Player movement & offensive and defensive rapm estimates

[Crow]

I took a brief and not totally satisfactory look at the frequency with which players with a -1 or worse estimate on offensive and defensive RAPM moved recently by trade or just not being re-signed. The initial finding was that far more players with -1 or worse on offensive RAPM estimate moved one way or the other than those as weak on defensive estimate. But there are probably more such players on offense than defense so I haven't computed rates and the time element probably needs to be handled more precisely. This initial snapshot though fits with my perception of insider focus and ability to discern quality by whatever means is stronger with offense than defense. More could be done with the data but without much or any insight from the inside it probably won't mean much.