Football by the Numbers: Running backs and mileage

Key issues when projecting running backs are at what point are they simply too old to bet on and how do you account for the team environment when assessing their true skill level?

For a back like Arian Foster, assessing true skill is merely a mental exercise. He is in an elite running environment in Houston and that’s not changing for him (or us) in 2013. But it’s more important for a back who is changing teams, like Steven Jackson, who also happens to now be 30 years old and someone who has had the 11th most carries before turning 30 in league history.
So it’s with Jackson where our models converge. But the idea is to use them more broadly, of course. And to be very clear regarding the use of these models (and this applies to last week’s column on quarterbacks and YPA, too ), the model isn’t the final word. It just tells us what the baseline is. It sets the bar at what is most bettable. You can bet against the model, but you have to know that you are expecting the unlikely to happen, to whatever degree, and then have some appropriately strong basis for betting the other way. You can end up right after doing this, of course. But you should be paid to gamble, meaning the player in question should be attainable at a price that’s lower than your projection. If you draft a player like Jackson around where your projection is -- a projection, remember, that violates our historical models -- then you are actually PAYING to gamble. And that’s stupid because there is, thus far, more risk than upside.
First, let’s look at the age-30-plus performance of all the 2,000-plus carry backs in this era, which we’ll set as beginning in 1993.
How did they perform in their age 30-plus seasons compared to through their Age 29 seasons. Furthermore, was relative rushing average at age 29 an indicator of future performance?
There are 23 backs in our sample who played 1,097 age 30-plus games, averaging 4.03 yard per carry as a group versus 4.26 YPC through their age 29 season, about a quarter of a yard decline. If that happens to Jackson, he’ll be just south of 4.0 this year. Seven of the 23 were better after age 30 than before, led by Tiki Barber at plus-0.44 yards per carry. Barber was better at age 29 in YPC. So, too were two other outliers -- Curtis Martin and Thomas Jones. Ricky Williams did not have an age 29 season. Fred Taylor and Ricky Watters and Warrick Dunn also defied father time -- and all performed relatively poorly at age 29.
But other great all-time rushers cratered terribly at age 30 and older. In order of biggest decline: Shaun Alexander, Eric Dickerson, Jamal Lewis, Ahman Geen, Barry Sanders, LaDainian Tomlinson, Edgerrin James, Jerome Bettis and Eddie George. All dropped at least a third of a yard is per-carry average.
Yards rushing per game, our bread and butter with running backs, dipped even more precipitously, down 29.9% after age 30. Worse, two of our outliers on the plus side in age 30-plus yards per rush, Williams and Taylor, were down 42 and 27 rushing yards per game, respectively. So it’s fair to say that only Barber really got better when it comes to annualized fantasy football value. And 16 of our 22 backs lost more than 20 yards per game, led by James’s minus-58.7.
Furthermore, if we extend our sample to the top 50 backs of all-time, they gained 1,000-plus rushing yards in their age 30 season 33% of the time versus 54% in the age 29 season (peak was about 75% in ages 26-27).
The 2,000-plus career carry backs did average 648 carries after their age 29 season. The median was 616.
Jackson owners care about volume. But at some point, rushing performance impacts rushing opportunity. It’s fair to complain that I’ve lumped all seasons after age 30 together rather than just looking at the age 30 season. Jackson owners aren’t betting on him in 2014 and 2015, after all.
Jackson, though, was much worse last year than his 4.1 yards per carry indicate. Teammate Daryl Richardson averaged 4.8 yards per carry on 98 totes. So Jackson averaged 0.8 yards LESS per carry than his backup (after accounting for rounding), which was third worst among the 22 starting backs who had a teammate with at least 75 carries. The only two starting backs worse than him last year, relative to a running back teammate, Mikel LeShoure (1.34 yards per carry worse) and Rashad Jennings (minus-2.01) are hardly great company (and Jennings was only starting due to Maurice Jones Drew’s injury). In other words, he’s likely already begun his decline.
Jackson also has been FAR worse in carries per rushing touchdown than any of the other backs in the sample and most starting backs period -- a touchdown every 42.8 carries versus the pre-age-30 average of the group of one rushing TD every 32.4 carries. Post-30, the group’s average increased about 12% to a TD every 36.3 carries. That sets the expectation for Jackson this year at about a TD every 48.1 carries.
But, the thinking goes, the rushing environment in Atlanta is so much better than what it was in St. Louis. The touchdown environment sure seems to be (whether that translates into a like number of additional goal-line runs is another matter). But the rushing environment aside from that does not appear to be. After all, the Falcons were 29th in yards per rushing attempt last year. Blame that all on Michael Turner if you want, but Turner had 38 broken tackles in less attempts than Jackson, who had just 22. He also had slightly more breakaway runs per carry (nine in 222 versus 10 in 258 for Jackson).
So our MODEL PROJECTION for Jackson in 2013: 218 carries, 869 yards, five rushing TDs. Adjust those numbers upward for how much better you think Jackson is than the runners in the group and for the environment in Atlanta, but be careful.
Also remember that a case can be made that Turner was at least as good as Jackson as a runner last year and Turner’s backup Jacquizz Rodgers, out-performed Turner in yards per rush.
BONUS COVERAGE: Let’s look at the worst starting running backs relative to the performance of backups with at least 75 carries last year and the best backups relative to the starter (i.e., they averaged more yards per rush than their team’s starting back).
Best starters: C.J. Spiller (plus-2.21 YPR), Adrian Peterson (plus-2.17), Jamaal Charles (plus-1.65), Matt Forte (plus-0.8), Reggie Bush (plus-0.77), DeAngelo Williams (plus-0.65), Marshawn Lynch (plus-0.62), Ryan Mathews (plus-0.57).
Worst returning starters: Jackson (minus-0.8 vs. backup Richardson), LeSean McCoy (minus-0.7, Bryce Brown), Mark Ingram (minus-0.64, Pierre Thomas), Ray Rice (minus-0.48, Bernard Pierce).
BONUS BONUS COVERAGE: Foster did not have a single backup at 75-plus carries. But his two top backups combined to average 5.1, which would make Foster (4.1) even worse, relatively speaking, than Jackson was in 2012 for the Rams.