pick is RARE.

Ahman Green*

Brian Westbrook*

Clinton Portis*

Domanick Williams*

Frank Gore*

Click to expand...

According to my sources (and also Wikipedia), Ahman Green had 455 carries in his final two college seasons at Nebraska.Also, just for presentational purposes, I'd change that cutoff to 350 or something. The 300 cutoff makes Portis's 297 (and Gore's 286) look a little suspicious. But, in quickly looking at the others, it appears that you don't need to cut it that fine. There is a big gap (I think) between Portis and the next highest guy.

Doug Drinen said:

Also, just for presentational purposes, I'd change that cutoff to 350 or something. The 300 cutoff makes Portis's 297 (and Gore's 286) look a little suspicious. But, in quickly looking at the others, it appears that you don't need to cut it that fine. There is a big gap (I think) between Portis and the next highest guy.

Click to expand...

Thanks Doug.Funny... I had it at 350, but decided to be more conservative. If Brian Calhoun turns out to be better than Kevin Smith I'll be sorry I did.Will look into Green. I've got him at 296 for some reason.

do players drafted by the Chargers in 2008 not count?

Marcus Thomas - 213 lbs, 5th round, UTEP, 24 -> small back group 3 (backup and role player)

power/speed metric:

(28.9 * 1429)/(4.59 ^ 4) = 41298.1 / 443.86 = 93

Jacob Hester - 226 lbs, 3rd round (pick 69), 23 on May 8 -> big back group 3 (backup/role-player)

power/speed metric:

(31.5 * 1429)/(4.61 ^ 4) = 45013.5 / 451.65 = 99.66

been waiting awhile for this stuff wdcrob, nice work

wdcrob said:

Again... to push back just a bit... don't you find it a little odd that the five smaller RBs who made it big after being drafted outside the first round were a Div III player, and four players with fewer than 300 carries in the previous two years? Fewer than 300 carries in two years of college for a high draft pick is RARE.Ahman Green*Brian Westbrook*Clinton Portis*Domanick Williams*Frank Gore*And doesn't it at least make you curious that these three highly touted guys passed all the tests except the one related to force?Cedric BensonDeShaun FosterWilliam GreenThose groupings should 'feel' similar to you (that's the best I can explain it - there's literally a recognizable feeling when I see similar data points, patterns, etc).And where are the false negatives? Where are the drafted players this system EXCLUDES that turned out to be great? Shouldn't there be at least one?

Click to expand...

Your system is designed to best fit the data from 1998-2007 only. I understand you may only have combine data to 1998. I have to go back one year to find one false negative. Warrick Dunn. And I don't need his combine data, because your rules would exclude him based on weight before you get to that point. Just in the years right before your start point, you would exclude Charlie Garner (weight-light), Terrell Davis (BCS, after pick 110), Dorsey Levens (BCS, after pick 110), Terry Allen (BCS, after pick 110), Jamal Anderson (non-BCS, 7th round) and Jerome Bettis (weight-heavy). We don't need to know the combine stats for those guys to know that. I suspect some of the other guys would be false positives who were drafted in the first round but failed during the decade of the 90's. I have no doubt that early picks outperform late picks in general, and success rates for later picks is relatively low. But your absolute rules do not fit what happened in the few years right before your chosen cutoff.

wdcrob said:

Chase Stuart said:

wdcrob said:

Market Inefficiencies: there's a bias against small school backs, backs who show up on the radar late or for whom there's not much track record

Click to expand...

Once again, I really like your work, so I don't mean to be overly critical.But I'm not sure this is true, and it goes against a lot of the (unpublished) work I've done.Tomlinson and Faulk were small school backs that went in the top five. Antowain and Tony Smith were first rounders, too.

Click to expand...

Thanks Chase. Again... I'm willing to be wrong here. And the BCS sample is so small that it's entirely possible I am.ETA... I do think it helps alot if you isolate each of the variables by looking at only the RBs who failed on only a single metric.

Click to expand...

Here would be my main two questions for you.1) What type of players are systemically under or overvalued in the draft?2) Why are they improperly rated?

