The Curse of the 370 is, I think, an idea which says that running backs that accumulate 370 or more carries in one season will underperform in the following season. Some part of that idea fits under the “Duh. Of course.” category, as running backs who accumulate a lot of carries in one year are bound for some mean reversion. But the idea as a whole is one that can be tested with data.
To start with, it’s probably worth noting that each carry that a running back accumulates in one season is valuable for the following season’s fantasy results. Running backs who carry the ball 200 times in Season N, will generally outscore in Season N+1 running backs who carried the ball 150 times in Season N. Carries probably mean that the team likes what the running back does and trusts him to carry the ball. So if the Curse exists, then what you would expect to see is that at 370 carries, there is a point of diminishing returns. Instead of each additional carry meaning more fantasy points in the following year, each carry would mean fewer points.
At the bottom of this post there is a module that lets us test the idea of the Curse. The graph shows carries or touches from one season on the X axis and then shows total fantasy points in the following season on the Y axis. There is also a blue smoothed trend line which should help us identify the point at which running backs encounter diminishing returns.
When all that we consider is rushing carries, it certainly looks like the Curse exists and not only that it exists, but the point of diminishing returns might be closer to 350 carries. But if the Curse actually exists, then wouldn’t we also expect that it would apply to running backs that compile 370 combined touches between rushes and receptions? The module will also let us test that idea. Because I was curious as to whether accumulating touches might differ for running backs of various ages, I also added the ability to filter based on age.
Play around with the graph and see if it confirms what you already believed about running backs and usage.