$30k challenge to Interpreter’s “Team Bayes”

[Analytics]

...In fact I doubt we would disagree at all in terms of what's wrong with the Team Bayes' papers, or even with the hierarchy of problematic assumptions. I simply chose to approach this challenge from a place that, to my mind, is most easily arbitrated by the math and least easily mucked up by poor analogies and special pleadings.

I doubt we would disagree, either.

Your approach reminds me of a documentary I saw years ago about Alan Turing at Bletchley Park during WWII, trying to break the code of the Nazi's Enigma machine. One of the other codebreakers said she was studying a coded document and and trying to figure out, for example, what the coded letter "A" really represented. Turing said that was exactly backwards--what she should be doing is eliminating all of the letters A could not represent. Taking this latter approach got you to a narrowed-down problem and to a solution much faster.

If the points of "evidence" the Dales parade are perfectly correlated, the odds of the Book of Mormon being historical drops from a gazzilion-to-one to only 50-to-one. That's a sweeping step towards bringing this back down to reality.

[Dr Moore]

Your approach reminds me of a documentary I saw years ago about Alan Turing at Bletchley Park during WWII, trying to break the code of the Nazi's Enigma machine.

Jerry, stop. You had me at Turing.

[Dr Moore]

If the points of "evidence" the Dales parade are perfectly correlated, the odds of the Book of Mormon being historical drops from a gazzilion-to-one to only 50-to-one. That's a sweeping step towards bringing this back down to reality.

The Dales' article does have a perfectly correlated set of multiplied probabilities in the numerator of their un-written, made-up Bayesian conditionals. For each Bayesian point odds, if they had shown the formula, we would see in the numerator a function, P(Z) where Z = Joseph Smith thought something might be true about native American (Lehite, Nephite, Lamantie) culture based on ideas he'd heard, ideas he formulated in his mind, or things he read.

But they really gloss over that perfectly correlated factor in each Bayesian, maybe because they skipped showing the formulas.

For the Dales to make their case of statistical independence in their 131 Bayesian factors, they have to argue the negative 131 times -- that Joseph couldn't have heard or read the cultural feature of Z about native Americans, or have had a reasonable basis to "guess" Z in the Book of Mormon. (And after that, they have to go back through their analysis to make sure they haven't given Joseph credit for a "hit" if he guessed all 3 of 3 possible cultural phenomena... something we haven't really discussed but may obviate a large number of their correspondences)

Actually, it might be an interesting exercise to use Bayes on Joseph Smith himself. Say we call Y some cultural feature about the Maya, as per the Dales' framework. And as above, if Z = Y, then Joseph has a "hit." By asking questions such as:

1. If Y is written in a book Joseph owned, what is the probability Joseph read Y and remembered it?
2. Similarly, what are the odds Joseph missed Y, even if it were in a book he owned?
3. If Y were in a book not specifically known to be in Joseph's possession prior to 1829, what are the odds he read or heard about Y anyway?
4. If Y was part of Joseph's cultural stories about the Indians, what are the odds Joseph heard about Y?
5. If Y is one of many possible cultural features, what are the odds Joseph would have reasonably picked Y for the Lehites (Indians) out of all the other possibilities?

[Tom]

Episode 1 has been released: https://interpreterfoundation.org/estim ... vidence-1/

The TLDR:

It seems unlikely that a young man of Joseph Smith’s limited education could produce a book the length of the Book of Mormon as a first-time author.

Joseph Smith is a definite outlier among the nineteenth-century’s great authors, even without considering the extraordinary content of the book itself. The estimated probability that someone of Joseph Smith’s age and education would publish a book the size of the Book of Mormon as their first work is p = .0006.

Evidence Score = 3 (beliefs adjusted 3 orders of magnitude toward authenticity)

[Lem]

Episode 1 has been released: https://interpreterfoundation.org/estim ... vidence-1/

The TLDR:

It seems unlikely that a young man of Joseph Smith’s limited education could produce a book the length of the Book of Mormon as a first-time author.

Joseph Smith is a definite outlier among the nineteenth-century’s great authors, even without considering the extraordinary content of the book itself. The estimated probability that someone of Joseph Smith’s age and education would publish a book the size of the Book of Mormon as their first work is p = .0006.

Evidence Score = 3 (beliefs adjusted 3 orders of magnitude toward authenticity)

Oh dear. I just took a quick look at the new Episode 1. Is the Interpreter still insisting that the mathematician, Kyle Pratt I believe, reviewed and passed on this? That mathematician needs to take a quick exit from this before it impacts his professional reputation.

[drumdude]

The real TL:DR -

The Book of Mormon‘s length makes it 1000 times more likely to be authentic.

[Lem]

...Overall, however, it still seems quite unlikely that any nineteenth-century author would have produced a book with the length of the Book of Mormon as a first-time work....

Posterior Probability

So what does all this mean for our beliefs about the Book of Mormon? Now we get to plug all those values into Bayes’ formula and see what happens:

PH = Prior Probability of the Hypothesis (1 in 1040 chance of ancient authenticity, or p = 1.0 x 10-41)
CH = Consequent Probability of the Hypothesis (the probability of an ancient collection of records being as long as the Book of Mormon, or p = .535)
PA = Prior Probability of the Alternate Hypothesis (1 – 1.0 x 10-41 chance of Joseph Smith as author, a very very high initial estimate)
CA = Consequent Probability of the Alternate Hypothesis (our estimate of the probability of a first-time nineteenth-century author publishing a book as long as the Book of Mormon, given his education and age, or p = .00055)
PostProb = Posterior Probability (the new probability of Book of Mormon authenticity)


Conclusion
It may not look like much happened there. At first glance, you might think that we started with an incredibly small probability of an authentic Book of Mormon and ended up with an incredibly small probability of an authentic Book of Mormon. But after reviewing the evidence, our extreme skeptic would have increased his estimate of an authentic Book of Mormon almost a thousand-fold—a change of about three orders of magnitude.

How does one even respond to such nonsense? KR has concluded that the rareness of publishing a book of this length supports his hypothesis that it is truly an ancient document, provided by an angel who brought gold plates. What do you even say to such nonsense? Did the mathematician pass on this? Did he really think that such assumptions have credibility?

[Lem]

I will postulate that a book of this length by a first time author supports the hypothesis that aliens took control of Smith and his friends, and caused a document to be produced that was intended to contradict Coe's version of Mesoamerican history, specifically so the Dales could then argue ineffectively against Coe. Why? Because the aliens know Coe is correct, and because they are little shiits they wanted to embarrass the Dales when they tried a Bayesian analysis to contradict Coe. Using KR's logic and exact examples, prove me wrong.

[Gadianton]

Thanks for the observations, Lem, keep them coming.

[Dr Moore]

It's a quick read. I've seen analyses like this before, in terms of book length.

My first reactions:

1. Book of Mormon was a dictated story, not "written." Comparison with other authors isn't apples-apples - a better comparison is to look at sermonizers and preachers of the time, especially "first timers" where we have all of their speeches in written format, then add up all the sermons given over a period of ___ months (include preparation time? or just dictation time?), and compare word counts. The Book of Mormon is, after all, an oral history with sermons throughout.

(ETA: also, remove redundant phrases and Bible passages: it normalizes for original content and moves closer to a fair comparison)

2. Joseph WAS an outlier. Absolutely. Probably more than a 1 in 1000 kind of outlier, actually. 1 in a million, perhaps? How much an outlier?

Do we have any basis to rule out Joseph having some type of savant-like abilities? I argue we do not -- look at his pre-Book of Mormon history and it's clear he was special. How does a young ignoramus convince counties full of people that he alone can find buried treasure? That's one hell of a story-telling gift -- the creativity, memory, charisma. Amazing. So to make that case that Joseph was "just ordinary 3-year educated ignoramus" is just trying to prove a negative. You can't do it. And that breaks everything in this analysis. WHY? Because the probability of Joseph producing a book this long IF he had savant abilities is going to be MUCH higher than 1 in 1000.

Anyway, and this comes to my challenge and why I think Kyler is going to give up on $10k for doing the work, because he's so in love with his research and his numbers that he can't see the fundamental structural problems clearly. Joseph was an outlier and book length proves the point: in fact, that Joseph was an outlier IS THE CAUSAL correlated relationship between ALL arguments about what Joseph did. Therefore, we can NOT call various things Joseph did "statistically independent" from one another, no matter how much those "things" might be statistically uncorrelated in the general population. What did he read? How curious was he? What did he hear? Special guy - special things, right? Which means Kyler either has to stop at Joseph being 1 in 1000, or else figure out a way to control his data sets, normalize for Joseph being Joseph, or none of what comes next adds anything.

Where is Kyle Pratt?