I'll take some time tomorrow and try to write up a few paragraphs of text that tie this together in a simple series of statements that's a lot easier to read than what I posted today. I was hoping that seeing all the castoffs grouped together by the hurdle where they fell would help make the point that the criteria I was using to slice them were legitimate. YMMV

I'll also do a brief for-the-record writeup of rookies, and backs taken in the last few drafts who remain unknowns at this point.

speaking as someone who doesn't have the time to research all these guys, i absolutely love any sort of system that can at least get me in the right ballpark as to who to pick.

just reading through threads about these players, you will always find subjective anaylsis where each guy has his supporters and thinks "they'll be the next Tomlinson"

at least this type of analysis doesn't depend upon someones gut feelings, which you have to admit, the predictive power of which is pretty terrible.

Tatum Bell said:

at the complaints about "predictive powers". He only just came up with this system of classification, and he's already applied it to the 2008 rookies, so what else is he supposed to predict here? There's insufficient data as of yet to predict even 2009, so what's the problem here?Time will tell. Good work.

Click to expand...

You're right, let's wait until 2009 or 2010 and THEN decide whether or not the system has predictive power for 2008. That would be far more helpful.

Michael J Fox said:

Tatum Bell said:

at the complaints about "predictive powers". He only just came up with this system of classification, and he's already applied it to the 2008 rookies, so what else is he supposed to predict here? There's insufficient data as of yet to predict even 2009, so what's the problem here?

Time will tell. Good work.

Click to expand...

You're right, let's wait until 2009 or 2010 and THEN decide whether or not the system has predictive power for 2008. That would be far more helpful.

Click to expand...

Sorry if you're in a hurry, but you can't always get what you want right away. Regardless, how are you going to know before 2010 at whether the predictions for the 2008 draft class were any good? Hell, we're still debating the prospects of guys like Ronnie Brown and Laurence Maroney from three and two years ago, respectively, so we don't even have total confirmation for those draft classes. Time will tell.

amnesiac said:

speaking as someone who doesn't have the time to research all these guys, i absolutely love any sort of system that can at least get me in the right ballpark as to who to pick.just reading through threads about these players, you will always find subjective anaylsis where each guy has his supporters and thinks "they'll be the next Tomlinson"at least this type of analysis doesn't depend upon someones gut feelings, which you have to admit, the predictive power of which is pretty terrible.

Click to expand...

This is exactly why I started the WR and RB projects after reading Lewin's 2006 article on QBs. There are roughly 32,000 Michael J Foxes here who are happy to tell you what they THINK - and I wanted something that didn't rely on opinion. I wanted something 'objective.'

wdcrob said:

amnesiac said:

speaking as someone who doesn't have the time to research all these guys, i absolutely love any sort of system that can at least get me in the right ballpark as to who to pick.just reading through threads about these players, you will always find subjective anaylsis where each guy has his supporters and thinks "they'll be the next Tomlinson"at least this type of analysis doesn't depend upon someones gut feelings, which you have to admit, the predictive power of which is pretty terrible.

Click to expand...

This is exactly why I started the WR and RB projects after reading Lewin's 2006 article on QBs. There are roughly 32,000 Michael J Foxes here who are happy to tell you what they THINK - and I wanted something that didn't rely on opinion. I wanted something 'objective.'

Click to expand...

Wow, a call out, thx. For reference, there is only 1 Michael J Fox here. And despite my pushback, I love the stuff you've done. But unless it has predictive power, then it is simply subjectivity (in terms of structure of your model) masquerading as objectivity.

I haven't really thought this all the way through yet, so I may have to come back and edit. But the gist of my system in normal language goes like this...

In a perfect market all elite backs would be drafted in the first round, and all backs taken in the first round were perceived at the time of the combine and draft to be elite. That's the basic starting point for everything that follows.

What you see if you look only at 1st round draft picks is that those who are perfect on age, weight, speed and power outperform those who are not. Even using JKL's examples of Dunn and Bettis you see this. Sure they had nice careers, but compared to the other 1st round players who were perfect in every way? Average, at best.

