2015-16 Team win projections

Home for all your discussion of basketball statistical analysis.
Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2015-16 Team win projections

Post by Crow » Sat Apr 16, 2016 1:07 am

How many of the top 5 this season revealed minute projections earlier in thread? How many would be willing to do so now? Could be interesting and helpful. Without we are left wondering how big a success factor that was.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2015-16 Team win projections

Post by Crow » Sat Apr 16, 2016 1:10 am

Prior year pyth wins and regression has sometimes been touted as a possibly good method, building block or at least a measuring stick that other predictors had to best to be worth paying much attention to. Did not do well this season.

kmedved
Posts: 82
Joined: Thu Jul 03, 2014 9:18 pm

Re: 2015-16 Team win projections

Post by kmedved » Sat Apr 16, 2016 2:32 pm

I took my minutes projections entirely from a fantasy basketball site, Basketball Monster (https://basketballmonster.com/). You can see the minutes projections here if you'd like to backtest: https://docs.google.com/spreadsheets/d/ ... edit#gid=0

If anyone wants to share theirs, I can try plugging in another set of minutes projections to see if they'd improve or hurt my win projections.

AJbaskets
Posts: 11
Joined: Mon Feb 03, 2014 10:17 am

Re: 2015-16 Team win projections

Post by AJbaskets » Fri Apr 29, 2016 10:03 pm

Thanks to all on the congratulations.

Should note, my projections used a blend of PT-PM and JE''s RAPM. And in terms of "best" ratings, of course I used two years of data and added aging adjustments, expected mean regression and I run the team ratings through the schedule and so forth. So there alot of little factors besides the player ratings.

Also, I just ran a retrodiction and got a ave. error of 5.5 and RMSE of 6.53. NOP, PHX and POR were the most improved with actual minutes, Charlotte had the biggest increase in error, Clifford is a genius. Should write is up on Nylon soon.

caliban
Posts: 29
Joined: Mon Oct 26, 2015 8:48 am
Contact:

Re: 2015-16 Team win projections

Post by caliban » Tue May 03, 2016 7:20 am

Congrats to AJbaskets! well deserved.

and Thanks to Pelton for providing your minutes projection. Very helpful for a first timer.

bbstats
Posts: 224
Joined: Thu Apr 21, 2011 8:25 pm
Location: Boone, NC
Contact:

Re: 2015-16 Team win projections

Post by bbstats » Wed May 04, 2016 4:51 pm

FWIW, I averaged KP and BMonster's minutes projections for mine.

Long live average error!

bbstats
Posts: 224
Joined: Thu Apr 21, 2011 8:25 pm
Location: Boone, NC
Contact:

Re: 2015-16 Team win projections

Post by bbstats » Wed May 04, 2016 5:28 pm

Just ran my retrodiction and found the following fit on *Adjusted NetRating* (not wins):

0.573 x my BPM projection
0.375 x Jerry's Multiyear + Aging Curve RPM
0.0583 x Expected Teammate Usage
0.107 x Prior yr Usg% x TS%
-6.5


Definitely overfit, but very interesting.

Crow
Posts: 6188
Joined: Thu Apr 14, 2011 11:10 pm

Re: 2015-16 Team win projections

Post by Crow » Wed May 04, 2016 7:44 pm

Could you detail how 0.0583 x Expected Teammate Usage and
0.107 x Prior yr Usg% x TS%
-6.5 get incorporated (these don't appear to be formatted into a simple pts basis) and the logic that led you to include them?

Post Reply