Preface |
Old messages are archived here. Talkbacks are fine with me, as are friendly notices. You can e-mail me if you prefer, but see the disclaimer that is displayed on that page. |
I am an Eguor administrator. I promise to "offer a fair hearing to editors who present a well-documented case that they've been mishandled in some way." I do not emptily label myself so; if you need assistance, please ask and I will do my best to help. |
I am open to recall. If you think I am unfit to serve as an administrator, open a request for comment. If the consensus there is that I ought to resign, I will submit a new RFA which I must pass. See User:AGK/Recall. |
Workspace
MHP mediation resuming
Thanks AGK for the invitation. I will maybe come back in January. I have a lot of other things to catch up with. Concerning the mediation, there seem to be two main parties: the "simplists" and the "conditionalists". These reflect two main points of view of reliable souces on MHP. Sources who are university statistics teachers are mostly conditionalists. Most everyone else is a simplist. The interesting thing is that the conditionalists know by instinct or brainwashing that you *have* to solve MHP with conditional probability, but are unable to offer cogent reasons why this is legal/morally/rationally obligatory. The editors in the mediation have each stuck to their own point of view for a year or two, no-one is going to move. More importantly, no-one seems ready for anything like a compromise.
If you place the editors on a spectrum from extreme simplist to extreme conditionalist you will notice that the less extreme are making some attempt at compromise, or at least, at understanding each other's point of view. On the extreme there is total lack of interest in trying to understand the point of view of the other side. It seems to me that there will be no success as long as these extremists remain involved. It is like the Israeli's and the Palestinians. Most ordinary people will accept a bit of give and take and basically want peace and prosperity. But the extremists on each side have an interest in polarization (e.g to keep their own influential political positions).
My own point of view does not belong in either camp. I understand both points of view, and personally, I hold to another. In fact I see a synthesis, a "meta"-position. I am an experienced mathematical statistician, professor, president of the Dutch statistical society, member of the Dutch academy of sciences, see Richard Gill. Very interested these days in "science in society" and in communication between scientists and non-scientists, especially between lawyers and forensic experts. After months of mediation I wrote my own point of view up in a pair of papers, and I hardly have anything new to say. One of the papers was sollicited for publication in Springer's new International Encyclopaedia of Statistical Science, the other is going to appear shortly in Statistica Neerlandica, I am reading the proofs at the moment. For e-Prints, see [[1]] or my homepage [[2]]. I keep listening to the MHP discussions in order to hear if something new comes up. I'd be interested to be some kind of advisor or adjudicator or even a mediator but I have had quite enough of patiently explaining my points of view for the n'th time, in different words, trying for all I'm worth to understand the other side, only to be shouted at or abused.
The main problems of communication between the parties are semantics. Is the "solution" to a puzzle the answer, or the route by which you obtained it? Another big problem is differing interpretations of probability. This has kept philosophers and especially philosophers of science busy for several hundred years and no progress has been made at all. Personally (I think) I have a "higher point of view", ie, I see a synthesis, and I will probably write a paper about this sometime. Richard Gill (talk) 07:13, 11 December 2010 (UTC)
- Can we not present both points of view in the article? It seems to me that both schools of thought are widely supported, and so basically having two "answers" to the MHP explained in the course of the article would seem on the face of it to be what WP:NPOV would have happen. Am I incorrect in saying this? AGK 21:03, 11 December 2010 (UTC)
- The article has to have both points of view presented. And the published criticism of the simple solution has to be presented too, it seems to me. Next the editors will fight every inch - no sorry, every millimeter - of the way on the amount of space given to each, and the order, or proximity. Richard Gill (talk) 05:59, 12 December 2010 (UTC)
You're a funny guy, Richard. On your own talk page, on Saturday, you very clearly agreed with me that the MHP article gives UNDUE WEIGHT to the criticisms of the simple solutions.
- "(1) I agree entirely with you that the criticisms of the simple solutions are given UNDUE WEIGHT in the current article." Richard Gill (talk) 19:06, 11 December 2010 (UTC)
Now, you make a sarcastic remark about the editors who have been trying to correct that situation. Glkanter (talk) 06:09, 12 December 2010 (UTC)
- I think the editors who have been trying to correct the situation just referred to would have more success if they would show ability to understand the point of view of the other sides. If only they would separate the opinions which have been presented from the way they have been presented and the personalies of those who presented them.
