APBRmetrics

huevonkiller wrote:Sorry but having papers from "low impact" journals should not matter. What matters is the argument presented before you, not if you work for ESPN or some well known place.

First of all, I was talking about scientific journals (e.g. Nature or Science, to name the two biggest), what ESPN has to do with that is really hard to understand.
Second of all, I emphasize on listing to the arguments, that is really important. The argumentation has to be consistent and the drawn conclusions have to be verified. "Low impact journals" was never an argument against WP, not even once was it used that way. Berri is arguing performatives ad hominem, pointing out his publications in order to give his argumentation for WP value. That is an fallacy, nothing else. Even if Berri would have published his stuff usually in high impact journals, that wouldn't make his metric correct. But pointing out that Berri is in fact publishing his articles in low impact journals makes it even more of a testament that we should listen to the arguments and not relying our judgement on the academic degree or the quantity of the publications.
So, reading your answer it seems I haven't got that point across.

huevonkiller wrote: Some of the most famous journals or newspapers in the world are extremely partisan.

I guess that is more a hostile media effect (a very common cognitive bias) than anything else.

huevonkiller wrote: That said I agree with your position on WP, but you really veered off at the end of your previous post. Scientists wildly debate "mainstream" ideas all the time so you're wrong there.

Seriously, it makes no sense that you are saying "I'm wrong". How would you know? Seriously, I had discussion with people claiming the "greenhouse effect" wouldn't exist. Those people used some odd "papers" from some odd "journals" in order to "prove" their point. They had no idea that the "greenhouse effect" is a consequence of the Planck law. In order to deny the "greenhouse effect" it would need to be shown that Planck was incorrect. If someone could do that, it would mean a publication in Nature or Science, it would mean that guy would become a candidate for the Nobel prize, and yet, nobody was able to show it.
The isssue is that most of those people had no clue about the very basics, they had no clue about the physics and just argued with their own biased view. And I see similarities with Berri's followers. Well, that doesn't mean that WP can't become "mainstream", but that is also effected by the low amount of really skilled people working in that area.

huevonkiller wrote: xkonk is pointing out hypocrisy, not praising Berri.

That may be true, but as you pointed out the argumentation is important, and so far xkonk has yet to show a coherent argumentation for his position.

huevonkiller wrote: Indeed, Berri might hold some incorrect beliefs but APM also produces some of the wackiest results I've ever seen.

Well, that's why smart people recommend using RAPM instead. You and Berri are just way behind the development in that area.

... just argued with their own biased view. And I see similarities with Berri's followers...

Does he really have "followers"? Are they just grad students working for free? Or some more alter-egos?
As the "bizarro" entry in basketball stats, who would do such a thing freely?

Well, that's why smart people recommend using RAPM instead.

Wow - using a fallacy in a post that quotes fallacies. Like it!

Can we get some numbers up in here rather than a wall of words?

mystic wrote:

huevonkiller wrote: xkonk is pointing out hypocrisy, not praising Berri.

That may be true, but as you pointed out the argumentation is important, and so far xkonk has yet to show a coherent argumentation for his position.

I would have imagined that the large amount of time spent discussing WP was argument enough. How about if I point out that I don't continually read about PER on the board, even though it is at least as flawed as WP but is more prominent?

motherwell wrote: Wow - using a fallacy in a post that quotes fallacies. Like it!

You showed your incompetence often enough. That statement just adds to it.


xkonk, I agree that PER is flawed. But what do you want to discuss about PER when even the creator himself admits that it has flaws? There is the difference between Hollinger and Berri, and the main difference is that one can say that PER is bad at evaluating defensive players and has flaws, while the other one is using it as the ultimate tool to evaluate player, while in fact it fails at predictions.
So far I don't see any value added by you to this discussion by just pointing out that others are also not good. And in fact the bad quality of other work does not make the quality of Berri's work better at all. So, if you want to talk about PER and it flaws, open a new thread and start a discussion.

And an ad hominem - wicked! I love how you mentioned a fallacy in a post that was an attack not on the work, but on where the work was published - an ad hominem attack. I love that sort of thing (truly I do).

But pointing out that Berri is in fact publishing his articles in low impact journals makes it even more of a testament that we should listen to the arguments and not relying our judgement on the academic degree or the quantity of the publications.

I can't disagree with that more vehemently, and I think it sums up my problem 100% with all the online stats forums. We should never listen to arguments, we should review DATA and see if it supports the argument. If that isn't something we can all agree on, there is no point to sports statistic discussions. None at all.

