Didn't Tomlinson have 370 carries in 2002? Why isn't he mentioned?

Previous post--meant to say "mentioned with Eric Dickerson"

Do you guys have any theories why adding receptions doesn't fit the model? You have to admit that it seems facially correct that a running back's total workload can be found with carries and receptions. I understand the research (I think), but can't see why it works out that way.

I'm thinking EA Sports might as well just make LJ the coverboy for the '08 edition so as not to cause undue harm to another player.

4: My guess is that, on receptions, the running back isn't usually running head-on into a big defensive tackle. It's more likely he'd be tackled by a defensive back or a smaller linebacker.

6: They might also duck out of bounds w/o any contact more frequently.

#4: I would guess it has something to do with the types of RB being measured. There seem to be a lot of young backs who are heavily involved in the passing game and, having been proven effective, later get a shot at being the feature back as far as carries go. Tiki Barber and Brian Westbrook come to mind; I'm sure there are others.

Re #1:

Check out his 2004 season. Remember that it's not always the following season but could be the season after that. LT wasn't exempt, though he did stave it off for a year.

As for why the delay? My guess is that LT was young enough during his 2002 season (only 2nd year in league) that he was able to recover better than older players. ED's monstrous seasons were also earlier in his career.

#1

The Chargers didn't make the playoffs that year - the list looked at players who crossed the 370-carry mark in the postseason.

I don't know if there's data around for this kind of thing, but wouldn't RBs who duck out of bounds more often (like LT2, Tiki, Westbrook) be less likely to collapse? Guys on that list like Natrone, Eddie George, and Jamal Lewis pretty much made their living running directly into and over anyone in their way. Fun to watch, but seems like it'd ruin your career pretty quickly.

Clearly there is some correlation, but is there causation ? Does the heavy workload cause poor performance the next year, or is the heavy workload the result of superior performance(a career year) that can't possibly be replicated, and the decline the following year is just a regression to the mean.

It seems that early in your career there is less correlation. I'd imagine that bodes well for Larry Johnson, until you figure in 416 + another 30+ against the Colts next week...

I think the Chiefs have a decent chance too, so let's consider this. They probably won't make it past the second round, but assuming they do and LJ continues his current pace he'd probably finish with somewhere around 475-480 carries.

Yikes!

Wonder if it came up less with 14 game seasons, or if its # of carries per game or total per season which determines...

370 carries is just over 23 carries per game (23.125). If it were carries per game, that works out to 323 carries per season.

In looking at leader for carries in the 14 game seasons of the 70s...

in 1970, Ron Johnson of the Giants lead with 263 carris, 3.9 ypc. He got hurt the next season, but rebounded to lead the league in carries in '72.

1971 Floyd Little lead the NFL with 284 carries, 4.0 ypc. He continued strong the next two seasons, 4.0 and 3.8 ypc and didnt' drop off until 1974.

1972 the leading carrier was Ron Johnson of the Giants, 298 carries, avering 4 ypc. he dropped to 3.5 next year, and 2.2 in limited duty the following year.

1973, 74, 75 were all lead by OJ Simpson with 332, 270 and 329 carries each year, 6 ypc, 4.2 ypc, 5.5 ypc, in 77 he lead AFC (but not NFL) with 290 and 5.2 ypc, knee injury in 77 and was never same again).

Walter Payton lead in 1976 with 311 carries, 1390 yards, 4.5 ypc. He lead again in 1977 with 339 carries and 1852 ypc, 5.5 ypc. he dropped to only 4.1 ypc next year, and while he never got to 1800 yards again, certainly had a good number of years left.

Based on this, it appears that people didn't run the same back as much per game in the 14 game seasons... but even when they did, it had less effect.

Thus, the cumulative total per season seems to govern.

Of those teams only two made the playoffs, the Bills in 74 (15 carries for Simpson) and the Bears in 77 (19 carries for Payton).

Could it be that coaching philosphy has stemmed from the days of shorter seasons (or college with shorter still), and has yet to adapt to the 16 game season?

Aren't these the same people who said Tiki Barber HAD TO/MUST/WAS GOING TO drop off in 2006 because all RBs his age drop off production wise after 400+ plus touches?

Don't add me as a believer in this theory.

11: I'm pretty sure the answer to your question is "yes," but since you can't really factor regression to the mean out of anything, I think causation vs correlation for number of carries is irrelevant here.

What I want to know is, when does the decline start, i.e. does efficacy decrease as carries increase in a given season? I'd like to figure out how much I want to pay for LJ in playoffs auction draft, for one thing.

