Justin Wolfers is writing a series of columns on prediction markets for WSJ Online. Here’s his first.
I agree with Andy Gelman’s comment : http://www.stat.columbia.edu/~cook/movabletype/archives/2007/12/are_political_m.html
We should not compare polls to prediction markets since polls are not predictive but give a useful information to make predictions. How should we interpret a comparison between predicttive markets and polls if we assume that better use polls in their own forecast ?
Comments are closed.