1st rounders whose playing weights were too light or too heavy:

John Avery

Trung Candidate

T.J. Duckett

Ron Dayne

Curtis Enis

1st rounders who were too old:

Michael Bennett

Joseph Addai

DeAngelo Williams

Carnell Williams

Trung Canidate

Ronnie Brown

Larry Johnson

Robert Edwards

1st rounders who didn't generate enough force given their size:

Cedric Benson

William Green

1st rounders who's mass was distributed in such a way that they couldn't make full use of the force they generated:

John Avery

Trung Candidate

Chris Johnson

Reggie Bush

Darren McFadden?

1st rounders who were perfect on all measures:

Jamal Lewis

Edgerrin James

Steven Jackson

Rashard Mendenhall

Felix Jones?

Marshawn Lynch

Jonathan Stewart

Laurence Maroney

Willis McGahee

Kevin Jones

Thomas Jones

LaDainian Tomlinson

Ricky Williams

Adrian Peterson

Fred Taylor

Chris Perry

Deuce McAllister

Shaun Alexander

Keep in mind that all backs drafted in the 1st round were guys that the scouts saw and really liked. They could run the ball in college. So I think what's happening here is this:

Too light and too heavy is just a deal killer - those are just horrible picks. When your best-case scenario is Warrick Dunn and Jerome Bettis a 1st round pick is too much to spend. Sure Dunn and Bettis had nice careers, but they were not in the same class as the backs on the 'elite' list who we've got career totals on. Sorry, they just weren't. The worst-case back on the 'elite' list is Thomas Jones - who has had the misfortune of playing for three really ####ty running teams in his career. And his career stacks up favorably with the absolute best-case light/heavy backs - Dunn and Bettis.

Age... if players develop until they're around 23 and you're comparing the college career of guys who are 22 and 23 vs those who are 20 and 21 you have to project that the younger players haven't reached full maturity yet - and the older ones did so while still playing against inferior competition. I'm not sure why the cutoff here is so severe, but given that I used September 1st I suspect it may have something to do with the cutoff dates for school or league football or something similar. Regardless, players who put up big numbers in college as older players are overvalued. Given the differences between large and small backs (see a future post, I guess) I think age is less of a deal-killer for big backs than small.

Power... self explanatory. NFL RBs must be able to generate a minimum level of force (speed * mass) to be successful. The players who fail to do this do not succeed - regardless of where they were drafted (Benson, Green). There's probably some way to combine this figure with the one below into a single number since BMI is in both. I haven't played with that yet.

Weight/BMI... a compact frame allows a RB to maximize the force they can use at impact. Longer, leaner frames spread the effectiveness of the force across a bigger area. Denser RBs are more able to use the force they generate to break tackles and maintain their balance. This stuff is pretty simple H.S. physics - but I've forgotten all that now. You can be a medium sized back with a low BMI (Adrian Peterson) or a small back with a high BMI (Westbrook), but you can't be a small back with a low BMI and be a successful RUNNER. That's a theory at this point. This year's draft class, with four players who are low/low, should help confirm or trash the theory.

So that's the starting point. I'll try and explain why I think there's a real difference between big and small backs and also what backs who aren't drafted in the 1st round but turn out to be elite have in common in future posts.

Good stuff, wdcrob. Before I ask any more questions, how do you rate each RB's career? I see all your inputs, but I'd like to see how you measure their outputs.

I like the general idea and I think some of this stuff has merit. I've become a big believer in BMI and I don't think it's a coincidence that "surprise" stars like MJD, Barber, Gore, and Westbrook all had a BMI of 30+. I think BMI, draft position, and combine numbers can go a long way towards determining which backs have a reasonable chance of success. A lot of this stuff just comes down to common sense and can be simplified as follows:

Favor players who were early picks and who physically resemble successful NFL RB's in terms of body type and combine numbers.