- Another problem, which I have been puzzling about for a long time, is that notions like "solution" and "correct" actually mean something very different within different communities. The notion of "probability" varies greatly from community to community. Practical minded people skip logical steps and have no patience with subtleties, while academic mathematicians are professionally deformed to abhor this behaviour. How to come to a reasonable balance when 90% of the "reliable sources" are psychologists, journalists, popular science writers, while 10% are mathematical statisticians? Especially when both groups come to the conclusion "2/3, switch". Both groups will be reading the wikipedia pages and both groups must find what "their" sources say. And the real devotees will like to understand both lines of thought. Fortunately, thanks to the symmetry argument, the two kinds of solutions can be brought very close together and it is within everyone's reach, I think, to comprehend the differences. Then everyone can make their own choice, what they think is the best way to approach the problem.
I don't think the mediation will get any further as long as the "extremists" on both sides refuse to see this "middle way". Instead I am labeled as a "funny guy" who keeps contradicting himself and/or changing his mind. Criticism (in my opinion, fair and constructive criticism) is taken as sarcasm or rudeness. So I have given up. Richard Gill (talk) 19:04, 12 December 2010 (UTC)
- Glkanter: Your comment dated 06:09, 12 December 2010 was unhelpful.
Richard: Would it not be most fair and most mindful of NPOV to have a 50:50 split of the article's coverage of the two viewpoints? AGK 19:43, 12 December 2010 (UTC)
- AGK, would you please state what you understand 'the two viewpoints' to be? Because I cannot reduce that number below three. Nor do I find the various POVs are of equal prominence, in fact, I question whether they are all 'significant' as per Wikipedia policy. Glkanter (talk) 22:47, 12 December 2010 (UTC)
- I guess my response to Richard would have been better phrased had I pointed out that for Wikipedia purposes, where what the reliable sources have published, the prominence of any differing POVs, and the readers' experience are of import, "amount of space given to each, and the order, or proximity" are much more appropriate topics for mediation compared to the OR, and private, unsupportable interpretations of what sources 'really' mean, that has monopolized the mediation to date. Glkanter (talk) 23:38, 12 December 2010 (UTC)
- AGK, I think a good article would first cover the simple solutions. Then it would cover the conditional solutions. The part on the simple solutions shouldn't include any criticism of them, it should just report the solutions. Part of the coverage of the conditional solutions should include a discussion of the substantive difference between the results obtained in each approach. In the conditional solutions you do more work and you get for this a more comprehensive result. All this should be done factually, without taking sides, just reporting content. Whether the extra output is worth the extra work is up to readers to decide for themselves. And it depends on how well the writers can argue for the conditionalist point of view.
- Most of this discussion would be founded on the standard assumptions of car initiallly equally likely behind every door, host equally likely to open either door when he has a choice, the player's initial choice of door fixed, though some sources take the player's initial choice also completely random.
- I would not lay down a rule like 50:50, but let each part have its "natural" length determined by the amount of stuff "out there" which wikipedia editors want to include and can get at least a little support for including ... I think the present spectrum of editors is representative of the spectrum of readers. Writers should be aware that "less is more". If they write more on the conditional solution, less people will read it!
- After this main part of the article, other sections could come representing other approaches, e.g., in game theory and economics MHP is seen as a decision problem and is approached with the tools of decision theory. There can be a section on the controversy between the simplists and the conditionalists. There must be sections on common variants to MHP (e.g., the possible "host bias" version), and alternative modelling approaches. This automatically brings up the interpretation of probability. There are two main schools of thought on what probability means, namely the subjectivist (Bayesian) and the objectivist (frequentist). Is probability about our personal lack of knowledge, or does it reflect properties of the physical world? Epistemological or ontological? According to different probability mind-sets, different assumptions are more or less reasonable. Also according to the different mind-sets, the empirical content of the conclusion "2/3" is different, too. Richard Gill (talk) 06:47, 13 December 2010 (UTC)
- If the 'substantive difference between the results obtained in each approach' is discussed (based on which reliable sources?), in the Conditional solution section, it would seem NPOV to include, in the same section, the reliable sources that describe the symmetrical simple solutions as being equivalent to the conditional solutions. This would include Morgan, Gillman and the traditional mathematics (non-MHP) sources Richard previously provided links to. Other sources that offer both solutions, without criticisms of either, would also be discussed at this point, including Selvin and Carlton. Glkanter (talk) 07:59, 13 December 2010 (UTC)
