Taxes.
Getting older.
Political rancor.
A Woody Allen film featuring a male protagonist who can easily, unrealistically and wish-fulfillingly woo the younger female object of his desire.
These are the things that we can count on each year.
Add to the list running back fragility. It’s natural for yesterday’s NFL conqueror to be today’s fantasy corpse - one day, Chuck Norris will try to break a brick with his hand, and that brick will break his hand, know what I’m saying? - and this progression of Freudian now-here now-gone seems most applicable to the RB position. Each season, stud RBs suffer injuries or regress, backup and out-of-nowhere RBs emerge, and the carousel of random RB productivity continues its perpetual spin.
The 2014 season was no different. Whereas only three quarterbacks and four wide receivers to enter the NFL as no better than fifth-round selections had top-24 and top-36 campaigns at their respective positions last season, 15 late-round (LR) and undrafted free agent (UDFA) RBs had top-36 positional performances in 2014, good for 41.7 percent. That number is high - but it’s not an anomaly. In any given season, you’re not likely to find the next Tom Brady and Antonio Brown, but you do have a decent chance of finding the next C.J. Anderson - especially if you have an idea in advance about what the typical stud LR & UDFA RBs look like.
Consider this article your in-depth primer on “what the typical stud LR & UDFA RBs look like.”
Rightfully, you might ask this question: “Can RBs who emerge out of nowhere actually be profiled?”
Based on the seven years’ worth of extensive data that I have collected and analyzed - and I continue to collect and analyze more data all the time - I can say this: “Yes, the cohort of productive LR & UDFA RBs can definitely be profiled.” As a group, successful LR & UDFA RBs have clear characteristics when compared to other collections of RBs.
An Analysis of All RBs Drafted, 2008-2014
But, first, before looking specifically at productive LR & UDFA RBs, let’s look at all RBs in general, to understand the lay of the land.
Regardless of the impact NFL scouts actually have on the draft process, NFL teams have been highly predictable and unimaginative in the way that they have drafted rushers since at least the fantastic RB class of 2008.
Here are 13 general trends based solely on draft position. The higher a RB is drafted...
1) The more he tends to have NFL success.
2) The younger he tends to be.
3) The heavier he tends to be.
4) The faster he tends to be.
5) The more agile he tends to be.
6) The farther and higher he tends to be able to jump.
7) The more 225-lb. bench press repetitions he tends to be able to do.
8) The more productive his most productive season in college tends to be.
9) The younger he tends to be whenever he has his most productive college season.
10) The more games he tends to play in his most productive college season.
11) The more rushes, rushing yards and rushing touchdowns he tends to accrue in his most productive college season.
12) The more receptions, receiving yards and receiving TDs he tends to accrue in his most productive college season.
13) He tends to accrue even more punt and kick return TDs in his most productive college season.
These trends could be presented in a lot of ways, but maybe the best way to show you the numbers is to highlight the difference between first- and second-round RBs and all other drafted RBs.
| 2008-2014 RBs | Top-36 % | Rookie Age | Wt | 40 Time | ShSh | 3C | VJ | BrJ | BP |
|---|---|---|---|---|---|---|---|---|---|
| Rounds 1-2 | 66.67 | 22.06 | 214.38 | 4.50 | 4.23 | 6.89 | 34.97 | 120.74 | 21.56 |
| Rounds 3-7 | 22.73 | 22.75 | 213.30 | 4.54 | 4.28 | 7.04 | 34.66 | 119.01 | 20.05 |
Clearly, the RBs selected at the top of the draft have historically been more likely to have at least one season of NFL success - and upon entering the NFL they have been younger, bigger, faster in the 40-yard dash, more agile in the short shuttle and three-cone drill, more explosive in the vertical and broad jumps and stronger at the bench press.
