SITE NEWS: We are moving all of our site and company news into a single blog for Sports-Reference.com. We'll tag all PFR content, so you can quickly and easily find the content you want.
Also, our existing PFR blog rss feed will be redirected to the new site's feed.
Pro-Football-Reference.com » Sports Reference
For more from Chase and Jason, check out their work at Football Perspective and The Big Lead.
ShareThisQuestions (without answers) about running back workloads
Much electronic (and acutal) ink has been spilled over the years on the topic of running back workloads. This year's Pro Football Prospectus has an article about whether receptions and playoff workloads are worth adding to the list of considerations, concluding no and yes respectively.
The question is vaguely reminiscent of the debates over pitcher workload that rage in the baseball stat geek community. At least they used to rage back when I was keeping up; I assume they still do. In some sense, the situations are similar. In both cases, you're dealing with positions where serious --- often career-ending --- injuries are more likely than for others in the same sport. Also in both cases, it appears that some of the injuries or drops in effectiveness might plausibly have been caused by workload while others were most likely just dumb luck. On the other hand, the situations are different in some ways. Most notably, pitching injuries are about repetitive stress while running back injuries are about violent contact.
Or are they? It is certainly intuitive that a high workload for a running back would lead to a greater probability of injury or loss of effectiveness, but it's not clear to me what the exact link is. What exactly happens on those 350 carries that might cause problems the next year? Is it the hits? The cutting? Is it muscles? Joints?
Question 1 - forgetting the empirical evidence for a minute, what is the theoretical basis of these ideas? Does a high workload in Year N mean an increased likelihood of a torn ACL in Year N+1? Receivers do a lot of cutting. Linemen do a lot of rolling around in piles. I could be wrong, but I don't think players at those positions tear ligaments at the rate running backs do. But running backs do a lot of both (cutting and rolling in piles). Does that combination cause ligaments get "tired" in a way that five or six months of complete rest doesn't fix?
Also, should the injury question be separated from the loss-of-effectiveness question? If the medical argument for high workloads contributing to injury is that ligaments get slowly weakened by a high-workload season and don't have time to heal, then does that same medical argument explain loss of effectiveness as well?
Question 2 - Could we learn something by looking at not only raw numbers of carries but also the distribution of those carries?
Question 2A - if it is true that the offseason isn't a long enough time to heal whatever damage is done during the year, then should we weight late-in-the-season carries heavier than early-in-the-season carries? If so, this might help explain Jamal Lewis' torn ACL in training camp of 2001. In his rookie year (including playoffs), he had 286 carries from week 10 on. In particular, maybe postseason carries should not only be counted, they should be counted extra.
Question 2B - is a consistent N carries per game less damaging than the same number of carries accumulated in a less consistent way? That is, is a 30-carry game twice as damaging as a 15-carry game? Or moreso? Last year, Clinton Portis had 385 rushes and Larry Johnson had 336. But Johnson had five games of more than 30 carries and Portis had none. If you assume the first 20 rushes of each game are "free" (such assumptions are often made in baseball --- the damage starts to pile up only after 70 or 80 or 100 pitches --- I have no idea if those assumptions are justified), then you might argue that Johnson was the most overworked back in the NFL last year. Here are the leaders in "Rushes over 20"
RSH Ov20
====================================
Larry Johnson 2005 336 84
Edgerrin James 2005 373 74
Shaun Alexander 2005 430 71
Tiki Barber 2005 370 60
Willis McGahee 2005 325 53
Cadillac Williams 2005 308 52
Clinton Portis 2005 385 52
Rudi Johnson 2005 350 46
Thomas Jones 2005 334 43
LaDainian Tomlinson 2005 339 37
Julius Jones 2005 257 34
Domanick Davis 2005 230 32
Lamont Jordan 2005 272 29
Samkon Gado 2005 143 26
Fred Taylor 2005 202 25
Reuben Droughns 2005 309 23
If you look at Rushes Over 15, the leaders are Alexander, James, Johnson, and Portis. If you look at Rushes Over 25, it's Johnson, Alexander, James, and Cadillac Williams. Considering that Johnson also had most of his carries late in the year (though not as late as Shaun Alexander), might he possibly be the back most at risk of serious injury or decline this year?
Questions to get the discussion going: Who is more at risk of serious injury this year, Larry Johnson or Clinton Portis? Which of them is more at risk of severe decline? Also, whatever you estimate the probability of injury or decline for each to be, how much of that is attributable to their workloads from last season? In other words, say the following factors contribute to the chance of Portis getting hurt this year: his own unique physical characteristics (genetics, etc), his workload from last year, and dumb luck. What are the relative weights on those three (or however many) factors?
This entry was posted on Thursday, August 10th, 2006 at 4:42 am and is filed under General. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.
As to the last questions, I would say 1) dumb luck, 2) genetics, and 3) workload.
As to the first series of questions, one of my problems with your original workload study was that it lumped Tiki Barber, who had about 15 carries/game over several years early in his career, with a running back who may not have played as much in some seasons, but then had a much heavier workload over a short stretch of seasons. Maybe you could look at one-year, two-year, or three-year stretches and rates (attempts/GP) setting some sort of baseline for total attempts as well, and compare different groups in years N+1 and N+2 in things such as YPC difference and games played. Then we could see if the 24+ per game guys were affected more than the 18-20 per game guys.
As an analogy (though admittedly not a perfect one), picture guys who are doing 100 lbs bench presses twice a day. Assuming they are equal to begin with, If guy A is doing 15 reps a day 2x per day for 8 days, and one is doing 30 reps a day 2x per day for 8 days, which is more likely to cross the line and do some long term, even if not acute, harm. It's not a perfect example, because the guy doing more reps would be getting stronger so long as he was recovering properly and spacing the workouts, whereas I dont think runners get substantially physically stronger at running the more reps they get (they may get better at it, which is different).