My big problem with so many of the stats community's arguments is that they lack numbers. The criticisms of WP all start and end with unquantified assumptions - and the key word there is UNQUANTIFIED, as in there is no actual data presented. As an example, there are interaction affects in basketball (obviously), but what they are and how big they are is never presented with supporting data. Interaction affects are something you can, with enough data, regress and find. That is why, to point at the title of this rather epic thread, Berri changed the value of Defensive rebounds in his metric - because there is an interaction affect in play on that end (but not the offensive end) for rebounds.

TL;DR Any argument against or for stats should be presented as hypothesis THEN backed up with actual data. If there isn't both a theoretical basis and supporting data, the argument is just not worth listening to.

motherwell, first look up what deduction and falsification means within the philosophy of science. Second, the numbers were already presented, you just missed them or showed your inability to interpret the numbers. Third, check out the ORB and missed shots relationship for teams, then compare that with the ORB and missed shots for individual players. You might find out something. Fourth, check the ORB% and MOV for teams, then check out ORB% and WP48 for individual players. You will again find something really interesting. Fifth, numbers are numbers, the difference between good and bad science is determined by the interpretation. If you are unable to interpret the numbers, they are pretty useless.

motherwell wrote: My big problem with so many of the stats community's arguments is that they lack numbers. The criticisms of WP all start and end with unquantified assumptions - and the key word there is UNQUANTIFIED, as in there is no actual data presented. As an example, there are interaction affects in basketball (obviously), but what they are and how big they are is never presented with supporting data. Interaction affects are something you can, with enough data, regress and find. That is why, to point at the title of this rather epic thread, Berri changed the value of Defensive rebounds in his metric - because there is an interaction affect in play on that end (but not the offensive end) for rebounds.

This was done many, many times for rebounding, and it took years for OTHER people to convince Berri that his values were wrong. Just because you don't keep up with the opposition, doesn't mean it isn't out there.

Have you read the following?

http://sabermetricresearch.blogspot.com ... nding.html

http://www.countthebasket.com/blog/2008 ... g-returns/

mystic wrote: And do you really not understand the inconsistency in Berri's metric? Either we can determine the intrinsic values of the boxscore entries via regression or we can not do that. If we can, no adjustment should be necessary, if we can't, the hypothesis is rejected. The hypothesis is not proven correct, when we make adjustments contrary to the hypothesis. Do you understand that point?

Great point - one that I find VERY true. I never understood how regressions were used at the team level to find their value at the individual level - but then that "value" seems completely undermined by a position adjustment. As I said years ago - isn't that position adjustment pretty much an admission that the original premise doesn't work? If the regressions tell us that 19 of the top 20 NBA players are post players (almost half of which are bench players) - then either accept it as truth or reject it. As far as I'm concerned, further manipulation (ie position adjustments) of the results rejects the original hypothesis.

/Lawyered

mystic wrote:xkonk, I agree that PER is flawed. But what do you want to discuss about PER when even the creator himself admits that it has flaws? There is the difference between Hollinger and Berri, and the main difference is that one can say that PER is bad at evaluating defensive players and has flaws, while the other one is using it as the ultimate tool to evaluate player, while in fact it fails at predictions.
So far I don't see any value added by you to this discussion by just pointing out that others are also not good. And in fact the bad quality of other work does not make the quality of Berri's work better at all. So, if you want to talk about PER and it flaws, open a new thread and start a discussion.

I don't want to talk about PER; I also don't want to talk about WP if the content of it is 'man, look how wrong this is'. My point is, and has been throughout this thread, that it is obvious that 90%+ of the people here know that both are flawed. However, I still have to slog through posts that contribute nothing beyond once again pointing out that someone doesn't like WP. There isn't, as far as I can tell, a good reason to do so besides that Berri is unliked. Both metrics are flawed and have lots of discussion about why; PER is far more prominent than WP. If I knew about the various measures but nothing about the board, I would assume that everyone here would be outraged that ESPN provides a national outlet where they only use PER and APM as the representatives of the statistical community. That has not been my experience.

If that still isn't clear enough, maybe I can phrase my point as a question: Why are we still talking about WP?

I would assume that everyone here would be outraged that ESPN provides a national outlet where they only use PER and APM as the representatives of the statistical community.

I didn't even know that espn used adjusted plus-minus. That's pretty cool, actually.
PER is the child of their man Hollinger, so any other box-score metric would be in competition.
People aren't generally outraged, I'm guessing, because:
- PER and APM are a big advancement over per-game averages used elsewhere.
- We don't actually rely on ESPN's choice of stats for our information. Do we?