- Morgan and myself both explicitly have "simple+symmetry" gives "conditional" under the usual "no (known) host bias" assumptions". Selvin gives many solutions side by side with no discussion or comparison whatsoever, he is just interested in the first phase: convince his readers that you have to switch. Carlton and Rosenthal both criticize the simple solution, explicitly. Rosenhouse writes a whole chapter about this criticism, he seems to think it is pretty important, though he tries to be neutral. (I think he is unconvinced by either side's arguments, and is still puzzling about the question. He relates it to the interpretation of probability issue, that was at least very perspicacious of him). The simple solution "coincidentally" gives the right answer, in the opinion of Rosenthal and Carlton, but it does not deliver the conditional probability, which they both apparently think very, very important. Or in other words: they think the simple solution solves a different problem, the simple solutions-ists are just lucky to come to the right recommendation: "switch". The reason that you *ought* to consider the conditional probability is actually because in this way you can be certain that your decision rule will be optimal. Until you have done some more work to prove optimality, for instance by remarking that by symmetry the door numbers are irrelevant and only the roles of the doors are important, you have not "solved the problem" in the sense that a mathematician uses the phrase "solve the problem". This is Nijdam's hangup: he doesn't admit that "to solve a problem" means something different to laypersons than to mathematicians. Both laypersons and mathematicians write on MHP and both laypersons and mathematicians will read the MHP pages, both laypersons and mathematicians will edit them. All editors had better try hard to understand the other's culture, even if they don't care for it at all. Glkanter and to a lesser extent Martin similarly just don't care about the mathematician's hangups. Of course they don't have to care, nobody has to, but they will have to accomodate themselves to it somehow, since both mathematicians and laypersons will go on editing the MHP pages till Kingdom come, till the sky falls in. Well, I am going to enjoy the Christmas and New Year period, and only come back to MHP next year, if it seems there is anything new to say. Richard Gill (talk) 14:55, 13 December 2010 (UTC)
- To answer Glkanter's first question: I don't think that the pedagogical device of remarking that a probability of 2/3 means two times out of three, if you think of repetitions, or of fair betting odds of 2 to 1, if you have a more subjective picture of probability, constitutes "Own Research". "1+1=2" is not OR, is it? The simple solution is "2 times out of 3, the player initially chooses a goat, and on exactly those occasions he will go home with a car if he switches". The conditional solution is "2 times out of 3 that the player initially chose Door 1 and the host opened Door 3, the car is behind Door 2". Is this OR? Surely someone has written out these "translations" in the literature, and if they didn't do that before, the reason is because they are completely obvious to every reader who thinks he or she knows what the word probability means. Why is Glkanter against translating the solutions into plain English???? Richard Gill (talk) 15:04, 13 December 2010 (UTC)
Glkanter is against polluting the Solution sections with commentary/explanations as long as your two previous responses. Which would seem to be required if the criticisms and counter-points are addressed within them. That's why I promote a separate, later, 'Controversies' section. And, of course you know I must point out you are ignoring plain English with your summary of the various reliable sources. We can discuss OR at another time.
Why don't we take this to the mediation page? Glkanter (talk) 15:49, 13 December 2010 (UTC)
The Signpost: 13 December 2010
- Rencontres Wikimédia: Wikimedia and the cultural sector: two days of talks in Paris.
- WikiProject report: WikiProject Algae
- Features and admins: The best of the week
- Election report: The community has spoken
- Arbitration report: Requested amendment re Pseudoscience case
- Technology report: Bugs, Repairs, and Internal Operational News
- Test. AGK 12:44, 16 December 2010 (UTC)
World War II arbitration case
Hi AGK, could you please look into Communicat's recent comments at Wikipedia:Arbitration/Requests/Case/World War II/Workshop? They appear (to me, as an involved party) to be frequently crossing the line into personal abuse and the comments which aren't outright abusive are becoming increasingly aggressive. For instance: [3] (editors labeled "the POV-biased elements"),[4] ("well, it's your problem now, like it or not"), [5] ("in case you're more intellectually challenged than I thought") and [6] ("it's unacceptable for people like you"). Regards, Nick-D (talk) 22:41, 17 December 2010 (UTC)
- Direct remarks concerning the conduct, current or prior, of another Wikipedia editor is an inherent part of the Arbitration process—which is why it is so unpleasant. There is a fine line between an editor explaining his opinions regarding the conduct of another editor, and an editor making unhelpful and hurtful comments. Here is my opinion, respectively, on each of the diffs you cite:
- Does he refer to anybody specifically? So far as I can see, he is talking generally, and that's fine, in an Arbitration case at least.
- Not sure on that one. I don't see why it could be offensive.
- Unacceptable; comments can be made without resorting to personal attacks. I will formally warn him in response to this diff.
- Isn't he just making a point?
- Most of all this is part and parcel of the Arbitration process. If he takes things over the line, though, then come back to me. There are still constraints on the way one makes points about user conduct in a AC case. AGK [•] 18:52, 18 December 2010 (UTC)
I've sent you an email
.. about the MHP arbitration. Richard Gill (talk) 04:21, 18 December 2010 (UTC)