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

[Dr Moore]

Colleagues,

With pandemic travel limits, I saved 15 months of departmental T&E budget, and have been at a loss for where to deploy it. But good news, we have a project!

I have offered out $30,000 in possible donations to the Interpreter Foundation, $10,000 apiece in the name of the three contributors on, let’s call it “Team Bayes,” who have or are aggressively promoting a Mopologetic agenda based on achieving astronomically impossible odds-in-favor by multiplying probabilities.

In some of these arguments, probability multiplication is strangely being performed in the name of Bayesian analysis, even though no proper Bayesian analysis has been performed. In any event, it’s like pornography — looks nice, but a cheap imitation of the real thing.

Here, for your convenience, is the Interpeter’s Team Bayes.

John Gee
https://journal.interpreterfoundation.o ... f-abraham/

Bruce and Brian Dale
https://journal.interpreterfoundation.o ... t-guesser/

Kyler Rasmussen (reviewed by Dr Kyle Pratt)
https://interpreterfoundation.org/estim ... vidence-0/

I am putting our department’s hard-earned funds on the line to back the widely-held assertion that the mathematical treatments these authors have put forward amount to academic bankruptcy, garbage-in/garbage-out, intellectually dishonest, lazy, and intentionally misleading.

If they disagree, and apparently they each do, then the challenge is simple. Prove that the probabilities they’re multiplying together are all statistically independent from one another. That’s it.

Here is the foundational theory being abused by these Mopologists. Again, this part is not Bayes, it’s probability multiplication.

https://www.khanacademy.org/math/ap-sta ... ation-rule

When two events are independent, we can say that

P(A and B)=P(A)⋅P(B)

Be careful! This formula only applies to independent events.

To win the prize, submissions must satisfy the following conditions:
1. Provide a proof, with data, that each probability is statistically independent (or if you prefer, uncorrelated or mathematically orthogonal) to each of the other probabilities.
2. Submit these proofs in writing.
3. Pass review by a current BYU statistics or stochastics professor of my choosing.

As of this morning, Dan Peterson informed me that the challenge has been duly relayed! Good luck, Team Bayes!

[Everybody Wang Chung]

To win the prize, submissions must satisfy the following conditions:
1. Provide a proof, with data, that each probability is statistically independent (or if you prefer, uncorrelated or mathematically orthogonal) to each of the other probabilities.
2. Submit these proofs in writing.
3. Pass review by a current BYU statistics or stochastics professor of my choosing.

As of this morning, Dan Peterson informed me that the challenge has been duly relayed! Good luck, Team Bayes!

Brilliant.

[Gadianton]

Thank you Dr. Moore for your thrifty use of department funds. I admonish all faculty to go and do likewise as is sensible within your own stewardships.

I predict that Dr. Moore shall do for the left half of what DCP believes in, what James Randi did for the right half of DCP's beliefs.

[Doctor Scratch]

I predict that Gee and the Dales will turn down the offer, while Rasmussen will accept it. It would be a lot more meaningful if DCP were to put something up in return, but I doubt that he'll do it.

[Dr Moore]

I am sadly confident this challenge will go unclaimed.

One reason is painfully obvious once the endeavor begins.

What are the odds that a person with any combination of curiosity, memory, creativity, charisma, or savant-like abilities is born? (extremely rare)

Knowing such a person exists, with proven particular abilities, then the conditional probability of that person doing X, Y, and Z things that normal people may consider (1) highly unlikely and (2) totally uncorrelated from one another can, and does, equal 1.0.

These contributions by Team Bayes all must argue (and demonstrate) not only that things like Early Modern English, Hebraisms, book length, city-state politics, and ancient proper names occur independently of one another in literature, they also have to demonstrate that those things occur independently from each other in the mind of Joseph Smith.

Why? Because they all purport to multiply large numbers of conditional Bayesian probabilities: X was discovered and Joseph Smith wrote Y, which corresponds with X. The common correlated fact, and the common causal factor, is Joseph.

In short, they have to prove that Joseph Smith cannot have been sufficiently a savant, curious, creative, observant or charismatic enough for all of those things to enter and process together in his mind.

It is no less difficult than proving the non-existence of Kim Peek or Daniel Tammet or Magnus Carlsen.

[Philo Sofee]

Ah, Cassius University sponsoring an event of singular magnitude for the world to witness. Thank you good Dr. Moore for the excellence of the forwarding of this stellar event. Remember we all need to dress in our best University duds when attending (tux's not necessary however)... I shall bring the drinks to celebrate the victory, we shall all meet in the large 3 story study wherein all our secret research materials are housed, i.e., the faculty lounge. The swimming pool shall be cleaned before hand so a good time will be had by all, and yes, I have also made sure the hot tubs are in working order. The electric roof shall be opened up so the sun shines on us all.

[Dr Moore]

So far it appears that none of Team Bayes is up for the task. Maybe there are efforts in quiet, and after all the task is not insignificant.

I want to raise an important point about data rigor when proposing statistically significant analyses, such as put forth by Team Bayes.

They’ve all failed to offer even the slightest attempt at establishing control sets for any of their point probabilities. That means we have unquantified small sample size problems for each unique probability put forth by Team Bayes, to say nothing of their collective failures at showing statistical independence prior to multiplying their many made-up probabilities.

Small sample sizes in specialty Mopologetic qualitative “evidences” such as Early Modern English, Hebraisms, political models, etc etc, presents devastating compounded error because small sample size means low/no statistical significance, high rate of false positives, and inflated belief in estimation value. The compounding of statistically predictable errors makes all outcome statements from these analyses of zero mathemtical significance. Zero.

There is simply no statistical rigor in any of these Team Bayes submissions to the growing corpus of Mopologetic pornography.

[Philo Sofee]

So far it appears that none of Team Bayes is up for the task. Maybe there are efforts in quiet, and after all the task is not insignificant.

I want to raise an important point about data rigor when proposing statistically significant analyses, such as put forth by Team Bayes.

They’ve all failed to offer even the slightest attempt at establishing control sets for any of their point probabilities. That means we have unquantified small sample size problems for each unique probability put forth by Team Bayes, to say nothing of their collective failures at showing statistical independence prior to multiplying their many made-up probabilities.

Small sample sizes in specialty Mopologetic qualitative “evidences” such as Early Modern English, Hebraisms, political models, etc etc, presents devastating compounded error because small sample size means low/no statistical significance, high rate of false positives, and inflated belief in estimation value. The compounding of statistically predictable errors makes all outcome statements from these analyses of zero mathemtical significance. Zero.

There is simply no statistical rigor in any of these Team Bayes submissions to the growing corpus of Mopologetic pornography.

An excellent observation, and which, if I may, just add a slight judicial attempt at giving a bird's eye view of your observation. It is as if they are imagining that the hole they have to fill (I mean, come on a beginning prior at 0.0000000000000000000000000000000000000001) which is deeper than the earth is wide at diameter can be filled with what, oh say, 3-4 dozen singular grains of sand sized evidences? Even 10,000,000,000,000 full dump trucks won't even begin to budge the size of their prior, and they haven't yet grasped that.

[Everybody Wang Chung]

When asked if he would accept the 30k Challenge, KR recently responded:

As soon as he can explain to me how it would be possible to statistically demonstrate the orthogonality of things like DNA evidence and the presence of Chiasmus in the text, I'll get right on that.

It's a red herring, and I suspect he's very much aware of it. There are cases where independence can and should be demonstrated quantitatively. Such is generally not the case for the historical Bayesian approach adopted by Carrier and others. In the majority of cases, since it would make no sense to try to calculate correlations on the basis of data (what's the process for calculating the correlation between the probability of me contracting COVID and the probability of Dr. Moore having a had a haircut on the last week?), we can instead make informed and careful assumptions about whether two things are independent. This sort of thing is done in statistics quite often, and is generally uncontroversial if done in a responsible and transparent way.

You and your good friends are free to read the FAQ and the essays and decide for yourselves if what I'm doing is reasonable or not. I have little desire to jump through hoops that don't actually exist.

https://www.patheos.com/blogs/danpeters ... qus_thread

[Dr Moore]

It’s a cop out. He wants to use the powerful tools of probability multiplication and Bayes, without following the rules.