That's my basic philosophy. It's easy to apply. This year Rashard Mendenhall and Jonathan Stewart are the only flawless prospects. Ray Rice and Felix Jones are borderline. Matt Forte, Kevin Smith, and Tashard Choice have fringe starter potential, but are probably going to be mediocre in the NFL. Darren McFadden, Chris Johnson, Jamaal Charles, and Steve Slaton might make some plays and contribute, but the deck is stacked against them ever becoming featured backs. Everyone else can be ignored until very late in your drafts, where you might want to take a flyer on someone like Torain, Omon, Parmele, or Hightower.

rob is right about the number of bizarre backs in this draft class. I think what happens to these guys will say a lot about the importance of weight and BMI:

Darren McFadden - adequate weight/low BMI

Chris Johnson - low weight/low BMI

Jamaal Charles - low weight/low BMI

Ray Rice - low weight/high BMI

It's rare to have two first round picks with such abnormal builds. I think both guys have a chance to contribute, but I don't think either will be a star. My hunch is that BMI trumps weight and McFadden will disappoint owners who are expecting a true franchise back. I view him more as a high-powered COP back. He may even be a complete bust. Chris Johnson has the most bust potential of this year's first round backs. If used properly, he'll provide a dynamic element as a pass catcher and part-time RB. He could be a miserable flop though. Charles is an exciting runner with an awful frame. Career backup IMO.

Ray Rice is the guy who most intrigues me here. I've always been very high on what I've seen on the football field. He LOOKS like a RB is supposed to look. Couple that with good combine numbers and an ideal BMI, and I think he has the most star potential of any RB not picked in the first round. But can a 200 pound power back succeed in the NFL? That's a huge unknown. There really hasn't been anyone quite like Rice in the past few years. MJD and Ryan Moats are probably the closest things.

I have a hunch that if anyone will emerge from the second tier and be viewed as a franchise type FF back 3 years from now, it will be Ray Rice.

wdcrob said:

2) Eliminate those over 240 pounds:

6) Take the remaining players and sort them according to draft position. The first rounders are elite and will generally be starters in the NFL until age 30+ barring injury:

Jamal Lewis

Click to expand...

Was Jamal under 240 at the time he was drafted? I know he's been as high as 250, right now he's listed at 245, but what was he when he was drafted?I like the research but it has a feel of looking at a conclusion and then going back to make data fit the conclusion. Still good to see patterns in the ethos.

Bracie Smathers said:

wdcrob said:

2) Eliminate those over 240 pounds:

6) Take the remaining players and sort them according to draft position. The first rounders are elite and will generally be starters in the NFL until age 30+ barring injury:

Jamal Lewis

Click to expand...

Was Jamal under 240 at the time he was drafted? I know he's been as high as 250, right now he's listed at 245, but what was he when he was drafted?I like the research but it has a feel of looking at a conclusion and then going back to make data fit the conclusion. Still good to see patterns in the ethos.

Click to expand...

Combine was 240.

EBF said:

Ray Rice is the guy who most intrigues me here. I've always been very high on what I've seen on the football field. He LOOKS like a RB is supposed to look. Couple that with good combine numbers and an ideal BMI, and I think he has the most star potential of any RB not picked in the first round. But can a 200 pound power back succeed in the NFL? That's a huge unknown. There really hasn't been anyone quite like Rice in the past few years. MJD and Ryan Moats are probably the closest things.

Click to expand...

Rice is interesting.My reasoning with him is that in some significant part what NFL scouts are evaluating when they look at small backs is their elusiveness. Guys the size of Rice aren't going to run over anyone, so they have to make defenders miss. Again, common sense, as you say. And if a guy with the measurables that Rice does also had an elite ability to make people miss he'd get drafted in the 1st round. He wouldn't fall. Most of the elite small backs to fall (Green is an exception, as Chase pointed out) are the ones where the scouts had a seed of doubt undermining their opinion about him as a runner. Portis, Gore, MJD, Westbrook.And with 700+ carries over his last two years at a BCS program that got a lot of attention, I think Rice is probably being evaluated correctly. They had a GREAT look at him, and there's nothing in his resume that'd make you stop and wonder about whether you had a good read if you were a scout.

wdcrob said:

EBF said:

Ray Rice is the guy who most intrigues me here. I've always been very high on what I've seen on the football field. He LOOKS like a RB is supposed to look. Couple that with good combine numbers and an ideal BMI, and I think he has the most star potential of any RB not picked in the first round. But can a 200 pound power back succeed in the NFL? That's a huge unknown. There really hasn't been anyone quite like Rice in the past few years. MJD and Ryan Moats are probably the closest things.

