2013 RAPM/IPV/etc.

[SideshowBob]

EDIT:

IPV's been updated.

http://talkingpracticeblog.com/

They've got offensive/defensive splits now.

Updated RAPM here

Here's what v-zero had to say:

EDIT 30/03/2013:

Updated with most up to date pbp data and a vast improvement on the priors.

The link in the OP will now take you to the most up to date version of RAPM for 2013 - significant improvements have been made in the implementation of the priors, data is up to date as of today on BBREF.

Spoiler:

http://www.apbr.org/metrics/viewtopic.php?f=2&t=8194&sid=3afe758334776e4b1780363bc03cb9ec


One of the guys on the APBRmetrics board built a prior-informed RAPM model (similar to the one we got used to in the last few years), and it was last updated at the all-star break.

This is basically a data dump as a result of incomplete work I'm doing - I know people are missing xRAPM and IPV (and most largely the lack of a good defensive metric) so hopefully this will help. It's built from 2000-01 onwards, with the playoffs, and with each year's ratings as the prior for the next. There is no box-score information involved, and new players to the league receive a prior according to their percentage of possible possessions played - so in 2000-01 everybody got a prior based on their percentage playing time, basically. Lambdas were found via cross-validation until it was clear that lambda settled (i.e. that the ratio of useful information contained in the prior to useful information contained in the in-season measurements became constant, which more or less happened in 2003-04). The ratings are such that possession weighted offensive ratings sum to zero, and possession weighted defensive ratings sum to zero, so that these can be interpreted with zero as the origin for for defence and offence.

Anyway, here's the link: https://docs.google.com/spreadsheet/ccc ... sp=sharing

N.B. I'm not suggesting anybody have too much faith in these, given that they aren't my end goal, but they should be as good as the estimates JE made with RiRAPM.

Here's the link to the data:

https://docs.google.com/spreadsheet/ccc?key=0AqnFn6ok911sdE1nNXNuUzFhZmFBbmU1UVBST2Vsd2c&usp=sharing#gid=0

[SideshowBob]

Garnett continues to look ridiculous

[E-Balla]

I almost forgot how much I hated prior informed RAPM. One look at the rankings and I remember immediately. Landry Fields is in the top 25? How the hell?

[kasino]

Apparently Amir Johnson is highly underrated
As much as anyone could be, never talked about third best player in the league

[tsherkin]

Minutes per game and starts should be weighted in if they arent already. Not just per-minute production, but a filter based on actual minutes played. Likewise a weight to the value of impact based on starting point.

Amir is a good player and i dont doubt he impacts our crap offense a lot, but the meaning of a huge boost to a crap offense and a moderate boost to a good one is not the same, especially as minutes are factored in. And the fact that he scores on orebs, in transition and the PnR, plus his 8,000-year long set jumper.

[Chocobanana]

Earl Clark is so low

[Rapcity_11]

Minutes per game and starts should be weighted in if they arent already. Not just per-minute production, but a filter based on actual minutes played. Likewise a weight to the value of impact based on starting point.

Nah, it's supposed to be per-possession. It's more informative that way and lends itself to additional manipulation like the ways you brought up. But the initial data needs to be per-possession.

Also, Amir Johnson might be the most underrated player in the league.

[tsherkin]

Right but it does tend to be easier to be more productive per-possession with a lower minutes load, especially with minimal work in isolation. Food for thought when comparing him to players with a different role, just as with ORTG.

[Dr Pepper]

Kobe Bryant is ranked no. 119 overall. How does that ranking compare to past seasons in this metric? One would assume it'd be higher than last season since Kobe has had a resurgence but rankings are relative

[CousinOfDeath]

Apparently KG is still playing otherworldly defense.

Luol Deng with a stellar -1.3 on offense. I thought it would be even lower.

[E-Balla]

Minutes per game and starts should be weighted in if they arent already. Not just per-minute production, but a filter based on actual minutes played. Likewise a weight to the value of impact based on starting point.

Nah, it's supposed to be per-possession. It's more informative that way and lends itself to additional manipulation like the ways you brought up. But the initial data needs to be per-possession.

Also, Amir Johnson might be the most underrated player in the league.

Led the Raps to a top 10 defense last year. Not the most but up there.

[tsherkin]

Led the Raps to a top 10 defense last year. Not the most but up there.

You mean we moved from 14th overall to 10th with him on the floor?

[ardee]

Tim Duncan as a negative on offense.... Incredible.

I think one'd better wait for the full season's worth of data before you jump all over this stat because the results are hilariously skewed.

I know KG is great but this metric has him as more impactful than '09 Wade or even his own '08 version and it's the highest defensive RAPM ever recorded IIRC, over anything Hakeem or D-Rob managed in the 90s, which is ridiculous because he's clearly not a generational level defender anymore.

[Doormatt]

Yeah I mean the rankings are interesting but you should probably take the actual numbers with a grain of salt. Not the biggest sample size.

Westbrook ahead of Durant is the most surprising thing.

[GSP]

Ive been campaigning Kg for Dpoy all year. I saw most of Celtics game and we are NOTHING on defense without him its amazing how much he anchors it.
Dirk at 5 is hilarious

[E-Balla]

Led the Raps to a top 10 defense last year. Not the most but up there.

You mean we moved from 14th overall to 10th with him on the floor?

Gave up a 103 DRTG last season while he was on the floor. Can't remember this season's numbers though.

[GSP]

WTF Kd has negative Drapm and Westbrook doesnt???????
Where is Harden

[Doormatt]

Harden is at 3.1 which is 35th in the league

[SideshowBob]

I think one'd better wait for the full season's worth of data before you jump all over this stat because the results are hilariously skewed.

While this is true, I believe mystic's talked about the fact that the ridge-regression method that is used to build this model allow it to become useful with relatively small sample sizes, and the 50 odd games up to the break should be more than sufficient.

I know KG is great but this metric has him as more impactful than '09 Wade or even his own '08 version and it's the highest defensive RAPM ever recorded IIRC, over anything Hakeem or D-Rob managed in the 90s, which is ridiculous because he's clearly not a generational level defender anymore.

First thing is that there are no comparable numbers for D-Rob or Hakeem. The figures on J.E.'s site for the 90s are a completely different stat, they're essentially built of a simulation of PbP from the 90s based on the box-score. I wouldn't really take ANYTHING from those numbers, and there's no basis to compare them to these. Also, remember that the listed 2000-2012 numbers are ALSO a different stat (xRAPM) which introduces a box-score prior. The comparison to be made here is with the old numbers from Engelmann's site that were taken down, that DavidStern managed to salvage.

Even with those numbers, the only way to compare across seasons is to standardize them in-year (Doc and mystic both harped on this quite a bit).

Where is Harden

Harden is at 3.1 which is 35th in the league

Remember that the model uses priors. Harden's number is based off his figure from last season, a figure that tried to explain his impact in a completely different role than this season. There's going to be some issues there. Further, given that he's in a new role, I would've expected less than stellar numbers anyway. Give him another season for a stat like this (and something like On/Off as well). There's no reason not to expect a jump next year.

Dirk at 5 is hilarious

Same issue with the priors. Dirk's impact was big-time in 2011 and 2012, and he hasn't piled up as many possessions this season, so that prior's got a relatively larger influence. Still, it's down from last year (he was +8.0 in the latest 2012 update) and he's down to a +6.6 already this year with only 1014 minutes played. In this case, more time might actually help.

[WhateverBro]

Garnett continues to prove how valuable he is defensively. I've been on the DPOY bandwagon the whole season and I'm not changing my stance.