4 ms·
>For example, lets say that everyone encounters an opportunity to become a billionaire at a perfectly constant rate of 1 opportunity/year [0] and the probabilit
by Dibby053 2y ago
>For example, lets say that everyone encounters an opportunity to become a billionaire at a perfectly constant rate of 1 opportunity/year [0] and the probability we can capitalise on it with our skills is normally distributed (ditto for ruinous events)
I think your assumption is wrong. The key idea is that the outcome of random lucky events depends on the amount of capital one already has. A lucky event doesn't add to capital, but multiplies it, so consecutive lucky events accumulate creating a long tail.
If you believed in Bitcoin in 2012 you probably wouldn't be rich today, if you were living paycheck to paycheck at the time. If by 2012 you already had some spare capital from getting lucky in the dotcom years, on the other hand...
- roenxi 2y agoI'm using a different type of lucky event to the study, but my argument is a statistical one so that difference doesn't worry me. You can substitute different types of events in if you prefer.