APBRmetrics

Thanks, this is really cool.

Great work!

One question: if you're a Tacoman, why 206 instead of 253?

I think this passage from the help page might explain why you guys are having trouble with glmnet: Note also that for "gaussian", glmnet standardizes y to have unit variance before computing its lambda sequence (and then unstandardizes the resulting coefficients); if you wish to reproduce/...

Most people should be accounting for this. Even the official +/- from the NBA does. In pbp data, I do it in the script deciding who's on the floor. There are a ton of weird nuances for these situations (for example, players subbed in during the possession but before the foul), but you can code it so...

I created a metric called "expected titles" a few years ago: http://shutupandjam.net/nba-ncaa-stats/ ... ed-impact/ ...hasn't been updated for a over a year, but it's still interesting I think.

The test I'm running is completely out of sample. For example, I am testing 2011 Data using 2010, 2009, 2008 RAPM/WS/WP/etc., That's not out of sample. BPM was found by regression onto J.E.'s 14 year rapm. The 14 year rapm includes 2011. So by definition 2011 is in sample whether you're using 2009 ...

I still believe I have the best box score metrics - but I am biased. I would put mine up against anybody's fwiw. The proof is in the pudding. :) Send me your data (preferably going back to the merger, or at least the turnover era), and I'll see how well it predicts out of sample. My email is neil.p...

Mike, I imagine you'd be interested to know that they're redoing their top 100 over at realgm right now. They've already done 1-4 (and 5 will end today, it's likely Shaq) - so they have 1.Jordan 2.Kareem 3.Russell 4.Wilt 5.Shaq Here's the current thread (for #5, they'll make a new one for #6): http:...

This is really cool and interesting, thanks

TP, is there any way that you could evaluate the deltas that you calculated versus the jump you'd expect from each player? For instance, Cousins was in his 4th season and 23 years old and Dragic was in his 6th season at 27 years old. Is there a way you could spit out a benchmark for how much you'd ...

This methodology is flawed because I'm pretty sure the RAPM you are using, ESPN J.E., already has steals factored in it so you are double counting the stat. "we can use a different measure of defensive skill, that is Defensive Regularized Adjusted Plus Minus, or DRAPM for short. This statistic...

In xRAPM, the prior is box score based stats. So if I understand it correctly, 2014 xRAPM uses 2013 box score stats as a prior? xRAPM uses multiyear data, weighted differently--X*2012,Y*2013,Z*2014. I believe J.E. uses a similar schema for both the box score data and the APM lineup data, with the d...

Lol, have you guys already given up on Otto Porter?

Very cool. To be clear, are you using npi rapm? Or something else (maybe rapm informed rapm)?

Can't say for sure without checking, but I'm guessing it's a possessions per 48 vs. possessions per game thing. I have no idea why there are three different ones though, perhaps some sites are more up to date than others.