11-09-2021, 12:00 PM
(11-09-2021, 11:41 AM)mavsluvr Wrote: I consider lineup data of the sort compiled by analytics departments a completely separate category of info than raw individual +/-, especially one-game individual +/-. Believe that multiple-player lineup data over a large enough sample can be quite useful in sorting out useful combinations of players, and as far as I know, all 30 teams use it (although, from what I understand, they have their own proprietary adjustments, and rarely make decisions based on raw data.)
(11-09-2021, 11:54 AM)mavsluvr Wrote: The only thing individual +/- tells you the truth about is what the point differential was when a particular player was on the floor (or not on the floor, as the case may be). It tells you absolutely nothing about how that player played. Doesn't tell you how many shots he made, how many he missed, whether he rebounded well, how many defensive plays he made, how well he played within the system, how he was progressing in areas of emphasis, anything that a coach would be looking at in evaluating the guy's performance in the game. Not. One. Thing. And Kidd said (I believe, honestly), that if you try to extract the truth about how a guy played by looking at a +/- stat, you're very likely to be misled.
Maybe I'm wrong to do so, but I typically use +/- to refer to "raw +/- and it's derivitive metrics". In fact, I think I'll change that habit.
And you may be right that JKidd waits for the math nerds to process the data and present it in various forms before he looks at it. Maybe so. You should ask him!!!