3 ms·
I'd assume that even if in every case the number of incorrect guesses is finite, the expected number of people that fail to guess their color is infinite. Am I
by im3w1l 10y ago
I'd assume that even if in every case the number of incorrect guesses is finite, the expected number of people that fail to guess their color is infinite.
Am I right about this?
- baddox 10y agoIm not sure what you mean by "expected number." Do you mean if you try to guess how many inmates the warden has managed to guarantee will fail? Per the article, the warden can guarantee that an "arbitrarily large finite number of them" will fail. But it's still always finite despite being unbounded. If you want to predict a lower bound on the number of failures the warden has guaranteed, you just have to guess a larger natural number than the warden. :)
- im3w1l 10y agoI mean if the warden picks a sequence randomly.
- daxfohl 10y agoInteresting observation! Yes, I think reduced to its essence, it boils down to: the expected value of "a finite number" is infinity. Which is strange in itself. So you end up with a sequence that converges to 100%, but that convergence never starts, but it certainly happens eventually. I guess a similar, less verbose thing would be "pick a random rational in (0, 1)". In decimal representation it'll repeat after "a finite number" of random digits, but "a finite number" again is expected to be infinity. So is a random rational really rational? Someone more educated on set theory will have to comment.
- waqf 10y agoIf you take any finite subset of N prisoners, it's easy to see that they won't do better than chance, so the expectation is for N/2 of them to guess wrongly. Since N can be arbitrarily large, your intuition is correct.