Sure. But if I handle snakes to prove they won’t bite me because God is real, and they don’t bite me—you do the math.
More seriously, though: the sentiment expressed in the quote is flawed, IMHO. Evidence isn’t always symmetrical. Any particular transitional fossil is reasonable evidence for evolution; not finding a particular transitional fossil isn’t strong evidence against it. A person perjuring themselves once is strong evidence against their honesty; a person once declining to perjure themselves is not strong evidence in favour of their honesty; et cetera.
I think this might have something to do with the prior, actually: The stronger your prior probability, the less evidence it should take to drastically reduce it.
Edit: Nope, that last conclusion is wrong. Never mind.
Right. Sensitivity does not equal specificity. Maher makes the mistake of assuming the rate of false positives and false negatives for the ‘snakebite test for god’ are equal. The transitional fossil test for evolution and the perjury test for honesty both have high false negative rates and low false positive rates.
Hm, I thought that reasoning argued against your own non-serious first paragraph rather than what Bill said. If the idea is “if God is real (and won’t let snakes bite me), then they won’t bite me”, then being bitten shows that the first part is false, but not being bitten doesn’t say anything about the first part being true or false.
Or if you don’t want to get hung up on formal logic, then it’s valid but very weak evidence, like a hypothesis not being falsified in a test.
What Bill Maher said was that if a person claims that ~Bite is significant evidence for God, they must admit that Bite is significant evidence for ~God. I’m saying I don’t think that’s accurate.
The sentiment that one should update on the evidence is obviously great, but I think we should keep an eye on the maths.
Sure. But if I handle snakes to prove they won’t bite me because God is real, and they don’t bite me—you do the math.
More seriously, though: the sentiment expressed in the quote is flawed, IMHO. Evidence isn’t always symmetrical. Any particular transitional fossil is reasonable evidence for evolution; not finding a particular transitional fossil isn’t strong evidence against it. A person perjuring themselves once is strong evidence against their honesty; a person once declining to perjure themselves is not strong evidence in favour of their honesty; et cetera.
I think this might have something to do with the prior, actually: The stronger your prior probability, the less evidence it should take to drastically reduce it.
Edit: Nope, that last conclusion is wrong. Never mind.
Right. Sensitivity does not equal specificity. Maher makes the mistake of assuming the rate of false positives and false negatives for the ‘snakebite test for god’ are equal. The transitional fossil test for evolution and the perjury test for honesty both have high false negative rates and low false positive rates.
Hm, I thought that reasoning argued against your own non-serious first paragraph rather than what Bill said. If the idea is “if God is real (and won’t let snakes bite me), then they won’t bite me”, then being bitten shows that the first part is false, but not being bitten doesn’t say anything about the first part being true or false.
Or if you don’t want to get hung up on formal logic, then it’s valid but very weak evidence, like a hypothesis not being falsified in a test.
What Bill Maher said was that if a person claims that ~Bite is significant evidence for God, they must admit that Bite is significant evidence for ~God. I’m saying I don’t think that’s accurate.
The sentiment that one should update on the evidence is obviously great, but I think we should keep an eye on the maths.
Fair enough, if the premise is that ¬Bite → God exists.
It’s more clearly apparent when you