To answer your question, I would feel better with multiple forms of evidence as a collective instead of a single piece.
EDIT: Also I'm not saying that this guy isn't cheating. I'm saying it is possible that 1000 coin flips could come up heads in a row and doing so isn't proof of cheating. Unlikely, but possible.
> I'm saying it is possible that 1000 coin flips could come up heads in a row and doing so isn't proof of cheating. Unlikely, but possible.
FWIW, I think the '1000 coin flips' thing was an exaggerated analogy used to illustrate the point; I doubt the evidence here is anywhere near that strong, even if it is extremely strong.
But the problem that people are trying to point out is that you seem to be asking for a literally impossible standard of proof (think about it: for any evidence you could possibly gather, there's a non-zero chance that it has all been fabricated, or you're consistently misreading the data, or your memory is playing tricks, or...).
If your point is not that 'proof requires literal 100% certainty', what is it? It seems that you might be bothered by the idea of convicting someone based on certain types of evidence? (Maybe you're intuiting that when someone claims statistical evidence at the '1000 coin flips' level, the actual chance of a false positive is much higher than the claimed amount, because of the possibility of human error or mendacity?)
> The only explanation for observing 1000 coin flips coming up as heads is that something other than random chance made that happen.
> Naturally, there could be several explanations - the overwhelmingly most likely is that he was cheating.
But you can't have it both ways, You can't say their could be several explanations and something other than random chance made that happen for the same thing.
I'm not trying to say its likely. Just thats its possible and because its possible that can not be the sole basis to prove guilt. You say "overwhelmingly most likely" and yes I agree its very unlikey but its still possible and because it is possible we shouldn't use that as the sole basis to prove guilt.
Use it as a pointer to start digging into them sure. find other things to build your case. But don't use that sole thing as the basis of guilt.
All I've been saying all this time is that its possible. And because it is possible that I personally won't "throw someone under the bus" because its improbable and not impossible.
You're still ignoring the point everyone is trying to make: whatever set of evidence you have, however strongly it points to someone's guilt, there are always possible alternative explanations. 100% certainty is a requirement that will never, ever be met.
FWIW, I think the '1000 coin flips' thing was an exaggerated analogy used to illustrate the point; I doubt the evidence here is anywhere near that strong, even if it is extremely strong.
But the problem that people are trying to point out is that you seem to be asking for a literally impossible standard of proof (think about it: for any evidence you could possibly gather, there's a non-zero chance that it has all been fabricated, or you're consistently misreading the data, or your memory is playing tricks, or...).
If your point is not that 'proof requires literal 100% certainty', what is it? It seems that you might be bothered by the idea of convicting someone based on certain types of evidence? (Maybe you're intuiting that when someone claims statistical evidence at the '1000 coin flips' level, the actual chance of a false positive is much higher than the claimed amount, because of the possibility of human error or mendacity?)