And here are how first- and second-round RBs and all other drafted RBs compare (on a per-game basis) when one looks at their best college seasons (again, on a per-game statistical basis):
| 2008-2014 RBs | Age | Gm | Att | RuYd | RuTD | Rec | ReYd | ReTD | RetTD |
|---|---|---|---|---|---|---|---|---|---|
| Rounds 1-2 | 20.86 | 12.67 | 21.09 | 123.48 | 1.30 | 1.97 | 19.36 | 0.14 | 0.02 |
| Rounds 3-7 | 21.11 | 12.02 | 18.09 | 100.89 | 1.11 | 1.82 | 17.99 | 0.12 | 0.01 |
The RBs selected at the top of the draft have their best collegiate seasons at a younger age, and in those seasons they play more games and on a per-game basis accumulate more touches and significantly more yards and TDs. Big surprise, I know.
Basically, if a NFL team did nothing but look at a RB prospect’s age, pre-draft measurements, college stats, and composite draft range in all the mock drafts available on the internet - if a team didn’t even look at his tape - the odds are decent that the team would draft the player with a pick comparable to the one it otherwise would’ve used to draft him anyway. I’m not saying that scouting is useless. Some guys do it really well. I’m just saying that, when it comes to the RBs drafted since 2008, it’s possible that the tape confirms the numbers just as much as the numbers confirm the tape. To me, that makes sense.
Introducing the Workhorse Metric
Readers of RotoViz, which is where I usually peddle my written thoughts, will likely know about the Workhorse Metric, which measures the extent to which any college runner in his most productive season has dominated the representative non-quarterback rushing production on his team. Basically, this market share metric for RBs (built upon the early work of Shawn Siegele and Ryan Rouillard) adjusts for the QB position, missed contests, injuries and game flow. Just like any other metric, it’s imperfect, but I’ve had success with it.
I bring it up at this point for a few reasons:
1) It gives us another perspective through which to analyze rushing production. Raw statistics are instructive and predictive, but they may not be the best representation of rushing production available to us.
2) As one might expect, the general production trends for draft position that we’ve seen above hold true for the Workhorse Metric. The higher a RB is drafted, the higher his Workhorse Score (WS) tends to be and the younger he tends to be when he has his best workhorse season.
3) Also - and this is an important point - the higher a RB is drafted, the likelier that his most productive college season by raw statistics and most productive college season by the Workhorse Metric are the same season. Or, phrased another way, the higher a RB is drafted, the more obvious it tends to be which college season is his best as a rusher.
And here is how first- and second-round RBs and all other drafted RBs compare when one looks at their best college workhorse campaigns:
| 2008-2014 RBs | Age | MS Car | MS Yd | MS TD | Tot WS | Same Season % |
|---|---|---|---|---|---|---|
| Rounds 1-2 | 20.75 | 71.71 | 73.13 | 78.76 | 75.95 | 80.56 |
| Rounds 3-7 | 20.97 | 64.67 | 66.84 | 71.03 | 68.87 | 67.27 |
The RBs selected at the top of the draft have their best collegiate workhorse seasons at a younger age, and in those seasons they have a higher percentage of their teams’ representative carries, rushing yards and rushing TDs - and about four-fifths of these RBs have their best workhorse season in their best raw statistical season, whereas that is true for only about two-thirds of all other drafted RBs.
So the general workhorse-draft position trends fit in with all the other trends we’ve seen so far - but that’s not really why I mentioned the Workhorse Metric. Instead, I mentioned it because of what it can tell us about LR & UDFA RBs who have NFL success.
As mentioned early, this last season 15 LR & UDFA RBs had top-36 positional campaigns. On average, the LR & UDFA RB to have a top-36 season in 2014 had an 80.91 WS - significantly higher than the average 75.95 WS we see in first- and second-round RBs drafted from 2008 to 2014. While most LR & UDFA RBs never have NFL success, LR & UDFA RBs with high Workhorse Scores (WS) have elevated chances of becoming fantasy contributors in comparison to their peers.
Against the backdrop of draft-related trends that we’ve established, the cohort of stud LR & UDFA RBs has some very clear and easily identifiable characteristics, and I contend that the cohort’s high WS is the most significant of these characteristics - but other important cohort characteristics exist as well.