Click to expand...

Rice is interesting.My reasoning with him is that in some significant part what NFL scouts are evaluating when they look at small backs is their elusiveness. Guys the size of Rice aren't going to run over anyone, so they have to make defenders miss. Again, common sense, as you say.

Click to expand...

But is Rice really a "small" back?He doesn't play like one. He's short, but he isn't small. He has a high BMI and a strong base. I don't think scouts dinged him for a lack of elusiveness so much as a lack of ideal height. If he had been 5'10" 220 then he very well could've been a first roun pick. So I'm not really worried about the fact that scouts saw a lot of him and still deemed him a second round talent.The real question isn't whether has first round talent, but whether a 5'8" 205 pound power RB can start in today's NFL. I really can't answer that. I think the success of MJD and Westbrook is cause for optimism.

EBF said:

wdcrob said:

EBF said:

Ray Rice is the guy who most intrigues me here. I've always been very high on what I've seen on the football field. He LOOKS like a RB is supposed to look. Couple that with good combine numbers and an ideal BMI, and I think he has the most star potential of any RB not picked in the first round. But can a 200 pound power back succeed in the NFL? That's a huge unknown. There really hasn't been anyone quite like Rice in the past few years. MJD and Ryan Moats are probably the closest things.

Click to expand...

Rice is interesting.My reasoning with him is that in some significant part what NFL scouts are evaluating when they look at small backs is their elusiveness. Guys the size of Rice aren't going to run over anyone, so they have to make defenders miss. Again, common sense, as you say.

Click to expand...

But is Rice really a "small" back?He doesn't play like one. He's short, but he isn't small. He has a high BMI and a strong base. I don't think scouts dinged him for a lack of elusiveness or so much as a lack of ideal height. If he had been 5'10" 220 then he very well could've been a first roun pick. So I'm not really worried about the fact that scouts saw a lot of him and still deemed him a second round talent.The real question isn't whether has first round talent, but whether a 5'8" 205 pound power RB can start in today's NFL. I really can't answer that. I think the success of MJD and Westbrook is cause for optimism.

Click to expand...

It really is a fair question. And with such small sample sizes there's no way of knowing if I'm right. This is all definitely under development. But by putting a theory out there we can at least ask great questions like the one you've got and then track what happens to Rice to tweak things around the edges.Having said...Here are 'small' backs who qualified as elite except that they weren't 1st round picks and also didn't have any real questions around them in terms of their ability to be evaluated (i.e. they attended major conference schools and logged more than 300 carries their last two years):Amos ZereoueJ.J. ArringtonMaurice MorrisJ.R. RedmondThese four backs were in almost identical situations to Rice - good measurables, good age, good weight, speed and BMI. And Zereoue, Arrington and Morris all had chances to shine and came up a little short (so to speak). They weren't necessarily bad backs, they just weren't legitimate NFL #1s. I think they're the most likely parallels to Rice - and that he's roughly equivalent to them.ETA: if he was 220 he wouldn't necessarily have to be elusive. Somewhere right around there guys can be effective as plowhorses - just point them at the hole and let them bash away. But at 205 he can't run people over - he has to at least make defenders unable to get a clean shot on him.

Like stated above his BMI is built to take a pounding, concern is his weight. As far as elusive, he does not have great breakaway speed but he does have good speed.

His money will be made running inside, his peers have called him one of the quickest players they have ever seen. Also goods news in the short term so far the Ravens love his receiving out of the backfield, Cam may use him quite a bit on passing downs this year.

wdcrob said:

EBF said:

wdcrob said:

EBF said:

Ray Rice is the guy who most intrigues me here. I've always been very high on what I've seen on the football field. He LOOKS like a RB is supposed to look. Couple that with good combine numbers and an ideal BMI, and I think he has the most star potential of any RB not picked in the first round. But can a 200 pound power back succeed in the NFL? That's a huge unknown. There really hasn't been anyone quite like Rice in the past few years. MJD and Ryan Moats are probably the closest things.

Click to expand...

