So if someone wins the lottery, we can conclude the game must be rigged, because it's more likely for civilization to collapse than for someone to win? Is that what you're saying?
.gif "What was that? :wwt")
I think we've been through this before, Dave. But tell me again. At what level of improbability (P-value) would you reject a hypothesis? Barring an impossible observation, there is a non-zero chance that the hypothesis could still be true at any P-value.
As to the fatal flaw in your analogy:
The purpose of probability is to deal with incomplete knowledge.
Where is the incomplete knowledge in your analogy? Our knowledge of the lottery is complete. There is no need to test any hypothesis about how someone won the lottery. We know how someone won the lottery.
You, however, can offer no explanation why that particular brain is mysteriously assigned to be 'you'. It just is. A brain tautologically happens to have lit up your jungle. That's all you know about it. It's not even the identical brain that originally lit up your jungle shortly after your birth.
The unique brain assumption is just that - an assumption. So why shouldn't the unique brain assumption be tested, and if it flunks the test, why shouldn't the corollary be favored?
Should you choose to test the offending assumption, why shouldn't you test it from the viewpoint that provides the information to test it, instead of flapping around up in the sky chirping that you don't see any reason to question the assumption from up there, when in fact what you don't see is anything at all. Because you're in the wrong place.
The corollary to the unique brain assumption is no less counterintuitive, doesn't flunk the observational test, and therefore should be favored.
