I do think that the significance of my unlikelihood is the weakest link in my argument, but I do still think that it's totally significant.
It really can't be both. You're trumping up the semblance of significance out of nothing more than an assertion that the outcome -- whatever it was -- would have been highly improbable in retrospect. That's the essence of the Texas sharpshooter fallacy.
Also, the most statistically savvy member of our thread (Caveman) says that the sharpshooter fallacy doesn't apply.
No, desperately glomming onto whomever expresses disagreement with your critics is not a defense. You have, on more than one occasion, said that you do not understand what he says, hence your characterization of him is uinformed. And you have passed on numerous requests that you demonstrate
your knowledge of the Texas sharpshooter fallacy. You obviously don't understand it, so you grasp at straws.
Whatever, if you don't mind, I'd like to postpone further discussion of the sharpshooter issue until I know with what specifically else you're finding fault.
This is what you do. You propose to focus on one subissue until it's resolved, posturing this as a systematic method of examining a claim. Except that here it
is resolved, and the resolution is that your proof fails on the point of
post hoc rationalization. So what you really propose is to focus on one subissue until refutation is imminent, then frantically change horses to avoid that disaster. You shift abruptly between individual points, pretending that each abandoned point is no longer operative once you leave it behind in favor of your new subissue.
This is why I proposed a breadth-first examination. You can't play your shell game if you're forced to confront the entirety of the errors in your proof all at once. And you know full well what
they are, so stop asking. You also know full well -- having
admitted to it -- that you cannot sustain the debate at its most fundamental level and must instead rely on foisted ground rules that forestall any meaningful challenge to your proof. In any universe, that means you lose the debate.
For one thing, do you accept that Bayesian statistics does apply to re-evaluating OOFLam?
I do not agree that you're using Bayesian statistics correctly to effect your proof, nor that it would be the appropriate method given the data you have at hand. I've told you specifically why. At this point, asking for the objections to be repeated is just rude.
Further, no one accepts that your proof is valid. You have shown it to quite a number of statisticians, including in person to experts you yourself have selected. They have all told you that you're wrong. Instead of taking their corrections to heart, you abandon them and carry on your merry way. The evidence shows you have absolutely no interest in whether your proof actually has any statistical validity. You have shown interest only in whether you can fool people into thinking it works.