Let’s examine them.
Successful LR & UDFA RBs to Enter the NFL Since 2008
Here are some of the clear characteristics of the LR & UDFA RBs who have entered the NFL since 2008 and had at least one top-36 positional campaign in their NFL careers:
1) Successful LR & UDFA RBs tend to be big - which I love, because in general I simply believe that bigger runners are better equipped to be workhorse RBs. I like big backs: Call it the Sir Mix-A-Lot Bias.
In fact, the successful LR & UDFA RB cohort by far has the highest average per-player weight of any subgroup. Not all LR & UDFA RBs need to look like Arian Foster, LeGarrette Blount, Joique Bell, or BenJarvus Green-Ellis to become successful NFL runners - Ahmad Bradshaw and Danny Woodhead have had solid careers despite entering the NFL weighing less than 200 lbs. - but in general these guys tend to look more like Rashad Jennings and even Mike Tolbert than Andre Ellington and Denard Robinson.
2) Successful LR & UDFA RBs tend to be comparatively unexceptional as athletes. Other than size (and decent ability at the bench press), not much distinguishes these guys in their pre-draft workouts. In fact, the successful LR & UDFA RB cohort is the slowest, least agile and least explosive of any subgroup when you look at the raw numbers. Again, think of guys like Foster, Blount, or even Alfred Morris. At best, you can say that these guys are good athletes for their size. At worst, you can try to overlook their lack of athleticism by making some explanatory excuse, such as, “When you’re big and strong enough to run over people, you don’t need to be exceptionally fast, agile, or explosive.” And for what it’s worth I kind of buy this excuse.
Also, I think it’s possible that these successful LR & UDFA runners make up for their relative lack of athleticism by being exceptional football players - they have good vision, take good angles, have a firm understanding of their offensive system, etc. Just think of Isaiah Crowell: Based on his pre-draft drills, he is an unexceptional athlete, but many of the film guys loved him entering the draft, not because they thought he was fast but because they thought he was a good player. So when we look for the next LR & UDFA RBs to break out, it’s not important that guys be very athletic. In general, they just need to be athletic enough (for their size) to be decent football players.
3) Successful LR & UDFA RBs tend to be comparatively old as rookies. In general, these are not guys who declare early for the NFL draft. In fact, the successful LR & UDFA RB cohort is almost the oldest of any subgroup. For every Crowell, who was a 21-year-old rookie, we can find a Blount, Bell and Matt Asiata, all of whom were 24-year-old rookies.
While I generally prefer younger RBs, the age of these successful LR & UDFA rushers doesn’t really bother me. For one, if they were younger, they probably wouldn’t be drafted as low and thus as easily available in fantasy leagues. Secondly, the ages of these players can easily be put in context. For instance…
4) Successful LR & UDFA RBs tend to spend some of their college careers at small school football programs. Almost all of the successful UDFA RBs and almost half of the LR RBs to enter the NFL since 2008 played at non-Power Five (P5) or non-Automatic Qualifying (AQ) schools for at least a part of their college careers. In fact, the successful LR & UDFA RB cohort has the highest percentage of players to compete at small school programs out of all the subgroups. For every Foster, who spent all of his college career competing in the Southeastern Conference, we can find a Crowell and a Chris Ivory, both of whom played at major Football Bowl Subdivision (FBS) schools before dropping to a lower level of competition; or a Blount and an Anderson, both of whom played in junior college before transferring to major FBS schools; or a Bell and a Woodhead, both of whom toiled against Division II competition for their entire careers before clawing their way into the NFL.
The reason many of these players are cursorily undervalued by the NFL in the first place - their small-school pedigree - is the same reason these stud LR & UDFA RBs are older as rookies: Either because they have transferred from FBS schools, or started at JCs before transferring to major football programs, or have spent all of their careers at non-P5 or non-AQ institutions. Almost none of these RBs are in a favorable position to declare early for the NFL draft, as they haven’t played enough games against decent competition or garnered enough attention from scouts to create the pre-draft hype that enables them to leave school earlier than they must. It’s true that most of these RBs are old rookies - and generally older rookies aren’t great players to roster - but in the case of successful LR & UDFA RBs the older age can be explained satisfactorily.