Minor quibbles about thresholds and correlation coefficients aside, this is very powerful information (and it has been since the initial article). My question is: how long will it take for this information to reach and influence NFL decision-makers?

While there's still a lot of debate among major league baseball teams regarding pitch counts, the change in usage by all major league baseball teams since the pitcher abuse articles were published by Baseball Prospectus (and Rob Neyer at ESPN.com) has been drastic - the reduction in Category IV and V starts per Baseball Prospectus' metrics (see link) is stunning. On first glance, it appears that there were only three Category V starts during all of 2006.

I believe that Aaron has mentioned in the past that, from his experience, NFL personnel have generally been receptive to this type of information. If NFL personnel are even more willing than their baseball counterparts to rethink their ways, it's possible that we've seen the last 400-carry season.

Some questions I had on another thread on this subject - just ideas for future research:

- Whether (similar to pitches for starting pitchers) it’s not the overall carry count that dooms an RB as much as games with a ton of carries. For example, maybe every carry over 20 in a game is more significant.

- Whether you can improve the predictive power by looking at a back’s height, weight, age, or career carry total.

- Whether you can improve the predictive power by looking at the distribution of runs by the RB. For example, maybe carries up the middle are more correlated with breaking down than carries that end out of bounds. Or perhaps carries for no gain are worse than carries for 7 yards.

Quick follow-up point to #16. See the link for the Pitcher Abuse Points data for 1996. Just glancing through, I saw over 50 Category V starts just ten years ago, versus just 3 such starts in 2006.

16

It might be working. This year, only one RB went over the 370 mark - LJ, who shattered it.

The next closest is Tomlinson, who 348. Marty also doesn't give Tomlinson pre-season carries.

Maybe instead of it being a new idea, or teams aren't aware of it, that Herm Edwards is just an idiot?

Re: #16--This is probably the reason that 2 RB schemes have become the norm in the NFL recently.

I have a feeling that the most effective to reach idiot coaches such as Edwards is to send a copy of this article to every sports agent in the country. If coaches are shortening the careers of the agents' players, that means less money for the agents. It would appear to be in the interest of agents to negotiate somewhat prohibitive performance bonuses for carries over 350--bonuses that would accurately compensate players for having their careers shortened by overwork. At least then Herm Edwards would understand that overworking a back like LJ has negative consequences for his team (i.e.: it would impact the team's sallary cap numbers).

The thing is, maybe in some circumstances it is worth it for a coach to pound the life out of a player in order to make a superbowl run. But since the player has no control over how many times he is asked to run in a season, he should at least have a mechanism for being compensated when the coach works him beyond what would be normally expected.

I remember Davis(DEN) blew out his knee trying to tackle the ball carrier during a turnover. That maybe a outlier. Do injuries occur when the RB is carrying, receiving the ball, running a route, or when a turnover occurs?

21: I don't think the type of injury is relevant. While I think most of this effect can be explained by regression to the mean (#16), and obviously the more carries someone gets, the more chances they have to get hurt, most injuries in football happen when a player is fatigued and not moving at full speed/ability. A player with a lot of carries will be more tired and have more accumulated stress on their body, so even when not carrying the ball, they are more apt to get injured in other situations.

I agree there is a negative correlation between workload and future performance (it just seems logical), but it would be great to see another impact variable factored in:

AGE - I assume there is a positive correlation between age and the ability to recover from a 350+ carry season, this should be accounted for.

Also lifetime workload might be an interesting point to look at in the future.

Is there any way you guys could sort the game logs by how often players are actually tackled (as opposed to going out of bounds, scoring a TD, etc.)? That might have a better correlation, as getting tackled logically seems to be the biggest contributor to this sort of "wear-and-tear" factor.

This is just a thought, and might be hopelessly misguided, but does the playing surface have an impact? Intuitively, running on old-style Astro-Turf would seem much 'harder' on the body than grass, or modern Field-Turf.

All the backs listed played in the last 25 years or so, when Astro-Turf was relatively common. It's possible that the replacement of Astro-Turf fields by Field-Turf will raise the '370 threshold' because it's a more friendly surface for players.

It's also entirely possible I'm talking through my hat.

#23: I wrote a short study in PFP 2K6 on the correlation between lifetime workload and production. Short answer: yes, there is a negative correlation between a high number of career carries entering a season and yards per carry in that season. See, e.g., Edgerrin James this season.

