Miracle of the Shroud IV - The Phantom Image

[JayUtah]

"Consistent with two different hypotheses" pretty much is just a different wording "Supports both hypotheses". As such it cant favour one over the other:

Correct. What Bayesian methods let you do is set aside the qualitative aspect of objective consistency and assign degrees of a priori belief to various elements of the model, saying a priori that you believe one datum is more or less supportive of various outcomes. Naturally that quantifies a set of preconceptions that thereafter pervade everything you hope to conclude from such a model. From those priors you can determine how the shape of subsequent data alters the probability of new outcomes, under that set of preceonceptions. It's very useful in driving iterative processes that incorporate quantifiable uncertainty. It's utterly useless in determining whether limestone on the cloth means it's an authentic 1st Century burial cloth, or any other affirmative question that must be decided by production of evidence, not conjuration.

Jabba fed his model a skewed set of beliefs that naturally reflected his preconception that limestone means authenticity. The skew was so egregious that variance around the 10-percent estimate for prior probability of authenticity mattered only slightly. Naturally the model worked out the way he planned it would, suggesting -- according to him, in an objective sense -- that limestone supports authenticity more than a medieval origin.

This is how unscrupulous people have misused Bayes for decades to convey a false sense of rigor. At best the computation means, "If you already believed strongly that limestone on the cloth meant it was authentic, then it would be prudent to believe that actual discovery of limestone would tend to reinforce even a scant probability it was authentic." Bayesian methods do not determine what is objectively true or advisable; they determine only what is iteratively likely or advisable under the a priori beliefs, from step to step, as new data arises.

In actual Bayesian inference that's not considered tautological because we watch how the numbers change among several hypotheses of varying uncertainty, as we apply new data. In the logical, probative sense it's highly tautological (and therefore useless). The "If you already believed..." predicate applies throughout, and that's what constitutes, in the logical sense, either a begged question or a circular argument, depending on your taxonomy.

Further, Jabba wants to bifurcate the question so that he can sweep important details under the ... cloth. "The cloth is authentic" means a certain specific set of things. "The cloth is not authentic" can mean any number of possibly contradictory things. It is this artificially created ambiguity that Jabba is intentionally using to "educate" his guesses. Because, he says, we can't know specifically how an inauthentic cloth might possibly acquire limestone, it's proper to relegate that to a low probability. In the Bayesian sense, Jabba has simply latched onto one value of θ in L(θ|x) and assumed that the rest of the likelihood function would behave as he needed it to.

And of course Jabba assiduously does not want to discuss his priors. This is the mark of someone who isn't simply fumbling around with Bayes but someone who has had experience misapplying Bayes and therefore cognizant of its inherent weaknesses. It's one thing to realize that one's foray into math was misguided. It's another thing to realize it and keep trying to hide it. One is an innocent mistake; the other is deliberate deception.

Jabba?

 

[Garrette]

Right. How then can it support one over the other?

His hypothesis is this:

Observation X supports both Hypothesis A and Hypothesis B, but it does not support them equally.

For instance:

X = My immediate environment includes an Acacia tree

Hypothesis A = My immediate environment is the Sahel

Hypothesis B = My immediate environment is Kentucky, United States

As Acacia trees are not common in Kentucky but are dominant in the Sahel, the probability of A is greater than the probability of B, at least according to Jabba.

What he is intentionally missing, of course, is that
(1) In the specific instance of limestone, A is not more probable than B and is, in fact, less likely, and

(2) There are more than 2 hypotheses. There is "the Sahel," and "not the Sahel," including acacia trees in arboretums and greenhouses and intentionally planted where not indigenous.

(3) There is other, narrowing, information that affects the probability of A, i.e., the C14 dating, or to stay with the analogy the fact that I've never been to the Sahel and the people under the Acacia tree speak with a Kentucky accent.

