http://www.telegraph.co.uk/earth/main.jhtml?xml=/earth/2008/08/18/scibetting118.xml Telegraph.co.uk Predicting the future - with the power of betting By Paul Parsons Last Updated: 3:01pm BST 18/08/2008 If you believe the polls, the race for the White House between Barack Obama and John McCain is neck and neck. But if you believe the scientists, Obama is ahead by a whopping 25 per cent. And not only that - without any specialist knowledge, the same techniques can be used to tell you how many medals Britain will win in Beijing, or even what the weather in Florida is going to be like. The idea is simple: to ask people to place their bets. In the case of the US elections, the experiment is called the Iowa Electronic Market, run by the University of Iowa. It works just like a financial market: investors are invited to buy virtual "shares" in the two major political parties, each priced between zero and $1. After the election, traders holding stock in the winning party receive $1 per share. Those who backed the loser receive nothing. During the campaign, traders can buy and sell their shares, with market forces driving the prices to reflect each party's chances of being elected - the current trading value of the Democrats, $0.62, corresponds to a 62 per cent chance that they will win. Unlike an opinion poll, investors in a market must choose who they think is going to win, not who they necessarily intend to vote for, and must vouch for that candidate with hard cash. Since the IEM was founded in 1988, this has proved vastly superior to traditional opinion polls. "Comparing markets and polls across entire elections," says Professor Forrest Nelson, one of the researchers, "the market is closer to the actual outcome than the polls about 74 per cent of the time." advertisement The idea that betting can distil the disparate and sometimes contradictory information held by members of a group was first reported by English scientist Francis Galton over 100 years ago, in the journal Nature. In 1906, he visited a country fair at which passers-by were invited to bet on the weight of an ox. Each guess cost sixpence, and there were prizes offered for the closest. None of the 787 individual guesses - some made by experienced farmers and butchers - came close to the true weight of 1198 pounds. But their average, 1197 pounds, was almost dead on. In 1945, Austrian economist Friedrich Hayek applied Galton's finding more widely. He argued that price is an effective way to aggregate public opinion on all sorts of issues - from the value of commodities to who will be leading the country. Betting on elections became a popular pastime in the early 20th century. But it's not the only field where speculative markets are proving remarkably accurate. Companies such as Google, Microsoft and Hewlett-Packard use internal betting markets to guide their business decisions: HP's consistently out-predicts official forecasts of the company's sales. Betting is now used to predict the future of climate change, movie box-office takings and even US military action. In 2006, the Hollywood Stock Exchange - a market based on the box-office success of movies - correctly predicted 32 of 39 major Oscar nominations, and seven of the eight top winners. Meanwhile, the futures market for the price of orange juice has earned a reputation as a more reliable predictor of the weather in Florida, where 98 per cent of the US orange crop is grown, than the National Weather Service. And when the US Navy was unable to pinpoint the wreck of USS Scorpion - a nuclear submarine that sank in the Atlantic following an accident in 1968 - it asked salvage experts to bet on the location. The wreck was later found within 200 metres of their consensus view. James Surowiecki, author of The Wisdom of Crowds, which examines markets and other examples of group intelligence, thinks the key to an effective betting market is allowing traders to think and act independently. In traditional stock markets, this isn't always possible, as investors shout their deals within earshot of other traders. This can lead to a "bandwagon effect" - copycat trading that drives stock prices away from their true values. Psychologist Richard Wiseman, of the University of Hertfordshire, agrees. He draws a parallel with the "ask the audience" lifeline on quiz show Who Wants To Be a Millionaire? (which, incidentally, gets the right answer 91 per cent of the time - compared to just 65 per cent for "phone a friend"). "In Millionaire, it is vital that people can't see how everyone else is voting," says Wiseman. "If that happens, you very quickly get sheep-like behaviour rather than people's actual opinion." But markets have also attracted controversy. In 2003, the Pentagon's Defense Advanced Research Projects Agency tried to set up what it called the Policy Analysis Market - a betting market to trade on US defence issues. The idea was that the moment the share price of, say, an al-Qa'eda attack, began to climb, then security could be stepped up. This "terror futures" market, as it became known, was condemned by Washington. "The idea of a federal betting parlour on atrocities and terrorism is ridiculous and it's grotesque," said Senator Ron Wyden of Oregon. The idea was scrapped, and John Poindexter, the official responsible, resigned - though not before futures exchange Tradesports ran a market on his departure date. In July 2003, it predicted he would be gone by the end of August; he officially retired from the agency on August 12. Tradesports had itself been criticised for running a market in February the same year, on whether the US would attack Iraq. Professor Eric Zitzewitz, a markets expert at Dartmouth College, New Hampshire, and colleagues used the results to argue against going to war. "We were able to estimate before the war began that markets thought war was a bad idea," he says. "While that estimate didn't end up influencing policy, the potential for markets to inform decisions in the future seems like a significant offsetting benefit to consider." This idea of using markets to shape foreign policy may sound unorthodox. But Professor Robin Hanson, an economist at George Mason University in Virginia, has an even more radical suggestion: he's put forward an entire system of government based on the predictions of betting markets. He calls the idea a "futarchy": voters get to invest in key policy decisions, and only the policies that end up with the highest share of the market become law. Elected representatives would continue to define and manage national welfare, but their policy decisions would be made by market speculators. "In futarchy, democracy would continue to say what we want, but betting markets would now say how to get it," explains Professor Hanson. He believes solid policy-making would be guaranteed, as only voters with a sufficiently informed opinion would put their money where their mouth was. "Those who know they are not relevant experts shut up," he says. "And those who do not know this eventually lose their money and then shut up." Investors would bet on whether particular policies would raise or lower the country's overall wellbeing. Hanson suggests that a good measure of this, by which the bets could later be settled, might be a country's gross domestic product. "People in high-GDP nations tend to be richer and better off than those in low-GDP nations," he says. So only those policies predicted to maximise GDP would be implemented. "With the full attention of our elected representatives, we should be able to do even better, for example by including happiness, inequality, health, leisure, and environmental measures." It's a proposal that's been greeted with some scepticism. Professor Tyler Cowen, also of George Mason University, thinks that the problem of bad governance is far too complex to be solved simply by making predictions of how policy decisions may or may not turn out. "I don't agree with the futarchy idea," he says. "The record of prediction markets is a strong one, but I wouldn't want to use them to run an entire government." Perhaps it's best to start small - for example, with the Beijing Olympics. A website called Hubdub is running a series of markets at hubdub.com/olympics on everything from how many different countries will win Olympic medals to whether Britain will beat Australia in the medal table. The good news regarding the latter is that the weekend's gold rush has sent the odds up by 29 per cent - the bad news is that they're still only at 41 per cent. And you don't want to know what the punters at sports exchange Betfair are saying about the Ashes…