Put in epistemic terms, before I ask the person what time it is, let's say my degree of belief that it's 4:15 is 75%. Then I ask the person what time it is and they tell me it's 4:25. My belief that it's 4:15 now drops to like 33%. The anecdotal evidence alone, absent any other evidence (the person is a stranger after all) is enough to disconfirm my original belief by quite a lot. Anything that confirms or disconfirms a belief is evidence.
And if they tell you that it's 4:65, you wouldn't believe them for an instant.
You would automatically assume that? Even rolling four in a row is pretty remarkable. I would certainly consider the possibility they were remembering it correctly and weren't lying. In the case of the nurse reporting the veridical account, too many people here won't even consider the possibility she may be correct. I don't consider that very skeptical. Skepticism cuts both ways: we don't automatically believe people and we don't automatically disbelieve them, either. Too many people forget the latter.
We don't automatically disbelieve them. We apply Bayesian analysis. We evaluate new evidence in light of existing evidence.
In this case, we have an anecdote of an anecdote. We have two layers of confirmation bias, misreporting and confabulation, no corroboration, no confirmable details.
Prior probability: Almost zero.
Value of evidence: Almost zero.
Adjusted probability: Almost zero.
If there were a 5% or 10% chance her report was accurate, we'd be
buried in hard evidence of the supernatural nature of NDEs.
Instead, we have none whatsoever.
It's difficult to evaluate exactly how unlikely it is that any of this nonsense is true. Certainly events with only a 1 in 10
9 chance happen all the time, and are perfectly ordinary and verifiable, so it's far, far less than that. Indeed, we regularly measure and assess events that happen one time in 10
17, and these are
still common enough that we record them actually happening, and they can present real engineering problems on an individual basis, never mind for the entire planet (another factor of about 10
10).
Based on that, I'd put all supernatural claims together being somewhere below 1 in 10
25 of
any being true, and any one claim below 1 in 10
30.
So, not 5% or 10%, but I'd be willing to grant you perhaps 0.0000000000000000000000000001%.