APBRmetrics

Didn't immediately find where JE did this (anyone have that thread link?) but it has been a few years. Would be interesting to see again, with offensive & defensive splits.

I think the link below might be what you're referring to:

http://www.apbr.org/metrics/viewtopic.php?f=2&t=8485

Yep. My search terms weren't quite right. Thanks.

I hope J.E. updates the Coaching RAPM. At the time, the stat was criticized for liking Scott Brooks so much. We have more time now with Durant, Westbrook, and Harden away from him so I expect his number to decline. But, the Wizards improved with him so he will get credit for that.

I respect the creativity in trying something new, but I could never buy the logic behind this working, when there is no "Off" for coaches for the entire season for almost all of them. To replicate the Off, it has to rely on a multi year sample allowing player movement on and off his team and how much previous player value was expected to improve his team vs how much they did (the difference being the coach's contribution), or the coach himself switching teams, but sample size wise player movement alone accounts for barely any variety compared to the countless lineup permutations APM works with for real players