I don't know that any carries are free, but I suspect it is not linear and the difference between 10 and 15 is not as great as between 30 and 35. As a thought experiment, if the Chiefs made Larry Johnson carry the ball on every offensive play for a game (60+), wouldnt we expect him to have a much, much higher risk of injury because of fatigue at the end of that game or in the next few weeks.
From the theoretical perspective, I think the fact that RBs get hit often is a contributing factor, but I'd be interested in knowing if the fact that they often get hit by much bigger players also comes into play. IOW, a WR does a lot of cutting, doesn't get hit often, but when they do, they usually get hit by someone roughly the same size. OL don't cut much, but they get hit every play, again usually by players the same size. RBs cut a lot, get hit a lot, and often by players much bigger than them.
So if we could distinguish between runs that don't end in a hit (out of bounds or TD), runs that end in a tackle by a DB, and runs that end in a tackle by a DL or LB, I'd expect (theoretically) the last to be more punishing. Of course, "ending in a hit" is itself problematic, but I don't imagine we could further distinguish between breaking tackles against various defensive positions.
"So if we could distinguish between runs that don’t end in a hit (out of bounds or TD), runs that end in a tackle by a DB, and runs that end in a tackle by a DL or LB, I’d expect (theoretically) the last to be more punishing."
Sounds like a great idea, let's start breaking down some game tape. I'll take the AFC, you take the NFC. Seriously, though, that would be a great statistic to have, so all we need to do is convince the NFL (or whoever is in charge of that stuff) to put it in the boxscores.
These are all very good questions, and I agree that we (and football coaches) need to be very cautious about jumping to hasty conclusions about RB abuse/overuse and whether it increases the risk for injury or decreased performance. I believe baseball people have overreacted somewhat to some initial incomplete research on pitch counts, a conclusion underscored by Bill James in the recent Neyer/James book. Pitch counts are indeed still very controversial even among sabermetricians.
Among the conclusions of Carl Prine's research (unpublished, I believe, but described in detail in the linked FO thread) is that high carry backs are at higher risk of both catastrophic injuries and joint deterioration. But I wonder whether that risk increases exponentially (rather than linearly) after X number of carries, which would be relevant to JKL's hypothetical 60+ carry game. If so, a solution might be to save your best RB for high-leverage situations, similar to your ace relief pitcher, though not necessarily just at the end of the game since you can substitute more freely in football.
http://www.footballoutsiders.com/ramblings.php?p=2650&cat=1
physical characteristics certainly must include age. there is no arguement that recovery times for a 23 year old is short than for a 31 year old, is there? I assume this holds from both week to week recovery times and season to season recovery times. I also think that once a player is a super star (and is paid accordingly) they are much more likely to hang them up sooner in the decline period. So, this might be some sort of desire factor - unrelated to the injury question, but if you are trying do decide who to keep in december, you gotta figure the Willis McGahee's of the world are working a lot harder to come back than the Priest Holmes'. just a guess but not everyone has a record to catch like Emmitt Smith did (he wasn't terrible in his final year either). Bettis could probably get a one year contract somewhere - but why should he. Antwain Smith probably has a mortgage payment. what were we talking about?
Has anyone considered the possibility that the defenses a running back faces might factor into the injury rate, i.e. that they are at a higher risk of injury against certain teams, or certain defensive schemes? Just a thought.
Also, could the offensive line play a role in the injury rate?
Finally, could the playcalling have any effect? For example, is it more dangerous to run up the middle, or off towards the sidelines?
Jim A, I am currently (probably temporarily) boycotting ESPN Insider, so I can't read the Schatz article on Page 2.
Does Prine's research suggest that more carries means greater catastrophic injury risk per carry, or just greater catastrophic injury risk in absolute terms. The latter would seem too trivial to mention, since injuries obviously occur more often on the field than they do on the bench (using carries as a proxy for playing time).
But the former point, if true, would be interesting.
I can see how joint issues would be cumulative, so more carries would result in greater injury risk (per carry) overall. But if we're talking about torn ligaments as opposed to joint/muscle issues, a cumulative effect would surprise me a bit.
Oops. Formatting blunder. Only the first two lines in quotes. (I must have ommitted a slash.)
dude how do you comment on your own comment. that is AWESOME
ok, what is going on - this is just weird.
doug, maurile broke your website.
well, it was fun while it lasted.
see ya!
Maurile, you don't have to click through to ESPN.com, just read the comments in the link I provided. There's a lot of noise to wade through, so mostly just read the posts authored by Carl. Posts 44, 58, and 97 are the most relevant. Apparently the risk becomes higher after 300+ or so touches (carries + receptions) in a season.
Alex, I'm pretty sure I remember Carl writing that he found no evidence that certain defenses "cause" injuries at higher rates. At the team level, they are pretty random except for being higher on AstroTurf.
Just reading through this, I think the question that really stands out to me is 'what is an injury?' Many of these guys have minor surgeries during the off season to 'clean up' their joints. It would seem to me - based on no evidence whatsoever - that the accumulation of these types of injuries is more likely to be the result of additional workload.
The catastrophic injuries seem to me to be more based upon specific types of contact than wear.
[...] In 2004 including the regular season and playoffs Dillon carried the ball 375 times at the age of 30. Research has shown (Doug Drinen writes here and here) that this is the age at which running back begin to slow down and 375 carries is a high amount for a back at any age. It was not surprising to see Dillon get hurt last year. He ended up playing in eleven games and averaging nineteen carries per game. I’d be surprised to see Dillon carry the ball 150 times this season. [...]