Rice is interesting.My reasoning with him is that in some significant part what NFL scouts are evaluating when they look at small backs is their elusiveness. Guys the size of Rice aren't going to run over anyone, so they have to make defenders miss. Again, common sense, as you say.

Click to expand...

But is Rice really a "small" back?He doesn't play like one. He's short, but he isn't small. He has a high BMI and a strong base. I don't think scouts dinged him for a lack of elusiveness or so much as a lack of ideal height. If he had been 5'10" 220 then he very well could've been a first roun pick. So I'm not really worried about the fact that scouts saw a lot of him and still deemed him a second round talent.The real question isn't whether has first round talent, but whether a 5'8" 205 pound power RB can start in today's NFL. I really can't answer that. I think the success of MJD and Westbrook is cause for optimism.

Click to expand...

It really is a fair question. And with such small sample sizes there's no way of knowing if I'm right. This is all definitely under development. But by putting a theory out there we can at least ask great questions like the one you've got and then track what happens to Rice to tweak things around the edges.Having said...Here are 'small' backs who qualified as elite except that they weren't 1st round picks and also didn't have any real questions around them in terms of their ability to be evaluated (i.e. they attended major conference schools and logged more than 300 carries their last two years):Amos ZereoueJ.J. ArringtonMaurice MorrisJ.R. RedmondThese four backs were in almost identical situations to Rice - good measurables, good age, good weight, speed and BMI. And Zereoue, Arrington and Morris all had chances to shine and came up a little short (so to speak). They weren't necessarily bad backs, they just weren't legitimate NFL #1s. I think they're the most likely parallels to Rice - and that he's roughly equivalent to them.

Click to expand...

Time might prove you right, but my subjective opinion is that he's better than those guys. I view him as closer to MJD or Westbrook, although I obviously won't guarantee anywhere near the same kind of success. I know you might respond with the argument that he doesn't have the small school or limited # of carries excuse that those guys have, but no 5'8" 200 pound RB is going to get picked in the first round unless he's a total freak like Barry Sanders. IMO there's almost nothing Rice could've done in college to go much higher than he did.

EBF said:

Time might prove you right, but my subjective opinion is that he's better than those guys. I view him as closer to MJD or Westbrook, although I obviously won't guarantee anywhere near the same kind of success. I know you might respond with the argument that he doesn't have the small school or limited # of carries excuse that those guys have, but no 5'8" 200 pound RB is going to get picked in the first round unless he's a total freak like Barry Sanders. IMO there's almost nothing Rice could've done in college to go much higher than he did.

Click to expand...

You really could be right. We're talking about five guys - so there are lots of reasons they might not have panned out that are specific to each situation - and the claims that I'm overfitting the data would be proved right.On the other hand, Felix Jones and Chris Johnson managed to get drafted in the first round in the same year Rice was drafted. btw... small back, very limited carries AND drafted in the 1st round? If Felix Jones can get/keep his weight at 208/210 he could be very good indeed.Ditto that for Chris Johnson - except that he's ALSO a small school guy. So I'd say he's insanely talented in terms of running skill. I just don't think he can get to a playing weight that'd let him be a #1 RB.Will all be very interesting to watch. Thanks for the interesting discussion!

wdcrob said:

On the other hand, Felix Jones and Chris Johnson managed to get drafted in the first round in the same year Rice was drafted.

Click to expand...

I don't like lumping all three of those guys together like they're one and the same. They're not. Ray Rice and Felix Jones/Chris Johnson are TOTALLY different animals. Almost polar opposites in terms of playing style. Just because they have a similar weight doesn't mean they should be filed into the same broad category.Ray Rice is a power runner. Built very thick. Runs extremely hard and appears to have excellent leg drive.

http://www.youtube.com/watch?v=5JWPbzMUKX8

Felix Jones is more of a finesse back. More elusive than Rice, but not as stout or sturdy.

http://www.youtube.com/watch?v=4miK7GnuSAI

Johnson is pretty much a pure speed back. Not much in the way of moves, but he can blaze in the open field.

