Top values on new metric called "Individual Player Values"

[Mike G]

Thru games of 12/08 ...

Code: Select all

Rank    Player          IPV
   1   Kevin Love       7.1
   2   Paul George      6.8
   3   Kevin Durant     6.2
   4   Stephen Curry    6.0
   5   DeMarcus Cousins 5.4
   6   Dwyane Wade      4.8
   7   Dirk Nowitzki    4.8
   8   Reggie Jackson   4.7
   9   Roy Hibbert      4.4
  10   Chris Paul       4.1

What's wrong with this picture?

[talkingpractice]

Thru games of 12/08 ...

Code: Select all

What's wrong with this picture?[/quote]

I too am quite surprised by the absence of the King in that list, and I practically put him on top as #0 for the heck of it (he's just on the outside of the top 10 right now).   Don't blame the messenger, blame RAPM (when used with small samples).

[talkingpractice]

And lol at me, as I just realized you were referring to Reggie being in the top 10 list, and not the King missing from it.

1, Reggie is literally #1 in the NBA in our pure NPI RAPM right now.
2, MIA is outscoring opponents by like 5-7ppg with royalty on the court, and by like 13-15ppg when royalty rests.

Rather obviously, each of these trends is not that likely to continue.

[Mike G]

The more egregious lapse is LeBron not being in the top ten. And then it's less unthinkable that someone like Reggie Jackson could be temporarily associated with great plus-minus.

Back in the day, we would say any ranking system that didn't have Shaq at #1 could be summarily dismissed. No looking to fix it, just start over. LeBron has that status now, but of course we make allowances for the slow stabilization of plus-minus results over a season.

Looking at LeBron's game log, I see that when he has rested 16 minutes or more, his team has won by 16 points or more.
They're 11-1 when he plays < 35.5 minutes; 6-5 when he plays more.

[bbstats]

My little SPM and +/- blend, per 100:

Code: Select all

Player	             Team    adjRTG
1 paul,chris           lac      7.36
2 george,paul          ind      7.04
3 james,lebron         mia      6.73
4 davis,anthony        nor      6.36
5 curry,stephen        gsw      5.69
6 durant,kevin         okl      5.37
7 love,kevin           min      5.25
8 mills,patty          san      4.63
9 drummond,andre       det      4.51
10 nowitzki,dirk       dal      4.49

[Crow]

People can of course share what they want but top ten lists leave me wanting for the rest of the estimates. The players with weaker APM than expected by conventional thinking are often the findings that I find the most provocative and potentially helpful.

[talkingpractice]

any ranking system that didn't have Shaq at #1 could be summarily dismissed.

Our published values are not "player values", and never claimed to be (as no RAPM-based model would be at this point in the season). They simply are what they are. We'll put a larger explanation of what's in the model on the site in the next day or so, and then anyone can consider/ignore the data as the season progresses, as they see fit.

To say that NPI RAPM for example should be "summarily dismissed", even at this point in the season, does not make any sense to me. But again, I'm wrong tons, so who knows.

[v-zero]

In terms of the correlation between his playing time and outcomes....It seems pretty clear to me that there would be a negative correlation because he starts, and if he starts and performs well then there will be no reason to play him for long, as the game will be in hand.

[Mike G]

Our published values are not "player values", and never claimed to be...

"Individual Player Values" are not "player values", and never claimed to be? OK ...

[talkingpractice]

Our published values are not "player values", and never claimed to be...

"Individual Player Values" are not "player values", and never claimed to be? OK ...

LOL, touche. I meant to say "player rankings", and not player values. Sadly, as it was the source of some fun, having added data from the past few nights, Reggie has slipped to 17th and Lebron is now 12th, in the sense of 'rankings' (which again they are not).

We've gotten a bunch of q/a. I'm going to answer here quickly, and then we'll just update the site every few days going forward. We can often answer specific q/a on specific players not on the site (or on the methodology we're using) via email/Twitter, depending on who is asking.

- Added a brief description of the method to the site, and introduced our FORPM prior (a random forest based SPM). Its also below.

Individual Player Value (“IPV”) is a stabilized in-season RAPM model which uses a robust machine learning based SPM metric (“FORPM”) as a prior for RAPM. There is no previous season information used, to put it on par with other in-season metrics such as NPI RAPM, ASPM, or PER. The choice of a FORPM metric as prior (using an ensemble consisting of random forest regressions and gradient boosting), rather than a traditional SPM metric, was made in part to eliminate discretion in variable selection, with the goal of making IPV a “pure” metric. In addition, a properly specified FORPM model (fit to SOS-adjusted PD) performs much better out of sample than more plain vanilla regression-based models (especially with regard to ‘defense’). Due to not using any previous year info or an aging/experience curve, these values should be considered as descriptive more so than as predictive. The model is based on ‘basketball’, and not on ‘offense’ nor on ‘defense’, and as such there is only one coefficient for each player. This is again both for purity of the metric, and due to this approach being more predictive out of sample. Individual Player Values here are not meant to imply ‘player rankings’, nor are they meant to imply that they are the players value if he were to be traded, or have his role changed on his team.

