Thanks for the lesson, Jay. It has me thinking about a structured problem-solving method that included educated guesses my old company used for product development (OT here).
Examples of proper use of the kind of reasoning Jabba proposes to use are most certainly not off-topic. And I use statistical inference all the time in my own job.
Now, what are the chances Jabba will a) read your post and if he does, b) learn from it?
This isn't the first time I've taken pains to try to teach Jabba and it won't be the first (or last) time that he's ignored it completely. I predict he'll do the same this time -- just carry on as if nothing happened.
He's done the same to all the experts he's consulted, even the ones he chose himself. Here he's trying to play the notion that the only people who object to his method are the people in this forum, whom he can safely ignore since they're "just" skeptics. Remember that his overall strategy seems to be to prove that skeptics are entrenched troglodytes, unable and unwilling to expand their minds. But he fails to realize he's wandering into a field he admits he doesn't understand -- and fails to remember that experts outside this forum have told him his approach is wrong. Much of Jabba's argument here and in the Shroud thread revolved around asserting that the alleged stubbornness and dishonesty of his critics were what kept him from proving his point, not his own shortcomings.
Narrowly-tailored requests for production are part of the fringe playbook. "Find, in an arbitrarily limited field of sources, an authoritative statement that directly and concisely says exactly what you mean in so many words." I'm sure if I scoured the library I might find such a statement, and I'm sure Jabba would rather have me do that than write posts here. But the fact that underconstrained models do not produce useful inference is a conclusion one draws from long and careful study of degrees of freedom, and from practical experience. It's the sort of thing that's a product of expertise and what separates expertise from mere Google-fu. The sort of statement, "You can't guess at both the priors and the likelihoods" is generally made only when someone wrongly tries to do it. It's the application of expertise to a specific problem. Textbooks aren't filled with lists of "You can't do that" statements. Knowing that you can't do something is what comes from having read and comprehended the whole book.
So while some people will read and understand my post, Jabba will most likely just pretend it never happened. It's a reminder to him that he's trying to do stuff he clearly doesn't know how to do, and a reminder that he can't write off his critics as ideologically compromised pundits. So if pressed, I predict he'll just say he has to take some time to study it (and we'll never hear any more about it). Or he'll find some trivial reason to not have to address it, such as that it doesn't fit his narrowly-minted criteria or that I'm "unfriendly" or some other such nonsense.
But he'll be back.
