3 ms·
> That is how it works: it's regression to the mean. If you expect a major hurricane every n years and you didn't have one this year, the expected likelihood of
by lotsofpulp 3y ago
> That is how it works: it's regression to the mean. If you expect a major hurricane every n years and you didn't have one this year, the expected likelihood of one next year (1/n) is greater than what you actually got this year (0).
Not if hurricanes are independently and identically distributed. Which I think you acknowledge could be the case in your last paragraph, but I have not ever come across anything linking hurricane to one another.
https://en.wikipedia.org/wiki/Gambler%27s_fallacy https://en.wikipedia.org/wiki/Gambler%27s_fallacy
- adhesive_wombat 3y agoNo, it's not a gambler's fallacy. You know you got zero this year, it already happened. You know it's expected 1/n next year (assuming independence, etc), just like every year. So the expected amount of hurricanes next year is more than what you did get this year, which was 0. In total you now (as of now, year 1 end) expect only 0 + 1/n total hurricanes over the two years (this year and next). That is 1/2n per year: less that you would have predicted at the start of this year, which would have been 2/n over two years for 1/n per year. A gambler's fallacy is thinking getting 0 this year means it's more likely than 1/n next year to act to retroactively "correct" 0 this year.