8) To construct a probability distribution on outcomes, you have to *know the space of outcomes*. You can have guesses about the space...
@CoreXXIII Vast oversimplification: *If you can reduce the problem to a calculation*, use Bayes. But that first step is almost everything.
-
-
@St_Rev Even there, Bayes is often useless in statistics. Different priors always produce different results unless there is a wealth of data -
@St_Rev An when there is a wealth of data, the data is often so inconsistently produced that statistical comparison is useless in general. -
@CoreXXIII Yeah. LWers like to invoke...agh, forget the name, it's the theorem about rational agents w/access to same data agreeing -
@CoreXXIII Ignoring the fact that it's essentially only relevant to gods and oracles. -
@St_Rev I dunno what that is but I'm gonna name it the economists' fallacy. -
@CoreXXIII Aumann's agreement theorem. -
@St_Rev Ah, this is different than i was thinking. it reminds of the fights I've seen over Newcomb's paradox. -
@CoreXXIII Newcomb's only a paradox because its premise is inconsistent. Can't be arsed to pick it apart at the moment though.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
. Banned in Sweden. SubGenius, Zhuangist, white-hat troll. Defrocked mathematician. Brain problems.