Let’s see how first- and second-round RBs compare with the successful LR & UDFA RBs on these four points:
| 2008-2014 RBs | Rookie Age | Wt | 40 Time | ShSh | 3C | VJ | BrJ | BP | Small School % |
|---|---|---|---|---|---|---|---|---|---|
| Rounds 1-2 | 22.06 | 214.38 | 4.50 | 4.23 | 6.89 | 34.97 | 120.74 | 21.56 | 13.89 |
| Top-36 LR & UDFA | 22.77 | 218.11 | 4.59 | 4.33 | 7.08 | 33.40 | 117.00 | 21.59 | 63.64 |
True to form, the stud LR & UDFA RBs are significantly older, slower, less agile and less explosive than the RBs selected at the top of the draft - but they are also significantly bigger, and whereas only about one-eighth of the first- and second-round RBs spend time at non-P5 and non-AQ schools (and only one of these RBs has spent time at JC) almost two-thirds of the LR & UDFA studs have played for small school programs. Clearly, one can see the emerging outline of an older, small school runner who isn’t an exceptional athlete but is certainly big enough to play in the NFL.
And what about the collegiate production of the LR & UDFA RBs? Get your popcorn ready - this gets exciting, especially if you have a very loose definition of the word “exciting.”
5) Successful LR & UDFA RBs tend to have production in their best statistical season that almost rivals the production of the first- and second-round RBs - and this is despite the fact that successful LR & UDFA RBs also tend to do little as receivers. In fact, the successful LR & UDFA RB cohort is second only to the first- and second-round RB cohort when one compares its production from best raw statistical season, even though the successful LR & UDFA RB cohort also has the fewest receptions and receiving yards in its best season out of any subgroup.
The perfect guy to think about in this context is Morris: A guy who does almost nothing as a receiver but who in his best raw statistical season had production that was just a notch below that of average first- and second-round RBs and certainly much better than that of average third-, fourth- and late-round RBs - and that’s especially impressive considering that his best season by raw statistics occurred in his first season as a starter, when he was a redshirt sophomore—which leads me to this…
6) Successful LR & UDFA RBs tend to have their best collegiate seasons at a younger age than other RBs. At RotoViz, Shawn Siegele has highlighted the importance of breakout age, and my RotoViz Radio cohost Jon Moore has recently talked with Josh Norris about the utility of age-adjusted metrics, and in the example of stud LR & UDFA RBs we can see once again the benefit of considering age - especially when it’s the age of a younger player with good production. To wit, the successful LR & UDFA RB cohort has its best season at the youngest age of any subgroup - and this includes best season by both raw stats and the Workhorse Metric.
That successful LR & UDFA RBs tend to do well at a young age gives added support to the notion that their advanced rookie age is less important than it otherwise might be. In general, successful LR & UDFA RBs are not guys who just happen to dominate at the end of college because they are older than their competition. Usually their story has more texture. Some of these guys are production wunderkinds of sorts (Bell and Woodhead, for instance) who continue to do well throughout college. Others are guys who do well at a young age, and then for whatever reason - injury, a change in coaching staff, a new backfield committee, etc. - they suffer a sizable productivity regression after their best season (a la Foster at Tennessee), and then, because of the NFL’s collective recency bias, they enter the league as prospects valued less than they otherwise might have been if they had peaked closer to the end of their college careers. Either way, it’s significant that these RBs are at their best at a young age.
