Compounding Knowledge

Re:Compounding Knowledge by dwighttk on Wednesday February 06, 2019 @08:07AM (#19094769)

Not over the time period he has been successful. Or at least the longer he lasts as a successful person doing what he does the less likely he's just acting at random. It's possible to get 1000000 heads in a row flipping a fair coin, but around 20 I'm going to be suspicious.

edit: With Buffet, he's not actually getting everything right, so I guess a better metaphor would be like 70% heads over 1000000 flips... more likely under random circumstances, but still pretty suspicious.

[ Parent | on HN ]

Re:Compounding Knowledge by T-hawk on Wednesday February 06, 2019 @03:33PM (#19099578)

> 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.)

[ Parent | on HN ]

Re:Compounding Knowledge by dwighttk on Wednesday February 06, 2019 @07:26PM (#19101749)

Yeah, 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.

[ Parent | on HN ]