It seems to me like theres no way there would be a correlation between touches and rushing yards next year. I think you guys should check the corrolelation between touches and yards (recieving + rushing) the next year.

LJ should renegotiate his contract. Immediately.

I have to say, if LJ does get gobbled up by his carries, then it HAS to be Herm Edwards fault. The official NFL depth chart list 4 rbs on the chiefs roster. Sure, the oft injured Micheal Bennet is one of them, but he only had 36 carries all year! LJ gets that in one game! Heck, Dee Brown is the third string, what wrong with giving hime a series or two at some point each game?
What's Derrick Ross even doing on the roster if he's only gonna get 7 carries all year?
I agree that this is something GM's are gonna hafta get involved in. It's important to protect your investment in players, that means getting backup RB's and making the coach use them.

This might sound silly but why not count pre-season carries as well? They play tackle in August, right?

I agree with most of the posts. Age is significant here. Curtis Martin, for example, should have stopped being an effective back due to age anyway. Plus, at first sight, bigger power backs seem to be more affected by tons of carries than more elusive backs. Emmit Smith, for one, was famous for never being a clean target for tacklers. What does it matter if you're carrying the ball if you're not getting hit too hard? It's a similar thing to receptions in that respect.

Thanks for this. As a relatively new reader, I appreciate being pointed in the direction of some of the basic and necessary articles around the site.

I think Emmit Smiths longevity can be attributed in part to the fact that all he really had to do was run straight ahead through gaping holes most of the time. Hard to hurt a knee or an ankle when you don't have to turn.

(The preceeding was meant mostly in jest and is part of a long-standing anti Emmit Smith agenda I have that is fueled primarily by envy and jealousy. It isn't intended to be taken seriously.)

It's an interesting article. I'd like to know what's the correlation coefficient between age and production?

I think you'd have to start at age 25 or so though, to account for the fact the many backs go from back-ups to starters in their first couple years in the league.

It seems like there might be room here for some sort of multiple regression model to try and tease out the impact of several factors at once.

Sterling Sharpe makes a similar here, only he set his mark at 390 carries. But basically came to the same conclusions.

#34: That's also in my article from PFP 2K6. Don't have it handy to give you specific numbers, but I found a negative correlation between age and both total yards and YPC. See, e.g., Warrick Dunn this season. And to a much, much lesser extent Tiki Barber, thanks in part to Washington's complete inability to tackle him on Saturday night.

I didn't run multiple regressions or anything like that, but that's something to keep in mind (most likely for someone else to do) for PFP 2K7.

Bill Parcells is touching on this very topic in his press conference today. He basically said (on Jones/Barber splitting workloads) that there are some backs in the league that can do 400 carries, but they are very rare, and even if he had one of those, he'd be likely ruining the guys's career in riding the one horse...

Bill Parcells is touching on this very topic in his press conference today. He basically said (on Jones/Barber splitting workloads) that there are some backs in the league that can do 400 carries, but they are very rare, and even if he had one of those, he'd be likely ruining the guys's career in riding the one horse...

#1: Didn’t Tomlinson have 370 carries in 2002? Why isn’t he mentioned?

This brings up one other important point. Pre-Season Carries.

I don't think Tomlinson has ever carried more than 5 times in a pre-season... He is rested during the pre-season and this year the S.D. Coaching Staff has limited his practice time during the regular season to keep him fresh as well. Additionally, he has over the last couple years started avoiding the hit rather than getting an extra yard or two on long runs down the sideline.

I would think that pre-season carries added might be of use in this analysis. LT's 370 carry season did not include pre-season carries...

How many of these correlate with contract years? That is, are the downturns after big-carry seasons correlated with whether or not they just signed a fat contract and can be lazy for a year?

i am no herm edwards apologist, but the fault isn't entirely his. i don't think he calls the plays in KC. (he didn't in NY.)

i think the play caller is the important person in this theory. for example, mike heimerdinger was the oc when tennesee used up eddie george and was also jets oc last year when they let martin play in meaningless games with a serious knee injury. when the head coach is a defensive guy, i would probably lay the blame on the oc.

most injuries in football happen when a player is fatigued and not moving at full speed/ability

If this were true, wouldn't we see a lot more injuries in the 4th quarters of games and also late in the season?

A lot of the questions people have brought up here are good, but I'm sure nailing down the primary idea was the first priority.

Maybe LJ is the current example of what we might call the Terrell Davis usage pattern. In an era of free agency, and with running back being a (relatively) fungible position, why not overuse your star while you still have him at a reasonable price and ride him as far as he'll take you? When it comes time for his big contract (and probable decline), let him go elsewhere, while you try the next RB. (By this reasoning, Seattle's mistake was re-signing Alexander.)

re: 19

Herm Edwards has abused his RBs ever since he became a head coach. I don't know whether he's just stupid, or he's maliciously ruining guys careers. I'd lean towards the former and give him the benefit of the doubt.

Apparently, he doesn't believe in using backup RBs much.

Aaron - Yuri already mentioned this briefly above, but Bill Parcells has talked about it this a few times this year in relation to his use of Jones/Barber. He rang the FO note so clearly today that I thought he must have read PFP or something. Here it is (as transcribed by Grizz of bloggingtheboys.com):

Are either Marion or Julius 25-carry-a-game backs?
That's difficult. If I say no, then you say he's minimizing their ability; if I say yes, and then do that with one, I might minimize his career. Very few backs can handle 350-400 carries a year; those players are a rare exception, not the rule. Just look around, if they are an exception, how long can they remain one. These guys have different styles of running. I've had a couple of backs that you could do that with, just keep handing them the ball, and then I've probably ruined a guy by doing that, I might've cut his career short. If I had minimized some of his stuff he might've lasted longer.

#16 Jivas: Note that PAP was formulated in 1997 and PAP^3 was formulated for BP2002 which means that that effect you're noting took at least 4 years, but more realistically, the 370 carry mark is equivalent to Rany's early PAP system in the late 90s. You can already see, if you have BP sortable stats (as a member) that cat 4/5 starts and stress levels were decreasing gradually from 96-01, but post-PAP^3 the drop has been drastic.

That is to say, expect this to seep in a few years before it will really start to sink in, something that would probably be abetted by a more accurate predictive metric than a simple line, after a few yaers of this idea soaking into the ground as it has with the LJ questions and Parcells, etc. I'm not suggesting identical models here, but the 370 mark, while a fantastic piece of information, isn't yet enough for teams to work from - a true predictive model with various risk levels is perfectly suited to business strategy approach, which is why PAP^3, Stress and the cat I-V start setup is so effective.

#24: While taking out OB carries seems effective, let me be first to say that theres usually 2-3 plays a game where someone is -railed- out of bounds and smashes into things, and I'd personally adjudge the trauma of those awkward landings and slams to basically make up for all the easy carries where you just get a shove. They're violent and full speed falls that are very dangerous to everyone around, and I think the risk level basically washes out even.

Herm Edwards meet Dusty Baker

Larry Johnson let me introduce Kerry Wood

I believe that LJ didn't start at Penn State until his senior year and this is first full year as an NFL starter. Maybe these carries won't bother him as much.

If you watch LT play, you know that he is very good at avoiding big collisions... he has the power to hurt tacklers with his shoulder and run for a couple extra yards but realizes how the Charger's offense is designed around him so he doesn't risk much. If he wants to get past a defender, he'll use his patented stiff arm or just juke them out of their shoes.

On many plays near the sideline, LT will just run out of bounds getting as much as he can. In big pile ups, LT will dive forward at knee level to get as much as he can rather than bullrush his way for a half yard or so. This is one of the main reasons why LT has been able to handle such a big workload. He knows how to not take contact.

LJ should renegotiate his contract. Immediately.

I wonder if a player would ever consider a holdout in the playoffs. That would take some sack.

How do punt and kick returns figure into this?

Has FO (or anyone, for that matter) done an examination of the Dunn/Alstott years under Tony Dungy? Why wasn't the one-two punch more effective? Offensive line issues? Too equal of a division of carries? It seems that this was a pretty good (albiet extreme) complement of backs, and a good way to avoid overuse -- especially since both backs had some pass-catching ability.

Has any research been done on number of carries over a career? And when RB's begin to slide a bit? In doing a quick look, I can't find a RB with a 15 TD season after 2,300 carries.
Riggins did have 24 TD's as he hit that number but that's it. It seem that many HOF RB's start to lose production around 2,300, with Sanders being a notable exception.

Any research on this? When is a running back getting old?

41: Even when the head coach doesn't run the offense, he still oversees the whole team. Herm just seems to like a one back system.

Terrell Davis' career ending ACL injury occured when he made a tackle after a Brian Griese interception. He might be still carrying 369 times today if that hadn't occurred. Anyway, I think it would be hard to believe that any running back would say, "don't give me the ball so much, I want to preserve my knees." A player with that type of mentaility would never make it in the NFL.