IlyaShpitser comments on Three more ways identity can be a curse - Less Wrong

40 Post author: gothgirl420666 28 April 2013 02:53AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (104)

You are viewing a single comment's thread. Show more comments above.

Comment author: maia 28 April 2013 02:46:01PM 4 points [-]

That is much less likely than the other way around, though. That's the point of regression to the mean.

But this is still evidence that his future posts will be at least somewhat above average! If not quite as good as this one.

Comment author: IlyaShpitser 28 April 2013 03:19:23PM *  4 points [-]

That is much less likely than the other way around, though.

What the posterior over the expected post quality of the OP is, and what one should say in public about a stranger are two separate issues. Do you understand that if I start following your posts and replying to good ones effectively saying "boy, this was a good post, but it's probably better than your average", then that behavior would be extremely patronizing?

Bayes theorem is just a one liner consequence of the definition of conditional probability. Having social graces is actually hard.. and yet "being Bayesian" seems to be a bigger deal than being a good sport.

Comment author: Mestroyer 28 April 2013 05:26:13PM 5 points [-]

Now that this has turned into a discussion about statistics, it's more important than politeness.

There are times when complaining about a social rule being broken is a worse infraction than breaking it in the first place. Your injunction to avoid pointing out flaws in someone's reasoning because it's impolite has left a far worse taste in my mouth than any insult by prior probability of the form: "you're probably about average quality for a poster, because one post isn't enough to prove otherwise."