 

[Garrette]

Correct. What Bayesian methods let you do is set aside the qualitative aspect of objective consistency and assign degrees of a priori belief to various elements of the model, saying a priori that you believe one datum is more or less supportive of various outcomes. Naturally that quantifies a set of preconceptions that thereafter pervade everything you hope to conclude from such a model. From those priors you can determine how the shape of subsequent data alters the probability of new outcomes, under that set of preceonceptions. It's very useful in driving iterative processes that incorporate quantifiable uncertainty. It's utterly useless in determining whether limestone on the cloth means it's an authentic 1st Century burial cloth, or any other affirmative question that must be decided by production of evidence, not conjuration.

Jabba fed his model a skewed set of beliefs that naturally reflected his preconception that limestone means authenticity. The skew was so egregious that variance around the 10-percent estimate for prior probability of authenticity mattered only slightly. Naturally the model worked out the way he planned it would, suggesting -- according to him, in an objective sense -- that limestone supports authenticity more than a medieval origin.

This is how unscrupulous people have misused Bayes for decades to convey a false sense of rigor. At best the computation means, "If you already believed strongly that limestone on the cloth meant it was authentic, then it would be prudent to believe that actual discovery of limestone would tend to reinforce even a scant probability it was authentic." Bayesian methods do not determine what is objectively true or advisable; they determine only what is iteratively likely or advisable under the a priori beliefs, from step to step, as new data arises.

In actual Bayesian inference that's not considered tautological because we watch how the numbers change among several hypotheses of varying uncertainty, as we apply new data. In the logical, probative sense it's highly tautological (and therefore useless). The "If you already believed..." predicate applies throughout, and that's what constitutes, in the logical sense, either a begged question or a circular argument, depending on your taxonomy.

Further, Jabba wants to bifurcate the question so that he can sweep important details under the ... cloth. "The cloth is authentic" means a certain specific set of things. "The cloth is not authentic" can mean any number of possibly contradictory things. It is this artificially created ambiguity that Jabba is intentionally using to "educate" his guesses. Because, he says, we can't know specifically how an inauthentic cloth might possibly acquire limestone, it's proper to relegate that to a low probability. In the Bayesian sense, Jabba has simply latched onto one value of θ in L(θ|x) and assumed that the rest of the likelihood function would behave as he needed it to.

And of course Jabba assiduously does not want to discuss his priors. This is the mark of someone who isn't simply fumbling around with Bayes but someone who has had experience misapplying Bayes and therefore cognizant of its inherent weaknesses. It's one thing to realize that one's foray into math was misguided. It's another thing to realize it and keep trying to hide it. One is an innocent mistake; the other is deliberate deception.
Jabba?

This. I abandoned the idea that Jabba is simply confused or in over his head long ago because of just such behavior. Of course, he may still be confused and in over his head (I think he is), but that does not exclude deception; it is obvious here.

 

[JayUtah]

By "consistent with two different hypotheses," I mean that the info does not eliminate either hypothesis.

And that is wholly incompatible with your invocation of Bayes. As I said numerous times, you're doing what so many other fringe theorists before you have done with Bayes. You're equivocating between qualitative and quantitative methods that do not have a compatible basis. hoping your critics won't catch you.

They have, so drop the charade.

Does law use the concept of "consistent" in the issue of consilience -- and if so, does it use my definition?

The law does not formally consider consilience; that is largely a scientific and historical concept. And no, consilience in any domain is not the a priori quantification you have attempted to foist.

There is no magical mumbo-jumbo of rhetoric that exempts you from supplying affirmative evidence for authenticity, if you wish to contend that a preponderance of evidence supports it. So kindly leave off the frantic handwaving and produce evidence.

 

[Jabba]

Blood/What difference does it make?/Bayesian Statistics

His hypothesis is this:

Observation X supports both Hypothesis A and Hypothesis B, but it does not support them equally.

For instance:

X = My immediate environment includes an Acacia tree

Hypothesis A = My immediate environment is the Sahel

Hypothesis B = My immediate environment is Kentucky, United States

As Acacia trees are not common in Kentucky but are dominant in the Sahel, the probability of A is greater than the probability of B, at least according to Jabba.

What he is intentionally missing, of course, is that
(1) In the specific instance of limestone, A is not more probable than B and is, in fact, less likely, and

(2) There are more than 2 hypotheses. There is "the Sahel," and "not the Sahel," including acacia trees in arboretums and greenhouses and intentionally planted where not indigenous.

(3) There is other, narrowing, information that affects the probability of A, i.e., the C14 dating, or to stay with the analogy the fact that I've never been to the Sahel and the people under the Acacia tree speak with a Kentucky accent.

Garrette,
- Good answer. But why do you think that the probability of so much limestone being on an authentic shroud would be less than the probability of it being on an inauthentic shroud?

 

[Garrette]

Garrette,
- Good answer. But why do you think that the probability of so much limestone being on an authentic shroud would be less than the probability of it being on an inauthentic shroud?

Long ago I said I was done with my interactions with you for reasons already given. This is my one and only exception to that vow:

Read the very well informed and articulated replies of the other posters here. There are literally dozens upon dozens of them.

I am not your Least Critical Poster.

 

[abaddon]

Garrette,
- Good answer. But why do you think that the probability of so much limestone being on an authentic shroud would be less than the probability of it being on an inauthentic shroud?

Good grief, Jabba, why are you so hung up on this.

Limestone was a commonplace structural building material at the time and particularly so for medieval churches and stately homes, the very places where the CIQ was stored for 700 years. It would be odd were there NO limestone on it. Not to mention gesso at all.

 

[Mojo]

Garrette,
- Good answer. But why do you think that the probability of so much limestone being on an authentic shroud would be less than the probability of it being on an inauthentic shroud?

He doesn't, as far as I can see. He seems to be saying saying that your hypothesis, that the shroud is authentic, is less likely to be correct than incorrect. And on the evidence, that would appear to be entirely correct.

 

[JayUtah]

But why do you think that the probability of so much limestone being on an authentic shroud would be less than the probability of it being on an inauthentic shroud?

Why do you think you have any basis whatsoever by which to quantify those beliefs? You're the one who did so, so you're the one on the hook to tell us how you did it and make us believe it's numerically sound. The fact that you refuse to do so suggests you're being deliberately deceptive about your reasons for invoking Bayes.

 

[Garrette]

He doesn't, as far as I can see. He seems to be saying saying that your hypothesis, that the shroud is authentic, is less likely to be correct than incorrect. And on the evidence, that would appear to be entirely correct.

Actually, I am, but not strenuously, for reasons similar to those Abaddon summarized.

The most telling thing about Jabba's question to me is that it was, in fact, to me as opposed to those who are obviously better informed and educated on the relevant facts and analytical methods. He doesn't want a discussion with those who know what they're talking about (not that I'm completely ignorant or incapable of holding my ground, but I am certainly not the equal of many others here on this topic).

Jabba's question to me and simultaneous ignoring of others is only the latest in the long list of damning evidence that he is interested only in obfuscation and not in honest discussion.

 

[abaddon]

Garrette,
- Good answer. But why do you think that the probability of so much limestone being on an authentic shroud would be less than the probability of it being on an inauthentic shroud?

And while I am at it, you have now placed squarely upon yourself an evidentiary Everest to climb.

Now you must prove that the CIQ only ever came in contact with limestone 2,000 years ago.

You must now demonstrate that at no time in the intervening 2,000 years was it kept in any building even partially composed of limestone.

You must now demonstrate how the CIQ was isolated from any limestone over the course of it's documented 700 year history.

You must now demonstrate that nobody who handled the CIQ had any contact with limestone.

You must now demonstrate that any limestone found originates in Jerusalem and nowhere else.

In short, you have increased your burden of proof by several orders of magnitude.

 

[The Sparrow]

I don't understand. MR. Jabba, are you concerned with the facts of whether the ciq is authentic or not, or with 'winning the debate' about whether or not it is authentic?

 

[Mojo]

I don't understand. MR. Jabba, are you concerned with the facts of whether the ciq is authentic or not, or with 'winning the debate' about whether or not it is authentic?

