2012-13 eWins

[Crow]

Any thoughts on what v-zero just spoke about Eternal / Chris?

I think you mentioned reading something about MCMC recently in your twitter feed.

[v-zero]

I don't have it available for download and have no plans to make it available at this time, so I understand I haven't provided any proof and that you are welcome to not believe me.

Well, my own boxscore metric showed a similar performance when tested last season (for 610 games (all games until March 11) with a RMSE of 11.19), even though I did not take the availability of the players into account. Thus, I actually believe you, I just would have liked to test the same sample.

I actually don't hate RAPM, but some around here (you included) seem to use it as a panacea and a crutch. "It doesn't regress well onto RAPM" seems to be a common statement, but regressing one uncertain model onto another to prove a point is pretty f*****g dumb.

How can you say "you included"? Where did I actually made a statement even remotely close to "It doesn't regress well onto RAPM"? You are pretty quick with a judgement, while again having no real proof for anything. It is rather sad, isn't it?

You keep mentioning RAPM as if it is definitely right. Earlier in this thread you questioned the value of Faried by stating as if it is fact that he damages his team, even though RAPM really has little to say about somebody who was a rookie getting relatively limited minutes last season. I have seen elsewhere you judging metrics based on how they predict RAPM values, and I'm not going to provide a load of links now, because I don't want to trawl through your posts looking for them. Perhaps 'common statement' wasn't fair, but I've seen it more than once and that's already too much.

People pass judgements all the time, and do it very quickly at that. If you read your posts in this thread they mostly involve the assumption that I don't know what I'm doing, rather than wondering whether you really understand, so do you think you haven't also passed judgement? Anyway, please for the sake of this thread and our own time can we end this?

[Crow]

v-zero is this paper of any value to you?

http://www.csss.washington.edu/Papers/wp74.pdf

[v-zero]

Looks interesting, but I can't say I have a purpose for it at the moment. This is a good paper to read for anybody interested in sports modelling: http://tolstoy.newcastle.edu.au/R/e8/he ... oles97.pdf
It's based around the English Premier League (soccer) but the methods therein are powerful and it is considered the seminal work on using MCMC and time-decayed pseudolikelihood in sports modelling.

[Crow]

Thanks. I'll read it over later.

[mystic]

You keep mentioning RAPM as if it is definitely right.

Where did I wrote that even once? Just give me one quote by me in which I say that RAPM is "definitely right" or something to that extend.

Earlier in this thread you questioned the value of Faried by stating as if it is fact that he damages his team, even though RAPM really has little to say about somebody who was a rookie getting relatively limited minutes last season.

So, you can't even read carefully enough, maybe that explains things here. At NO point did I say anything about RAPM and Faried in regard to him "damaging" his team. I didn't even say anything about him "damaging" the team at all, I just mentioned the FACT that the Nuggets with Faried on the court played about 5 points worse (per 100 possession that is) than with him on the bench.

I have seen elsewhere you judging metrics based on how they predict RAPM values, and I'm not going to provide a load of links now, because I don't want to trawl through your posts looking for them.

At no point did I ever judge a metric based on the ability to predict RAPM values. You are really off here. I judge a metric on its ability to explain and predict.

People pass judgements all the time, and do it very quickly at that.

Some may have the ability and some not. You are clearly in the latter group, obviously.

so do you think you haven't also passed judgement?

My judgement is restricted to a specific subject, the subject we were discussing. You claimed, you found a way to determine that coaches have no clue what actually helps them winning. And so far you haven't presented anything which would actually show causality here. Or to ask something directly: Do you believe that your model actually allows you to judge the coaching abilities?

[bbstats]

"So, you can't even read carefully enough, maybe that explains things here"

Cheese and rice, man. Take it down a notch.

[J.E.]

Wholeheartedly agree. The immature insults make me wish this board had an ignore button

[v-zero]

Wholeheartedly agree. The immature insults make me wish this board had an ignore button

At the risk of being put on several ignore lists, though I hope not, it is possible to put people on ignore by going to user CP and adding them to your 'foes' list.

[mystic]

"So, you can't even read carefully enough, maybe that explains things here"

Cheese and rice, man. Take it down a notch.

Wholeheartedly agree. The immature insults make me wish this board had an ignore button

Question to both of you, because it seems as if you think the statement was an insult: Why is stating an observation, that someone did not read carefully enough in order to get the point right, an insult? It wasn't meant as an insult (by no means), but I assumed that v-zero did not take the proper time to read the post, which made him believe I wrote something about RAPM being the proof that Faried is "damaging" his team.

[Bobbofitos]

Wholeheartedly agree. The immature insults make me wish this board had an ignore button

It's taken years for an insult to be hurled. It's... a little strange for this forum.

[Crow]

It is fairly easy to get annoyed about being mischaracterized or feeling that you have. You say one thing and folks think or assume more. That kind of tension can pass if you let it. It is harder to unwind, undo or correct the specific objections.

[J.E.]

Maybe "insult" was too strong of a word, but I'd really like the condescending BS to stop being posted.
Like
"Maybe that's something you missed during the conversation?"
"Is that something which came to your mind at least once?"
"You are pretty quick with a judgement, while again having no real proof for anything. It is rather sad, isn't it?"
"Some may have the ability and some not. You are clearly in the latter group, obviously."

That really has no place in this forum

Sorry for the derail. I'll stop now

[Mike G]

Cheese and rice, man. ...

Stronger language I never hope to hear.

And thanks, J.E.
Wondering now whether to split the thread and "unbury" the real topic.

[Mike G]

Speaking of Faried and the Nuggets -- Denver ranked near the bottom of MPG correlations with eWins, PER, and WS last season; and again so far this year.

Code: Select all

mpg     Den 2012         e484   PER   WS/48  
34.8   Ty Lawson         1.45   19.4   .157
33.6   Arron Afflalo      .88   14.7   .121
31.4   Danilo Gallinari  1.20   16.5   .140
27.5   Al Harrington      .97   15.3   .088
27.4   Andre Miller      1.04   14.8   .078

26.9   Wilson Chandler    .64    9.0  -.021
22.9   Rudy Fernandez     .64   13.3   .083
22.5   Kenneth Faried    1.74   21.9   .212
21.8   Corey Brewer       .73   13.7   .085
16.5   Kosta Koufos      1.24   17.3   .163

15.6   Timofey Mozgov     .94   13.2   .074
15.2   Chris Andersen    1.28   18.9   .183
9.9   Jordan Hamilton     .72   13.9   .055

306.    correlations:    .145   .052   .042

Missing are players traded midseason: Nene, McGee ...
And for this season, with numbers from a couple days ago:

Code: Select all

mpg      Den 2013        e484    PER    WS/48  
37.9    Lawson,Ty         .77    11.9    .041
37.2    Iguodala,Andre   1.11    14.6    .084
35.4    Gallinari,Danilo  .76    10.6    .059
30.3    Faried,Kenneth   1.72    22.9    .197
26.4    Miller,Andre     1.13    17.7    .136

22.4    Koufos,Kosta     1.11    15.9    .163
22.0    Brewer,Corey      .60    12.4    .090
19.4    Mcgee,Javale     1.36    24.3    .178
17.3    Chandler,Wilson   .00     3.8   -.076

248.     correlations:   .246   -.004    .011

eWins seems to "like" the top 3 more, and the bigs less, than PER and WS do.