User talk:Gitchygoomy

January 2017

Hello, and thank you for your contributions to Wikipedia. I noticed that you recently added commentary to an article. While Wikipedia welcomes editors' opinions on an article and how it could be changed, these comments are more appropriate for the article's accompanying talk page. If you post your comments there, other editors working on the same article will notice and respond to them, and your comments will not disrupt the flow of the article. However, keep in mind that even on the talk page of an article, you should limit your discussion to improving the article. Article talk pages are not the place to discuss opinions of the subject of articles, nor are such pages a forum. Thank you. MrOllie (talk) 15:55, 17 January 2017 (UTC)[reply]

Gitchygoomy, you are invited to the Teahouse!

Hi Gitchygoomy! Thanks for contributing to Wikipedia.
Be our guest at the Teahouse! The Teahouse is a friendly space where new editors can ask questions about contributing to Wikipedia and get help from experienced editors like Cordless Larry (talk).

Visit the Teahouse

We hope to see you there!

Delivered by HostBot on behalf of the Teahouse hosts

16:04, 17 January 2017 (UTC)

Bayesian probability

Your recent edits on Bayes factor and Bayesian probability seem unnecessarily pointy.

Yes there may still (perhaps) be a few Frequentist statisticians who may deny that Bayesian probabilities are "real" probabilities, but they're a dwindling bunch.

The reality is there's wide statistical acceptance of both points of view, and that each leads to a meaningful and useful notion of probability -- it's just not the case (if it ever was) that the only acceptable "statistical" probability is the long-run limit of frequency ratio. Jheald (talk) 18:45, 6 February 2017 (UTC)[reply]

I don't think it matters, in a scientific discussion, whether or not a point of view is popular or not; it matters if you have a reasonable argument. If you think I've made statements that are inaccurate or don't make sense, then I would gladly look at them again. Why do you put the term "real" in quotes? Why do Bayesians use the terms plausible and probable interchangably? What is a "Bayesian probability" in fact? If its just a belief, then how can we have an argument? --Gitchygoomy (talk) 19:21, 6 February 2017 (UTC)[reply]

The reason Bayesians use the word "probability", I suppose, is that Bayesians see their approach as standing in the direct tradition of de Moivre and Laplace and so on creating that subject; because it obeys the laws of probability; and because it conveys their best belief in how likely an event is to happen or to have happened, or a statement is to be true. But what is more to the point here, as Ladislav notes below, is that this is how the word is used, regardless of how any of us an individuals might or might not think the word ought to be used. Jheald (talk) 20:41, 6 February 2017 (UTC)--Gitchygoomy (talk) 21:48, 6 February 2017 (UTC)[reply]

But the problem is it doesn't obey the laws of probability; this is why it is mixed up with "belief," and why likely has to be put in quotation marks. --Gitchygoomy (talk) 21:48, 6 February 2017 (UTC)[reply]

Hi, Gitchygoomy, and welcome to the Wikipedia. I reverted many of personal opinions you added to the article. You should read WP:OR that can be of help. To explain it further: "it matters if you have a reasonable argument" is not a criterium here. In Wikipedia, verifiability is the criterium. Also, do not forget to find out how verifiability is defined by the Wikipedia standards. Ladislav Mecir (talk) 20:14, 6 February 2017 (UTC)[reply]

Hi Ladislav, The Wikipedia article isn't about how popular the Bayesian program is, but about what it says. My comment about a reasonable argument was directed toward Jheald's comment. --Gitchygoomy (talk) 21:39, 6 February 2017 (UTC)[reply]

Hello again. I reverted yet another addition of yours, due to many issues. To list a couple:

Your edit referred to Cox's theorem in a way discouraged by WP:MOS.
The Cox's theorem was meant to justify common probability laws from the objective Bayesian point of view. The theorem has been criticized for various reasons, for example, as you seem to note, since its assumptions are not acceptable for the critics. Many probabilists do not care, not needing common probability laws to be justified and accepting them as axioms. Other probabilists use different justifications, such as the Dutch book approach, or various modifications of Cox's theorem using different assumptions than R.T. Cox did. Once probability laws are accepted and used by probabilits, it does not matter which justification was used. You are welcome to list the notable criticisms of the Cox's theorem, but, please, do not do that in the Bayesian probability article, where they are off topic. All such criticisms should be added to the Cox's theorem article. Ladislav Mecir (talk) 09:14, 7 February 2017 (UTC)[reply]

I would ask you to reconsider allowing a controversial view to be simply referred to as "the rules of rationality and consistency" at least, make sure they are properly referred to Cox. --Gitchygoomy (talk) 11:31, 7 February 2017 (UTC)[reply]

Also, I suggest that terms like "probability" should not be allowed to be used after the manner of Humpty Dumpty in Alice in Wonderland, i.e. any way anyone wants. You must be aware, as I have recently become, that when it comes to controversial topics, Wikipedia is used for the purpose of propagandizing some view. This is fine if the content is coherent and clear, but not when it is vague and inconsistent. In this case, it should be fair play to point out sources and logical problems. In good faith and in search of information, I turned to the Wikipedia article on Bayes Theorem, only to find that it was frankly uninformative, due to this kind of manipulation. It was, frankly, unintelligible. An older version I found in a google search for information was more clear and accurate. I don't know why it was wholly replaced by a fuzzy and misleading alternative, but it is cause for concern. --Gitchygoomy (talk) 11:37, 7 February 2017 (UTC)[reply]

Re "Wikipedia is used for the purpose of propagandizing some view. This is fine if the content is coherent and clear" - no, propagandizing is not fine. We shall strive to present the subject from a Wikipedia:Neutral point of view not polluting the text by our own "coherent and clear" propaganda. Ladislav Mecir (talk) 06:33, 8 February 2017 (UTC)[reply]

I agree of course that propaganda has no place, but sometimes it's masked by vagueness and misuse of terms (like the case which I pointed out, of "rules of rationality and consistency" being used to mask a particular and arguably unrational set of assumptions). I believe that the article on Bayes theorem has been distorted by "propaganda"--Gitchygoomy (talk) 09:07, 8 February 2017 (UTC) but others could argue that it's not. It's only by making the arguments explicit that others can judge them. I know it's a fine line.[reply]

It seems to me that this article has been written not as a general discussion of motion perception, which is what one would expect, but as a conduit to promote a particular and in my opinion very sketchy and inaccurate point of view. I have made some changes which presented speculative, untested "models" as simple facts, as well as statements that are simply false.

Sampling error

The entry on Sampling Error contains an interesting qualifier:

"Exact measurement of sampling error is generally not feasible since the true population values are unknown; however, sampling error can often be estimated by probabilistic modeling of the sample."

The "can often be estimated by probabilistic modeling" part seems a bit strange. How often? When can we know that our estimates are valid, that they fall in the nice part of the "often valid" curve? Gitchygoomy (talk) 07:08, 17 August 2018 (UTC)[reply]