For some reason, "Berri" in a thread title still creates a lot of interest. A lot of not-Berri topics are discussed in these threads, which could be discussed in other threads; but some people just get energized by WP.

mystic wrote:
First of all, I was talking about scientific journals (e.g. Nature or Science, to name the two biggest), what ESPN has to do with that is really hard to understand.

No, not quite. What do greenhouse gases have to do with Berri? Come on, don't be obtuse we know what you were doing.

This is what you said:

"In fact your posts have huge similarities to posts written by people claiming climate change wouldn't exist. They also bringing up some "papers" published in low impact journals in order to make it seem as if that would be legit."

I'll spell it out for you.... ;]

You were trying to make an analogy and your prose was flawed. I think Berri is treated with a double-standard even if his WP statistic is silly.

Second of all, I emphasize on listing to the arguments, that is really important. The argumentation has to be consistent and the drawn conclusions have to be verified. "Low impact journals" was never an argument against WP, not even once was it
used that way.

You veered off topic and diluted your position, then I made a valid notation. Perhaps your analogy should be more refined next time.

Your attitude seems odd to me when it comes to outliers in WP.

Berri is arguing performatives ad hominem, pointing out his publications in order to give his argumentation for WP value. That is an fallacy, nothing else. Even if Berri would have published his stuff usually in high impact journals, that wouldn't make his metric correct. But pointing out that Berri is in fact publishing his articles in low impact journals makes it even more of a testament that we should listen to the arguments and not relying our judgement on the academic degree or the quantity of the publications. So, reading your answer it seems I haven't got that point across.

This is not what you said, you went on to make a grand generalization and politicized the thread, IMO.

Your comments on Berri are mostly spot on. But some of your comments are extremely distracting and draw attention away from your more salient positions.

Seriously, it makes no sense that you are saying "I'm wrong". How would you know? Seriously, I had discussion with people claiming the "greenhouse effect" wouldn't exist. Those people used some odd "papers" from some odd "journals" in order to "prove" their point. They had no idea that the "greenhouse effect" is a consequence of the Planck law. In order to deny the "greenhouse effect" it would need to be shown that Planck was incorrect. If someone could do that, it would mean a publication in Nature or Science, it would mean that guy would become a candidate for the Nobel prize, and yet, nobody was able to show it.

Again this is not what you said, scroll up if your memory is unclear.

Further those are not the arguments against the greenhouse effect either, you must not know all the counterarguments very well.

Many scientists dispute the magnitude, opportunity cost, benefits, damage, deaths, etc. of climate change.

The isssue is that most of those people had no clue about the very basics, they had no clue about the physics and just argued with their own biased view. And I see similarities with Berri's followers. Well, that doesn't mean that WP can't become "mainstream", but that is also effected by the low amount of really skilled people working in that area.

Perhaps in some cases, but not always. I'm not sure it even matters if it is real or not, there are more important problems in the world.

Many scientists have no basic understanding of economics, or opportunity costs. Their input is limited and incomplete, their warnings and hysteria can even be dangerous. There is plenty more to consider on this subject actually.

Simply having a Physics degree means nothing, since this is not just a "scientific" issue.

That may be true, but as you pointed out the argumentation is important, and so far xkonk has yet to show a coherent argumentation for his position.

Hmm that's not what xkonk said.

Well, that's why smart people recommend using RAPM instead. You and Berri are just way behind the development in that area.

RAPM can be just as noisy too.

motherwell wrote:

Well, that's why smart people recommend using RAPM instead.

Wow - using a fallacy in a post that quotes fallacies. Like it!

Can we get some numbers up in here rather than a wall of words?

I've always preferred RAPM to APM for the record.

I just think that when Peja leads the league in APM for an entire year, it throws off my respect for that stat. When similar ridiculous (every once in a while) results pop up in RAPM, it makes me chuckle as well.

bbstats wrote:Came in expecting to have to argue with huevon. But I agree with him for once!

Ironically, Wins Produced shows up in economics stuff (I remember feeling really gross reading it in my Sports Econ class)...to that end, I would say that WP is somewhat mainstream.

Sports Econ? Sounds like an incredibly fun class, I'm jealous. I've only taken bland economic courses so far.

Anybody have thoughts about this study?
http://weaksideawareness.wordpress.com/ ... /#comments
I haven't stared at it enough and am not really in the mood to do so right now but thought I'd ask if anybody had thoughts to make decoding quicker or make the task more interesting.