Here’s how the successful LR & UDFA RBs compare (on a per-game basis) with first- and second-round RBs when one looks at their best college seasons (on a per-game statistical basis):
| 2008-2014 RBs | Age | Gm | Att | RuYd | RuTD | Rec | ReYd | ReTD | RetTD |
|---|---|---|---|---|---|---|---|---|---|
| Rounds 1-2 | 20.86 | 12.67 | 21.09 | 123.48 | 1.30 | 1.97 | 19.36 | 0.14 | 0.02 |
| Top-36 LR & UDFA | 20.50 | 11.05 | 19.41 | 114.77 | 1.23 | 1.75 | 16.89 | 0.13 | 0.01 |
Because the successful LR & UDFA RBs in general went to non-P5 and non-AQ schools, they didn’t have the opportunity to play as many games per season, but, otherwise, based on this table they basically look like significantly younger versions of the first- and second-round RBs who touched the ball about two fewer times per game. In general, the two cohorts are not statistically very dissimilar. The RBs drafted at the top of the draft are clearly more productive in their best statistical season, but they’re not that much better than the LR & UDFA RBs who have NFL success.
And what about production by market share? Remember - the successful LR & UDFA RB cohort has its best season at the youngest age of any subgroup - and this is true for both raw stats and the Workhorse Metric. What does the WS for this cohort look like?
7) Successful LR & UDFA RBs tend to have higher WSs than other RBs. While the LR & UDFA RBs who had top-36 campaigns in 2014 had an admittedly high WS, it’s not uncommon for such RBs to outdo first- and second-round RBs. In fact, the successful LR & UDFA RB cohort has the highest average WS of any subgroup. Although these RBs do not distinguish themselves from their first- and second-round peers on the basis of their raw stats, stud LR & UDFA RBs absolutely separate themselves from all other RBs on the basis of how they perform in the Workhorse Metric. That’s why this metric is important.
Just look at Branden Oliver. He wasn’t exceptional in 2014, but as an UDFA rookie he did lead the Chargers backfield and finish the year as the No. 28 fantasy RB. How did he do that, as a 208-lb. rusher with limited athleticism? He had a 98.01 WS. Certainly, he needed some luck even to have the opportunity to lead his team in rushing - but all RBs need a degree of luck at some point, and once Oliver had his chance he did better than almost anyone thought he would. That wasn’t luck. That’s just what usually happens when a true workhorse gets a chance, even if the chance is only temporary and the workhorse isn’t big and fast.
And as useful as the Workhorse Metric is for analyzing LR & UDFA RBs, what I find most intriguing about it is this…
8) Successful LR & UDFA RBs tend to submit their best workhorse campaigns in seasons that are not their best statistical campaigns. In other words, with the average LR & UDFA RB, the season that most people might assume is obviously his best season might not actually be his best season. In fact, the successful LR & UDFA RB cohort of all subgroups has the lowest correlation between best statistical and best workhorse seasons. While the average RB of every other subgroup is likely to have his best statistical and workhorse campaigns in the same season, the average LR & UDFA RB is likely - at least historically - not to have both campaigns in the same season. When it comes to the stud LR & UDFA RBs, it’s very possible that their best seasons are precisely not those that people can look at and say, “Yep, that’s them at their best.”
The implications of this fact seem apparent. Even with all of the other explanations for why these productive players aren’t drafted until the fifth round, if at all, this reason might be the most important: Despite the strong raw production these stud LR & UDFA RBs accumulate in their best statistical seasons, these players still possesses hidden value that NFL decision makers probably do not see during the evaluation process - probably because these players are not being evaluated on the basis of their truly best seasons. At least that’s what the Workhorse Metric suggests.
Here’s how the successful LR & UDFA RBs compare with the first- and second-round RBs when one looks at their best college workhorse campaigns:
| 2008-2014 RBs | Age | MS Car | MS Yd | MS TD | Tot WS | Same Season % |
|---|---|---|---|---|---|---|
| Rounds 1-2 | 20.75 | 71.71 | 73.13 | 78.76 | 75.95 | 80.56 |
| Top-36 LR & UDFA | 20.41 | 66.79 | 71.04 | 82.28 | 76.66 | 36.36 |
At a younger age, with a significantly lower percentage of team carries, the LR & UDFA RBs edge out the first- and second-round RBs with a higher WS - and the anonymity of these strong workhorse campaigns is largely ensured by the fact that just over one-third of these workhorse performances occurs in the best statistical seasons. When these LR & UDFA RBs enter the NFL, they do so with peak college production rivaling that of the first- and second-round RBs drafted since 2008… and nobody knows it - until now.
