This article is part of our According to the Data series.
The running back position is extremely dependent on team context. Running back play will forever be tied to offensive line talent; one could probably argue the offensive line is more responsible for a running back's rushing yards than the back himself. There's all kinds of evidence of this, such as that running backs drafted in the late rounds have historically averaged slightly more YPC than those drafted in the first two rounds. That's an amazing stat; we know NFL teams aren't completely missing on running back talent, so a good portion of the late-round running back success must be due to factors outside of the running back's control.
Because a running back's talent accounts for only a small part of his output, we see stats like YPC fluctuate quite a bit from year to year. On average, only around 40 percent of a running back's YPC carries over from season to season, with the rest regressing toward the league mean of 4.25 YPC. That means that the most likely YPC for a running back coming off of a season with 5.0 YPC - an outstanding number - is around 4.55 YPC, or (5.0 * 0.4) + (4.25 * 0.6).
So while obviously we want our running backs to maintain high efficiency, it's very difficult over the long run. When a running back has an abnormally high YPC, he's bound to regress; it's not one of those stats that's easily maintained (outside
The running back position is extremely dependent on team context. Running back play will forever be tied to offensive line talent; one could probably argue the offensive line is more responsible for a running back's rushing yards than the back himself. There's all kinds of evidence of this, such as that running backs drafted in the late rounds have historically averaged slightly more YPC than those drafted in the first two rounds. That's an amazing stat; we know NFL teams aren't completely missing on running back talent, so a good portion of the late-round running back success must be due to factors outside of the running back's control.
Because a running back's talent accounts for only a small part of his output, we see stats like YPC fluctuate quite a bit from year to year. On average, only around 40 percent of a running back's YPC carries over from season to season, with the rest regressing toward the league mean of 4.25 YPC. That means that the most likely YPC for a running back coming off of a season with 5.0 YPC - an outstanding number - is around 4.55 YPC, or (5.0 * 0.4) + (4.25 * 0.6).
So while obviously we want our running backs to maintain high efficiency, it's very difficult over the long run. When a running back has an abnormally high YPC, he's bound to regress; it's not one of those stats that's easily maintained (outside of Jamaal Charles, whose career 5.8 YPC mark is unbelievable).
With that said, let's take a look at some running backs from last year who greatly overachieved or underperformed in terms of YPC. Next to their names, you'll see the most likely YPC finish for 2013 (based on a 40 percent year-to-year carryover) and whether or not they're likely to see at least the same number of carries as they did in 2012.
| RB | 2012 YPC | 2013 Projected YPC | 2012 Carries | Maintainable? |
| Adrian Peterson | 6.03 | 4.96 | 348 | N |
| C.J. Spiller | 6.01 | 4.95 | 207 | Y |
| Jamaal Charles | 5.29 | 4.67 | 285 | Y |
| Marshawn Lynch | 5.05 | 4.57 | 315 | N |
| Alfred Morris | 4.81 | 4.38 | 335 | N |
| Arian Foster | 4.06 | 4.17 | 351 | N |
| Benjarvus Green-Ellis | 3.94 | 4.13 | 278 | N |
| Trent Richardson | 3.56 | 3.97 | 267 | Y |
| Darren McFadden | 3.27 | 3.86 | 316 | Y |
So we have five players who were well above the league average in YPC and four below it. Since workload is such an important aspect of predicting fantasy points, what we're really searching for are backs who were highly efficient - and thus likely to regress - and won't be able to maintain a heavy workload. That's a recipe for a drop in production. We're also looking for talented backs coming off of poor seasons in terms of YPC, but whose workload will at least remain steady. Those backs should improve in 2013.
Looking at the chart, we see Peterson, Lynch and Morris were all extremely efficient in 2012, but they could all have a tough time replicating their overall workloads. Even if they do, it's likely that we'll see a drop in efficiency. Meanwhile, Spiller and Charles aren't as scary since their workloads - specifically Spiller's - could increase.
On the other end of the spectrum, McFadden and Richardson both represent potential improvements. Both are obviously talented and should be able to maintain their volume of carries in 2013. McFadden in particular will almost assuredly beat his 2012 average of just 3.27 YPC.
Green-Ellis and Foster could easily improve their YPC, but they don't necessarily offer too much value because they're going to see fewer carries - Foster because, well, 351, and Green-Ellis because of the presence of Giovani Bernard.
Since ADP is generally a reflection of prior-year stats (at least more than it should be), the expected changes in YPC aren't completely factored into ADP. Peterson is a special case, but there's a good chance Lynch and Morris are being over-drafted. Meanwhile, both Richardson and McFadden offer value in the back of the first and mid-third, respectively.
Jonathan Bales is the author of the Fantasy Football for Smart People series. He also runs the "Running the Numbers" blog at DallasCowboys.com and writes for the New York Times.
