This story is unavailable.

First of all, this piece is well-written and I appreciate your contributions to The Ringer.

You are comparing a quarter of a season to a bunch of complete seasons. As basic statistics teaches us, variability in a sample is a function of (among other things) the size of the sample. If you showed us the same graph you show us below but only for the first 40 games of the season, I can predict with near-certainty without even reviewing the data that there would be MANY seasons in which players over 35 had a negative WAR. Many of your points may still stand, but you owe your readers a disclaimer that you are dealing with relatively meaningless samples and that you will return with an update in 4 months.

Show your support

Clapping shows how much you appreciated Bradley Mankoff’s story.