A Brief Attempt to Model NBA 2k16 Ratings

Apr 8, 2015; Salt Lake City, UT, USA; Sacramento Kings center Sim Bhullar (32) towers over Utah Jazz forward Jack Cooley (45) and Sacramento Kings forward Omri Casspi (18) while waiting for a foul shot during the fourth quarter at EnergySolutions Arena. Utah Jazz on the game 103-91. Mandatory Credit: Chris Nicoll-USA TODAY Sports
Apr 8, 2015; Salt Lake City, UT, USA; Sacramento Kings center Sim Bhullar (32) towers over Utah Jazz forward Jack Cooley (45) and Sacramento Kings forward Omri Casspi (18) while waiting for a foul shot during the fourth quarter at EnergySolutions Arena. Utah Jazz on the game 103-91. Mandatory Credit: Chris Nicoll-USA TODAY Sports /
facebooktwitterreddit
Apr 8, 2015; Salt Lake City, UT, USA; Sacramento Kings center Sim Bhullar (32) towers over Utah Jazz forward Jack Cooley (45) and Sacramento Kings forward Omri Casspi (18) while waiting for a foul shot during the fourth quarter at EnergySolutions Arena. Utah Jazz on the game 103-91. Mandatory Credit: Chris Nicoll-USA TODAY Sports
Apr 8, 2015; Salt Lake City, UT, USA; Sacramento Kings center Sim Bhullar (32) towers over Utah Jazz forward Jack Cooley (45) and Sacramento Kings forward Omri Casspi (18) while waiting for a foul shot during the fourth quarter at EnergySolutions Arena. Utah Jazz on the game 103-91. Mandatory Credit: Chris Nicoll-USA TODAY Sports /

Christmas may be in December, but only on a certain day in Late September/early October do I wake up with the anticipation of a young child waiting to see what St. Nick placed under the Christmas tree. Why’s that? Because the latest installment of NBA 2K comes out, and within hours, I’m carefully maneuvering the Thunder to their elusive first title in Oklahoma City.

Unfortunately it’s currently August 6th and not September 29th (This year’s early release date!!!), so I attempted to temper my excitement by building a model to predict the NBA 2K16 ratings for players prior to the upcoming release of the game. Before delving into those ratings, let me cover my behind by discussing the limitations of the rashly made model.

  • The model seemed to be under-predicting in the 2k15 test data, so I’d suggest adding 3-4 points onto the predicted rating. I think this is because 2K adjusts their formula on top of which I’ve struggled to quantify defensive value in game terms.
  • On the topic of defensive value, great defensive centers are underrated, so perhaps think of these ratings as mostly offensive with a little bit of defensive influence. It’s by no means a perfect model.
  • Because the model is dependent of previous year(s) data, there are no predictions for incoming rookies. That’s a model I may work on later.
  • Some players have small sample sizes that affect their rating (Sim Bhullar may be good, but not this good), so take those ratings with a grain of salt.
  • Here are some of technical aspects of the model.
  • Again this is a working model, so I plan to make some adjustments in the future. Rim protection numbers, team success, and social media popularity are just a few factors that could be considered.
  • That all said, the ratings mostly pass the eye test!

Enjoy!

Data from yudelytics.silk.co