For example betting markets may be better than polling for predicting elections even though betting markets probably have a more biased population participating.