The Stud LR & UDFA RB Summary
I’ve thrown a lot at you about the LR & UDFA RBs who have NFL success. Allow me to summarize. Successful LR & UDFA RBs…
1) Tend to be big.
2) Need not be exceptionally athletic.
3) Often are relatively old as rookies.
4) Usually have spent time at small school football programs.
5) Are almost as statistically productive at their best in college as early-round RBs at their best.
6) Have their best college seasons at a relatively young age.
7) Are more productive at their metrical best in college than early-round RBs at their best.
8) Generally are at their statistical and metrical best in different seasons.
Remember, these are trends. It’s not necessary - it actually would be rare - for one single RB to have all eight of these characteristics. But if you’re looking at LR & UDFA RBs in the hope of finding the next rushing stud, you could probably do a lot worse than simply starting with guys who have at least a couple of these characteristics and then moving on from there.
Toward the 2015 Season
With the 2015 NFL Combine starting soon, I will periodically publish rookie RB profiles at RotoViz once we have full pre-draft workout information. The Workhorse Metric will likely form the backbone of these profiles, and each piece will include a full list of players who as prospects were comparable to the subject.
In the meantime, I’ve started looking at the 12 veteran sleeper workhorse RBs whom I believe have a chance of becoming solid fantasy contributors this upcoming season. For your convenience - and because I believe that overwriting is impossible - I’ve provided here the profile for my No. 1 veteran sleeper workhorse for 2015:
Khiry Robinson, 2013 UDFA, 88.97 WS
Given that Robinson was one of Shawn Siegele’s top-10 dynasty sleepers for 2014 - and given that Robinson flashed some ability last season before his injury and is now projected to be on a team lacking Mark Ingram and thus an established starter next season - I’m not being highly original in declaring Robinson my No. 1 workhorse sleeper for 2015, but why take a jump shot when a layup is available?
In two seasons in the NFL, Robinson has been a solid backup and committee RB, exhibiting the ability to contribute as both a runner and receiver. In other words, he could essentially replace Ingram and Pierre Thomas simultaneously. Most importantly, listed at 6’0” and 220 lbs., Robinson has the size to carry the load for New Orleans. What Ingram did in 2014 Robinson probably has the ability to do in 2015, when he’ll be 25 years old for all but the last few days of the season. Even though Robinson isn’t exceptionally athletic (he ran a 4.71 40-yard-dash at 206 lbs. at his 2013 pro day), RBs historically haven’t needed to be good athletes to succeed in Head Coach Sean Payton’s offense. Evidence Exhibits A & B: Ingram and Thomas.
And lest one think that Robinson’s success so far in the NFL is the fluky result of small sample size, peruse Robinson’s college career: As a redshirt senior at West Texas A&M in 2012, Robinson had 247 rushes for 1,621 yards and 15 touchdowns and 38 receptions for 430 yards and four more TDs, good for an 87.2 WS. And just in case you think that Robinson was a one-year wonder who benefitted from being a little older than his competition, here’s what he did in 10 games as a true freshman in 2008 at Mesabi Range Community & Technical College: 160 carries for 893 yards and 13 TDs for an 88.97 WS on top of which he added 15 receptions and a TD as a return man. And if you think that junior college stats are suspect at best and meaningless at worst, take that up with C.J. Anderson, LeGarrette Blount, Brandon Jacobs and Vick Ballard. When a stud JC RB gets to the NFL, it’s not unusual for him to have success. Robinson could be next.
Want to know who my remaining 11 sleeper workhorse RBs for 2015? Read the rest of Part 1 at RotoViz, and keep an eye out for Parts 2-4 of the series. And if you have questions about the Workhorse Metric or anything else fantasy related, feel free to hit me up on Twitter (@MattFtheOracle).
