News for hackers. Stuff that matters.

A comment on Nassim Talebs case against Nate Silver is bad math. Parent comment

Re:Nassim Talebs case against Nate Silver is bad math by et2o on Tuesday April 09, 2019 @05:15PM (#19618997)
Curious about this.

> It's how you can prove all non-Bayesian models are irrational

Can you expand or link?

> Error bars and spreads are already outside the Bayesian paradigm, which demands a single number

Not following at all. What do you mean?

[ Parent | on HN ]
Re:Nassim Talebs case against Nate Silver is bad math by ikeboy on Tuesday April 09, 2019 @05:21PM (#19619049)
https://plato.stanford.edu/entries/epistemology-bayesian/sup...

>Not following at all. What do you mean?

One assumption is that given a bet, you must choose to take a side of it. If you won't bet for something at X%, you must bet against it at 1-X%.

[ Parent | on HN ]
Re:Nassim Talebs case against Nate Silver is bad math by et2o on Tuesday April 09, 2019 @05:31PM (#19619141)
Ah, interesting link. I think I ran into some confusion since 'rational' and 'Bayesian' are somewhat overloaded terms.

When I think of Bayesian analysis, I always think of Bayesian models which attempt to model uncertainty and always have a posterior distribution (spread) of outcomes.

For example, in the Dutch Book page, instead of having a point-prior (P=0.51 (delta on P=0.51?)), each degree of belief could have some (different) distribution which could yield a posterior with a distribution and change the math here.

[ Parent | on HN ]
Re:Nassim Talebs case against Nate Silver is bad math by ikeboy on Tuesday April 09, 2019 @05:33PM (#19619158)
Yeah well there's a difference if you're estimating a parameter or predicting true/false on an event.
[ Parent | on HN ]
Re:Nassim Talebs case against Nate Silver is bad math by et2o on Tuesday April 09, 2019 @05:39PM (#19619219)
Not sure about that, think about the case of (Bayesian) logistic regression.
[ Parent | on HN ]