I recently began tinkering with a new boxscore based catch-all stat after reading a post on the APBR metrics board. The post, by a poster named v-zero, indicated that he had come up with a simple formula for including a usage and efficiency tradeoff in linear weights based metrics. This immediately got me thinking about Alternate Win Score, the best simple box-score based all-in-one stat for predicting future outcomes. The formula for Alternate Win Score, via Neil Paine of Basketball Prospectus is: (pts+0.7*orb+0.3*(reb-orb)+stl+0.5*blk+0.5*ast-0.7 * (fga-fg)-fg-0.35*(fta-ft)-0.5*ft-tov-0.5*pf)/mp. I quickly got excited about the possibilities of including the linear usage and efficiency trade off described by v-zero in that metric to create a better metric for rating players, relatively quickly.
So that’s what I did. Introducing, the rather boringly titled, Usage Adjusted Rating. I pulled the data, updated to include numbers from last night’s game, from the NBA.com/Stats page using the pace adjusted per 48 minutes numbers. I then set a minimum minutes cutoff of 50 (roughly 5 per game in the young season) and ran the numbers. Here they are:
(For your reference, league average UAR translates to roughly 5.5–5.6)
So far about 14–15% of the way through the season, Anthony Davis looks like the most productive player in the entire league on a per minute basis. He’s not even old enough to drink. Goodness. Also, Jordan Hill, look at you! Anyway, this is by no means a perfect metric. It still has all the same problems that plague other box-score based metrics- namely, not properly evaluating defense- but it’s another fun way to look at it. I plan to try to keep the numbers updated as close to daily as I can (they will be located in a new page in the menu of the site) and maybe I will make a weekly feature out of writing about the Usage Adjusted Ratings. Maybe I’ll even come up with a better name for it. Help me out in the comments!
Stats used for this post derived from NBA.com/Stats