Fine, prediction markets are gambling. They are better gambling propositions than traditional oddsmaker books which have higher margins/house take baked into them.

Big political outcomes is a helpful prediction market service, even/because insiders can cheat by betting on the outcome. Midterm outcomes, or (4% chance) of $5000 dividend cheque by March 31 2007 are specific probabilities that reveal pure BS.

Some horrible contracts are whether a speech will include some word, or outcomes that have already been predetermined (Oscars) with insiders who know 100% betting vs the clueless. Other terrible contracts include US-Iran ceasefire status. Too ambiguous, and resolution process is prone to corruption.

Still, these markets are more useful predictors than polls, because experts are analyzing with money instead of bias.