3 ms·
> but around 20 I'm going to be suspicious. There's a multiple-endpoints effect here. If there are 2^20 participants all doing that, one of them is going to f
by T-hawk 8y ago
> but around 20 I'm going to be suspicious.
There's a multiple-endpoints effect here. If there are 2^20 participants all doing that, one of them is going to flip 20 heads in a row. You only notice the one who did after the fact. Some participant has to be the most outlying result and that's the only one you notice.
(Not that that's necessarily the case for Buffett, but that's a general fallacy to be aware of.)
- dwighttk 8y agoYeah, I'd be less suspicious of 20 in a row out of a group of 1000000 flippers. But if any of them got 70% of 1000000 flips I'd be suspicious.