The Interpreter; Bayes Theorem; Nephites and Mayans

[_Lemmie]

Midge likes to refer to Billy Shears in the third person as "B.S."; Billy's recent comments, as noted by ihaq and honor, must have him frothing in frustration:

It is a huge anachronism that the Nephites started using the Anno Domini system at least 516 years before Dionysius Exiguus invented it;

and:

Your assertion that Dr. Coe is an undisputed expert on the Maya but is not an expert on the Book of Mormon seems to tacitly admit that the Maya and the Book of Mormon are mutually exclusive.

Honorentheos, you were wondering whether there was a "wow" cap on this paper, apparently not quite yet! You would think the authors would update their posterior odds of Shears absolutely shredding every single correspondence they bring up in the comments and just stop setting up these huge slams for him. Surely their 'skeptical priors' have been thoroughly overturned by now (which, weirdly, was a definition Bruce gave in the comments for the reason Bayes factors exist).

Any chance Billy Shears knows Professor Jenkins?

[_Lemmie]

Shears makes an excellent point about accurately considering both the numerator and denominator of the likelihood ratio, in his discussion of correspondence 3.18.

Recall that the authors' insistence that a "10% guess" results in a likelihood ratio of "0.1" directly requires that their LR denominator = 1:
p(evidence true given Book of Mormon fiction)/ p(evidence true given Book of Mormon true) = 0.1/1=0.1

The authors have objected strenuously to this observation, saying their denominator is NOT fixed, and thus allows for the evidence to be considered under the "not-H" hypothesis, that is, not-H = Book of Mormon non-fictional. However, their analysis does not support this as they, without exception, fall back into a "guess" of 1 in 10 (10% or 0.1) being directly interpreted as a LR of 0.1, which = 0.1/1, thus assuming a denominator = 1, and similarly for the other two LRs.

In the following excerpt, Shears takes them at their word that the denominator is not fixed at 1.

Billy Shears on May 29, 2019 at 11:58 am said:
Correspondence 3.18 “Calendars kept by holy men/priests”

First, on the numerator:

Emphasizing that the calendar was correct because a “just man” kept the record who proved his ability to accurately count because he was “cleansed every whit from his iniquity” and performed miracles in the name of Jesus seems like nothing more than an odd way of foreshadowing the visitation of Christ to a modern reader, and is a very Joseph Smith way of thinking (e.g.. Moroni 7:16-17).
So what is the probability Joseph would guess this, assuming the book is made up? I’ll say 0.10, but this should be interpreted in the comparison to the following probability.

and then the denominator:

Assuming this is based on Mesoamerican history:

Believing that the count of years since the birth of Christ is correct because the counter was a just man who performed miracles in the name of Jesus doesn’t fit very well in the Mayan world. If a Mayan wanted to convince us that a calendar date was correct he would talk about his training, his systems, his responsibilities as a priest, and his diligence. He wouldn’t talk about successfully performing miracles in the name of Christ.

That said, there is some overlap between a man who is believed because he performs miracles and a priest who in addition to keeping the calendars performs human sacrifices and cures diseases, so I won’t say this is a total anachronism. On the whole, I’ll give this a probability of 0.10.

and finally the likelihood ratio:

As a reminder, the likelihood ratio is the probability of the evidence assuming that the hypothesis is true divided by the probability of the evidence assuming that the hypothesis is false.

A man who is believed to be able to accurately count 33 years because he is a righteous miracle worker isn’t the same thing as a Maya Ah K’in who is the calendar keeper because that is a governmental office he inherited from his father. Thus, this isn’t an actual hit. But at the same time, a righteous miracle worker isn’t totally different from a priest either, so I don’t see it as a total anachronism that by itself weighs against historicity. So, I don’t see this tipping the scale one way or the other.

Likelihood ratio: 0.10 / 0.10 = 1.00

(a LR=1 does not change the posterior odds.)

[_Res Ipsa]

Lemmie, thanks for hammering on this point again. I finally get it.

[_Gadianton]

great update Lemmie. This is a surreal moment. I'd love to see Dale respond with a "no, Billy, here is an example of how we did our calculation..." or even, "no, Billy, here is an example of what a proper calculation would look like..."

Has Billy successfully primed the pump?

[_Lemmie]

Thanks Gadianton and Res Ipsa. Billy's really doing the heavy lifting, though, and I feel like we are seeing a Jenkins type event here, because he is not letting up. The total annihilation of the paper on every level continues.

Here he points out again the complete inconsistency in the author's methodology, as well as calling Correspondence 5:10 (likelihood ratio = 0.02) into question:

• Bruce E. Dale on May 28, 2019 at 1:54 pm said:

Richard,

You are welcome to do this test of the King James version of the Bible versus The Maya. I would be interested in the results.

But to make it a good comparison with our work, you would have to test claims of fact in both books. For example, the fact that both the Maya and the New Testament had a baptismal rite would count as a positive correspondence, but the fact that precious stones in the Bible are diamonds, pearls and that such precious stones are unknown in Mesoamerica (and unmentioned in The Maya) would count as a negative correspondence..

Bruce

----
Billy Shears on May 30, 2019 at 8:58 am said:

Hi Bruce,

You said, “For example, the fact that both the Maya and the New Testament had a baptismal rite would count as a positive correspondence, but the fact that precious stones in the Bible are diamonds, pearls and that such precious stones are unknown in Mesoamerica (and unmentioned in The Maya) would count as a negative correspondence.”

I’m confused on that last part. If the Bible mentions diamonds and pearls but The Maya does not, how can you count that as a negative correspondence? A basic tenet of your methodology that you have repeatedly defended in these comments is that “only statements of fact which are dealt with by both books can be rationally admitted to the analysis; on statements of fact where one or the other book is silent, we cannot factually assume either agreement or disagreement. There is no rational scientific basis for doing so.”

Thanks,

Billy

p.s. Pearls were unknown in Mesoamerica? 4 Nephi 1:24

Here is 4 ne 1:24:

24 And now, in this two hundred and first year there began to be among them those who were lifted up in apride, such as the wearing of costly apparel, and all manner of fine pearls, and of the fine things of the world.

The paper:

If Joseph Smith “guessed” the Book of Mormon, he would very probably have guessed “precious stones” to be the only precious stones he knew of, namely diamonds, rubies, and perhaps pearls.

Joseph Smith would not have “guessed” the precious stones to be jade, obsidian, turquoise or calcite. Nor would the names of those stones have meant anything to all but a very small fraction of those who read the Book of Mormon. (Cureloms and cumoms, anyone?)

But Joseph Smith made neither mistake. He (or rather the Book of Mormon authors) simply called them, quite accurately, “precious stones.

But obviously he did, as Billy noted, by incorrectly 'guessing' pearls, turning this 0.02 result into another 50+ result, at least, although I'll leave the final assessment to Billy, if he decides to document that one.

[_Gadianton]

It looks like Dale doesn't have the "precious stones" to respond to Billy, but he did take time for a thank you and smiley for John Minor, who commended Dale for his excellent work. He was unable to find any flaws in Dale's methodology. He says:

but I have used statistics extensively in my professional career as a flight test and evaluation expert, enough to follow the discussions. Again, I can find no fault in your methodology or your conclusion.

Wondering if I should cancel my vacation plans...

[_Bret Ripley]

... I can find no fault in your methodology or your conclusion.

I for one am grateful for this rare glimpse into the Interpreter's peer review process.

[_Lemmie]

Another part of John Minor's comment:

John Minor on May 31, 2019 at 4:04 pm said:

... What I find interesting is that despite your numerous attempts to clarify the limited scope of the objective of your research, your detractors continue to ignore that fact and choose to try to expand the discussion taking it far outside the scope of your paper.

Hmm...He must have missed Bruce Dale's (remarkably snide) comment here, in two back to back comments:

Bruce E. Dale on May 16, 2019 at 6:54 pm said:

...If you don’t want to take the Bayesian methodology seriously (or you want to avoid the necessary work—my null hypothesis regarding the motivations of some of the commentators), you can just lay Bayesian statistics aside for right now and simply compare the numbers of positive and negative correspondences: our 131 positives versus 6 negatives cited in Coe’s book (12 more negatives if you add in non-peer reviewed sources such as the Dehlin podcasts and the 1973 Dialogue article).

That math isn’t hard to do, is it?

....As long as Interpreter is willing to host this discussion, from this point onward I am going to discuss the evidence. I hope others will join me in focusing on the evidence.

As we examine the evidence, I think it will clarify why Billy Shears is wrong. Our methodology is not biased toward selecting for agreement, as Billy claims.

Instead, we are simply interested in finding out where the Book of Mormon and The Maya have something specific to say about a particular topic…and then evaluating whether the whether the Book of Mormon lines up with the facts as stated in The Maya.

So we are going to talk about the evidence (at least I am), since most of the critical commentators…… [in next comment] continued…

…most of the critical commentators seem really reluctant to do so.

Why are they reluctant to talk about the evidence summarized in our paper?

Bruce

John Minor must have missed that exchange, because that is where Billy Shears begins his systematic dismantling of the evidence, correspondence by correspondence, that I've been noting here for the last several pages.

Regarding Minor's comment about methodology and conclusions, many people have been very specific about how many problems there are on the mathematical side of this paper. It's a sea of statistical misnomers, incredibly naïve statements about the components of Bayesian and statistical analysis, and of course, multiplied conclusions based on a use of "independence" that no credible academic paper would ever use.

But back to Shears and others , I've started charting their arguments, so far I've noted the dismantling of four correspondences with LR=0.02, and one with LR=0.1, replacing them with 50+, 50+, 50+, 1, and 170. (There are at least 8 or more additional comments dismantling others.)

Even allowing for the very bad Bayes, the ludicrous independence multiplying, and the limited data set, the conclusion is not holding.

[_Lemmie]

It looks like Dale doesn't have the "precious stones" to respond to Billy, but he did take time for a thank you and smiley for John Minor, who commended Dale for his excellent work. He was unable to find any flaws in Dale's methodology. He says:

but I have used statistics extensively in my professional career as a flight test and evaluation expert, enough to follow the discussions. Again, I can find no fault in your methodology or your conclusion.

Wondering if I should cancel my vacation plans...

Back to that comment by Minor one more time, Gadianton. If a statistician really thinks the paper is valid, why wouldn't they post a substantive response? It wouldn't take much effort if the person really had a case, but this "trust me, I'm an expert" approach just smacks of a little too much time spent in MLM-Land.

[_honorentheos]

Two of my three comments were posted this morning but the following still shows up as awaiting moderation:

To carry on with this point, the fact the authors had to look outside of The Maya to find examples where Coe explicitly pointed out misses to be able to include any at all should have told the authors their approach was flawed. Each of the misses explicitly pointed out by Coe when asked for examples would have been easily identified through comparing The Maya with the Book of Mormon but would have required recognizing one can’t simply look for superficialities but needs to look at specifics of the comparative societies and environments to consistently be able to do so. Coe provided a sample of inaccuracies in an interview or paper, and the authors failed to recognize that these should have provided a template for use in identifying the misses in a consistent and wholly internal comparison between The Maya and the Book of Mormon. Had the authors remained consistent with their methodology and kept to the Maya they would have found zero misses. I credit them for recognizing how untenable that result would appear to an outside audience though many would guess the mathmatical conclusions in the paper are even worse than that and should have resulted in an equally introspective moment.