Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Because when you mix a lot of random variables together, the variance tends to decrease.


Exactly, and the variance of the maximum will decrease even faster than the variance of the mean.


This is very interesting, can you point me/us to any explanations about it?


I'd guess at the http://en.wikipedia.org/wiki/Central_limit_theorem - Combined distributions tend towards Gaussian (bell-curve)




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: