Well, prediction markets do better than polls, but actual financial markets are in part prediction markets and they do pretty badly at times and sometimes there's a guy who calls a bubble long before it pops. And that guy doesn't necessarily stake a huge fortune on that prediction. So yeah, talk is cheap, and yeah, on average people are more careful with their wallets than their mouths, but on the other hand, people sometimes have interesting things to say, and asking how much money they bet on being right is less interesting to me than asking how they arrived at their conclusions. Now in some cases I will ask someone how much money he'd put on being right, but I'll do that when I think they're either wrong or fairly uninformed and yet very opinionated.