cousin_it comments on Simpson's Paradox - Less Wrong

68 Post author: bentarm 12 January 2011 11:01PM

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

Comments (58)

You are viewing a single comment's thread.

Comment author: cousin_it 13 January 2011 07:16:53AM *  2 points [-]

This post is very similar to a comment I made back in November. I love the concept and have been noticing it a lot. Other examples of "paradoxes" that may be interesting to LW: Braess's, Sen's.

Comment author: prase 13 January 2011 12:41:05PM 0 points [-]

Isn't Sen's paradox essentially the same as the Prisoners' dilemma?

Comment author: JJ10DMAN 05 February 2011 10:46:28AM *  1 point [-]

Thanks for the links cousin it! Great reads.

Re: prase's reply: The Prisoners' dilemma is a legitimate dilemma. No matter how many times I read the page on Sen's paradox I can't interpret it as anything remotely sensical.

I kept editing this post again and again as I boiled down the problem (it's harder to explain something wrong than something correct), and I think I've got it down to one sentence:

If you just look at sorted lists of preferences without any comparative weights of given preferences, you're going to get paradoxes. Nash equilibrium exists because of weights. Sen's paradox does not exist because it precludes weights. If Bob just wants Alice to see a movie and doesn't much care about his own viewing of the film either way, and Alice just wants to spend time with Bob and doesn't much care if it's by watching a movie or not, then there's no paradox until a social scientist walks in the room.