I understand where you're coming from, but that's a dangerous line of reasoning. Just because a guy is good enough to get picked high despite having serious flaws doesn't change the fact that he has serious flaws. Cedric Benson was talented enough to picked in the top 5 despite horrendous workout numbers. Troy Williamson was talented enough to get picked in the top 10 despite having bad hands. But guess what? Benson is still a poor athlete and Williamson still has bad hands. A flaw is a flaw. That doesn't mean I hate Chris Johnson, but the guy is rail thin for a RB prospect. I don't even look at him as a RB really. To me he's a RB/WR hybrid. Interesting in PPR. Not so interesting in any other formats.

I don't mean to nitpick. I guess my real problem is with what constitutes a "small" back. I'm not really sure that Rice is a small back just like I'm not sure MJD is really a small back. I'm starting to think that BMI says a lot more about "size" than weight does, so I'm more inclined to group guys together on the basis of BMI than on the basis of weight.

When you do that Ray Rice is lumped in with Rashard Mendenhall and Tashard Choice instead of Chris Johnson and Felix Jones.

EBF said:

When you do that Ray Rice is lumped in with Rashard Mendenhall and Tashard Choice instead of Chris Johnson and Felix Jones.

Click to expand...

Agree re: flaws - that's the main point of my first post - elite backs are elite because they're flawless. So I also agree re: Johnson. Disagree re: Rice - just don't believe there are 205 pound power backs. Not just strong backs, but those who can make a living running people over instead of making them miss.Time will tell.

wdcrob said:

Disagree re: Rice - just don't believe there are 205 pound power backs. Not just strong backs, but those who can make a living running people over instead of making them miss.

Click to expand...

Rice is not solely a power back. He has 4.45 speed and is unbelievably quick. But at the same time, he is very strong and built like rock, so he is very tough to bring down. He WILL run some people over, but his quickness will also allow him to avoid them if need be. The only real knock on him has been his size. He will most likely play between 205 and 210, and is probably stronger than many of the backs that are bigger than him.Hi college career was outstanding, the college careers of most of this years backs pale in comparison. As far as his game goes, there is very little not to like. The sole reason he was not a first rounder was his height. I have no doubt he will be a solid starter at some point within the next couple years.

I think most of us would instinctively tend to value:

- backs who produced at big schools rather than small schools

- backs who were explosive rather than grinders

- backs who were not too old when they were drafted

- backs who are relatively compact and not oversized

Does this theory actually take us beyond that?

wdcrob said:

And doesn't it at least make you curious that these three highly touted guys passed all the tests except the one related to force?Cedric BensonDeShaun FosterWilliam GreenAnd where are the false negatives? Where are the drafted players this system EXCLUDES that turned out to be great? Shouldn't there be at least one?

Click to expand...

Foster was great, but had injury issues. The few times he was actually healthy, he showed tremendous ability. I'm not sure he's a good example for your system.

wdcrob said:

5) Create a power/speed metric using the following formula: (BMI * 1429)/(40yd time^4). Group all players with an index score below 100.0. I used combine measurements and 40-yard dash data as reported by www.nfldraftscout.com. Note that this metric is a variation of Bill Barnwell's index score that uses weight instead of BMI, and that it's dynamic. It should be updated as current weights are reported. Again, these are borderline starters and quality backs ups for the most part:Cedric Benson

Click to expand...

Since Benson didn't run the 40 at the combine, what did you use for him? NFLDraftScout lists him at 5'10 1/2'', 222 and lists 3 different 40 times - 4.48. 4.55, and 4.62. Benson passes the test using the 1st two times but falls just short using the slowest. Why would you choose the slowest?

BuckeyeArt said:

wdcrob said:

5) Create a power/speed metric using the following formula: (BMI * 1429)/(40yd time^4). Group all players with an index score below 100.0. I used combine measurements and 40-yard dash data as reported by www.nfldraftscout.com. Note that this metric is a variation of Bill Barnwell's index score that uses weight instead of BMI, and that it's dynamic. It should be updated as current weights are reported. Again, these are borderline starters and quality backs ups for the most part:Cedric Benson

Click to expand...

Since Benson didn't run the 40 at the combine, what did you use for him? NFLDraftScout lists him at 5'10 1/2'', 222 and lists 3 different 40 times - 4.48. 4.55, and 4.62. Benson passes the test using the 1st two times but falls just short using the slowest. Why would you choose the slowest?

Click to expand...

I tried to use only combine times. Since they didn't list any, I took the one that seemed most official - the time listed as his pro-day time.

:whitecorner: and some of the best board discussion ever.

nice work wcrob and other contributers.

Portis 26 said:

I think most of us would instinctively tend to value:- backs who produced at big schools rather than small schools- backs who were explosive rather than grinders- backs who were not too old when they were drafted- backs who are relatively compact and not oversizedDoes this theory actually take us beyond that?

Click to expand...

The best benefitof BMI language for me has been away of expressing the difference that a short player i.e. MJD is not always a small player versus a tall player McFadden being a big player. As a minimal numbers type of guy, trying to get a more specific metric which changes those instincive ideas into a numeric value does not help me as much as it could others. Also, considering the NFL generally values similiar things, looking at draft position, I think one can largely (not perfectly however) incorporate these notions into an evaluation model.

wdcrob said:

Sure Dunn and Bettis had nice careers, but they were not in the same class as the backs on the 'elite' list who we've got career totals on. Sorry, they just weren't.

Click to expand...

I think you are underestimating Jerome Bettis. Last I checked he was a first ballot Hall of Famer. Granted he can be viewed as a "compiler" to som degree, but isnt' there something to be said for an NFL back who was able to last long enough to "compile" those stats.

EBF said:

wdcrob said:

EBF said:

Ray Rice is the guy who most intrigues me here. I've always been very high on what I've seen on the football field. He LOOKS like a RB is supposed to look. Couple that with good combine numbers and an ideal BMI, and I think he has the most star potential of any RB not picked in the first round. But can a 200 pound power back succeed in the NFL? That's a huge unknown. There really hasn't been anyone quite like Rice in the past few years. MJD and Ryan Moats are probably the closest things.

Click to expand...

Rice is interesting.My reasoning with him is that in some significant part what NFL scouts are evaluating when they look at small backs is their elusiveness. Guys the size of Rice aren't going to run over anyone, so they have to make defenders miss. Again, common sense, as you say.

Click to expand...

But is Rice really a "small" back?He doesn't play like one. He's short, but he isn't small. He has a high BMI and a strong base. I don't think scouts dinged him for a lack of elusiveness so much as a lack of ideal height. If he had been 5'10" 220 then he very well could've been a first roun pick. So I'm not really worried about the fact that scouts saw a lot of him and still deemed him a second round talent.The real question isn't whether has first round talent, but whether a 5'8" 205 pound power RB can start in today's NFL. I really can't answer that. I think the success of MJD and Westbrook is cause for optimism.

Click to expand...

Also how many RBs are going to be drafted in round one of one draft class? Lets not forget that their are other positions in a draft class that garner attention - its not all about fantasy footbal in the NFL draft. In a year where there were no McFaddens, Stewarts, Mendenhalls, Jones, or Johnson it is possible a team would have taken Rice in round one. Its similar to a fantasy draft where some teams look for value and say "why take a RB in round 1, when I know I can get a similiar talent in round 2?"

EBF said:

wdcrob said:

On the other hand, Felix Jones and Chris Johnson managed to get drafted in the first round in the same year Rice was drafted.

Click to expand...

I don't like lumping all three of those guys together like they're one and the same. They're not. Ray Rice and Felix Jones/Chris Johnson are TOTALLY different animals. Almost polar opposites in terms of playing style. Just because they have a similar weight doesn't mean they should be filed into the same broad category.Ray Rice is a power runner. Built very thick. Runs extremely hard and appears to have excellent leg drive.

http://www.youtube.com/watch?v=5JWPbzMUKX8

Felix Jones is more of a finesse back. More elusive than Rice, but not as stout or sturdy.

http://www.youtube.com/watch?v=4miK7GnuSAI

Johnson is pretty much a pure speed back. Not much in the way of moves, but he can blaze in the open field.

Click to expand...

As far as Chris Johnson goes with the way Devin Hetser, Joe Cribbs and Leon Washington have changed games with their kick return abilities, isn't it possible that Johnson was drafted so early just as much for the possibilities he brings to the kicking game as his ability as a runner?