IEM Success Due to Collective Intelligence
October 4, 2004 | Sydney Morning HeraldIf you want to know who's likely to win next Saturday's federal election, don't consult the polls or ask an expert - ask a bookie. Of all the ways of predicting election results, the odds bookies quote on election eve are the most accurate. It may sound like superstition or just casual observation but there's actually a fair bit of science behind it, as research by economists and psychologists is demonstrating. What it's saying is that there's such a thing as "collective intelligence" whereby the average guess of the group is more reliable than the opinion of some great expert. The practical exploitation of the idea that punters' betting carried valuable information was pioneered by academics at the UNIVERSITY OF IOWA's business school. Their now well-known (but still small) Iowa Electronic Markets are online, real-money futures markets allowing you to make the equivalent of a bet on various future events, including elections. The IEM, too, has a most impressive forecasting record, though it erred in predicting that George Dubya would win the popular vote as well as the presidency in 2000. The market outperformed the polls in nine out of 15 cases.
Contact: Robert Forsythe