Neither. He's trying to filibuster.

 

[Christian Klippel]

In short, you have increased your burden of proof by several orders of magnitude.

Pff, nothing that some misapplication of Bayes can't fix! Move some decimal points thataway, pluck some nicer looking numbers out of the nether-regions, done!

Greetings,

Chris

 

[Maurice Ledifficile]

I don't understand. MR. Jabba, are you concerned with the facts of whether the ciq is authentic or not, or with 'winning the debate' about whether or not it is authentic?

Neither. For Jabba, Hugh, and other enthusiasts, the game is to manufacture and prolong debate. As long as there is a semblance of controversy, the hobby lives on.


Mojo is a ninja.

 

[abaddon]

Neither. For Jabba, Hugh, and other enthusiasts, the game is to manufacture and prolong debate. As long as there is a semblance of controversy, the hobby lives on.

Perhaps there is a different target but the effect is the same. BLAARGing. Maybe we should call this SLAARGing.

 

[abaddon]

Pff, nothing that some misapplication of Bayes can't fix! Move some decimal points thataway, pluck some nicer looking numbers out of the nether-regions, done!

Greetings,

Chris

Well, one must look at the practical applications of Bayes. When I bought my kitchen table, it did not wobble nor require a folded beermat under all four legs. Hence the prior probability of a stable table able to stand is 100%. Hence Bayes proves that my able stable table can still stand despite having only two legs and residing behind the shed.

Perhaps this looks like mockery, but it isn't. It is exactly the argument Jabba is attempting to foist on everyone with respect to the shroud, no more, no less.

 

[JayUtah]

He doesn't want a discussion with those who know what they're talking about...

That's been apparent for quite some time. Fringe argumentation frequently seeks collateral reasons for ignoring problematic critics, often upon allegations that a critic is uncivil, that his criticism doesn't rise to an appropriate level of attention amid a purported deluge of criticism, or that the proponent's ad hoc ground rules for debate are not being adhered to. It's part of the larger goal of creating an illusion that ideas have been tested and vindicated.

Jabba's question to me and simultaneous ignoring of others is only the latest in the long list of damning evidence that he is interested only in obfuscation and not in honest discussion.

The subtle reversals of his burden of proof are further examples. As Bayes requires, Jabba has codified his a priori beliefs as numerical values. His critics properly seized upon those immediately as the principal flaw in his proposed method. Jabba insists his values are "educated guesses," insinuating that they are reasonable and defraying all debate on their propriety. Abbadon and others have questioned the basis of his "guesses" and are now inexplicably being asked to accept a burden of proof that the applicability of limestone is not as Jabba has codified.

Throughout we see that his arguments are simply one method after another of deflecting any sort of intellectual responsibility for his claims.

 

[abaddon]

that his criticism doesn't rise to an appropriate level of attention amid a purported deluge of criticism,

Pardon my trimmage.

HF claimed to be deluged by moddy malarkey and the onerous requirement to respond to same. Having been here for years and with something in the region of 10k posts I know this is outright bollocks. So do you and/or anyone else who has been here for any time at all. We all know it. Mostly, unless you do something really egregious, the mods are invisible. Even supposing one were to do something egregious to warrant mod scrutiny at worst one might get a PM.


Should one wish to appeal, off you go to the appeals section. It is neither difficult nor onerous.

HF would have us believe that he has retired from the arena because of the burden of answering the flood of modmails. That simply is not true.

Sorry, but it needed to be said.

 

[Christian Klippel]

Well, one must look at the practical applications of Bayes. When I bought my kitchen table, it did not wobble nor require a folded beermat under all four legs. Hence the prior probability of a stable table able to stand is 100%. Hence Bayes proves that my able stable table can still stand despite having only two legs and residing behind the shed.

Perhaps this looks like mockery, but it isn't. It is exactly the argument Jabba is attempting to foist on everyone with respect to the shroud, no more, no less.

I beg to differ slightly. If the kitchen table example were Jabba's, it would be a table with no legs, hovering above your shed.

Greetings,

Chris