- How it differs from what we posted last year: Last year's published results also used data from previous seasons. The goal for this year was for IPV to be 'pure' (like pure APM). So, IPV is simply a robust SPM used as a prior for RAPM, for purposes of stabilization. As such, it should be much more useful than plain vanilla NPI RAPM. We're calculating it and reporting it, but its not our 'absolute opinion on player value' anymore than in-season FG% is the 'absolute opinion on shooting efficiency' of the guy who first decided to divide FG by FGA.

- Expanded from top 10 to top 20 today because i like the guy at #20 today. Again, these are not rankings, perse.

- Someone that we do not know asked whether or not we like Mike G (I swear this was a true question). Simple answer is that I/we have learned a ton from his posts over many years, and continue to. And he seems like a nice guy. He appears to not be a very large proponent of practicing (and much prefers the game itself). If there are olive branches in basketball, consider one extended. And again, we've learned a ton from his posts here over the years, and I respect his opinions on the game quite a lot.

- How large of a sample do we need: Presumably less of a sample than is needed for NPI, and ofc way more of a sample than is needed for a traditional SPM or boxscore metric.

- To answer Crow's question from yesterday, the notable people not doing so well in RAPM thus far (ie net negative), but doing very well in the FORPM portion of the model (and I assume in any other sort of vanilla SPM or boxscore model): The King, Howard, Bledsoe, Drummond, Kawhi.

- As I said in another thread a few weeks ago, if there's a team out there without the internal capability for this (currently) that would like a copy of our IPV values (and the underlying components), then the first team to write us can have it all every morning via email (for all players, obviously), totally gratis. At the end of the season, we'd provide the supporting documentation to verify/calculate the results internally (also gratis). I doubt any team will take us up on that, but the offer is out there now. If a team is concerned re our other basketball businesses, then this can be done via a 3rd party (someone very known in the analytics community who we talk to every day but is not affiliated with us), and with no direct contact between us and the team. We have no interest in any gains/profit/benefit of any kind here other than that 'it would be cool' to send values to an NBA team each morning, to setup a currently non-analytics team with FORPM/RAPM/IPV, etc

[nbacouchside]

My little SPM and +/- blend, per 100:

Code: Select all

Player	             Team    adjRTG
1 paul,chris           lac      7.36
2 george,paul          ind      7.04
3 james,lebron         mia      6.73
4 davis,anthony        nor      6.36
5 curry,stephen        gsw      5.69
6 durant,kevin         okl      5.37
7 love,kevin           min      5.25
8 mills,patty          san      4.63
9 drummond,andre       det      4.51
10 nowitzki,dirk       dal      4.49

Is this basically 80% SPR + 20% plus-minus?

[Crow]

Thanks for the tidbit about the notable people not doing so well in RAPM thus far.

[talkingpractice]

Thanks for the tidbit about the notable people not doing so well in RAPM thus far.

You're welcome, no problem at all. Here's some others with bad RAPM (NPI) thus far, and these guys and the others mentioned before are a really large chunk (almost all) of the guys whose RAPM is lower than you'd probably guess: Kyrie, Afflalo, Noah has a RAPM of like negative 9 million, Plumdog surprisingly very low, Henson is -9.5 million, Oladipo also really low. Harrison Barnes -10 million and lol how/why is he starting.

[Crow]

Barnes with a -15 raw on / off +/-. Still a weak rebounder, similar to my early look at it last season ( I guess he got a little better later last season but still finished in bottom 10% in this group for OR% and bottom 20% for DR%). 6th worst OR% for a guy 6-8+ and playing 15+ minutes out of 140 players this season. 10th worst on DR%, 6th worst overall. Golden State as a whole has a problem with OR% but alright on DR%.



"Oladipo also really low." Thanks for the additional information on him and others.


Might you share what Westbrook's and Lawson's estimates are right now? It has sort of come up at a team blog.

[talkingpractice]

New values up. Lebron up to #8 now as his RAPM made a nice jump up after the past few games (incl last night).

We added the FORPM and RAPM (NPI) numbers for each player too, rather than just IPV.

Crow I'll pm you now with the values you asked for. Sorry for the delay (just saw the question).