Yes, I projected player WAR for the entire NBA for the 2015-16 season. Yes, I’m insane. Here it is, player and team projections – in excel for easy sorting:
The plan was to post all this before season tip off on Tuesday, but I didn’t finish the projections until the wee hours of Thursday morning. Other events happened that delayed this post – but it’s finally here.
I have the thank Kevin Pelton immensely for his player minute projections for the entire league. I don’t even attempt to try to predict how coaches/gms will allot playing time – often the playing time many players get is beyond my ability to comprehend. It drives me so crazy I won’t even try. But, Kevin does it – so I blindly take his work (with permission of course) and use it to project all the player WAR.
My player projections aren’t the run-of-the-mill type, I don’t just weight some basic overall previous NBA season(s) stat (say, Win Shares) by some basic age adjustment – compile – and call it a day.
Every player in the last two NBA, Preseason, DLeague, and NCAA seasons is rated (normalized for league, adjusted for pace, etc.) – with every rating broken down into 13 smaller ratings (2ptMade, 2ptMiss, Oreb, Steals, etc.). Using all my historical ratings for each league, I created a conversion across all 13 rating subset between leagues. I also created an NBA age progression across every subset using more recently weighted 1980-2015 NBA regular season data.
Anyway – I include all the data in the spreadsheet (within reason, I know my spreadsheet confuse many) – so anyone can see how each player’s rating was generally derived. Remember, young players with certain skill sets (say, a very young high usage point guard – much higher career “curve” & eventual “peak”, big taper at the “old” years) will see their previous NBA adjusted rating take a bigger bump than a very different player (say, a low usage rebounding big – much smaller career “curve” or eventual “peak”, less severe taper at the “old” years).
I will eventually create a HUGE ranking of a ton of players based on CAREER projections (not just this current season) and skill set breakdowns some time in the future – but I’m wanting to include many more leagues (like, every major foreign league, NBA summer league, etc) in the data set. That is taking a TON of time, and I’m not nearly ready to produce that work for public consumption just yet.
Thanks to everyone who has been asking for these projections for your patience. I am very sorry I didn’t compile and get the work formatted before the opening tip. I am a perfectionist, and I often go off on “eureka” statistical tangents that cause me to take more time than I expect sometimes. I should have stopped what I was currently working on and dove into these projections about 5 days earlier than I did. Live and learn.
Finally, for those of you that only care about team predictions – here they are so you don’t have to look at the spreadsheet if you don’t want to. The records are based MAINLY on team rating – with adjustments for team historical statistical similarities for better roster balance (the Warriors had the best similarity scores relative to historical good teams, Minnesota the worst), and conference (generally East teams with a few more wins with weaker SoS). The spreadsheet has all the team rating breakdowns, go there if you want greater detail: