Predictive markets

Found this on Marty Kearns' fine Network-centric Advocacy blog, a reference to Yahoo's technology prediction market, Yahoo Tech Buzz. Prediction markets are based on the notion that crowd thinking in many cases is more accurate than "expert" thinking. James Surowiecki popularized this idea in his book The Wisdom of Crowds (see earlier posts on this blog). John Brunner also had something like this in his classic and remarkably prescient 1970s sci-fi novel Shockwave Rider.

The Yahoo market uses NewsFutures engine. See their "A Simple Example" page for how these markets work.

"Market" is accurate in the sense that people trade "shares" that have some either real money value (as in the case of the Iowa Electronic Markets political futures market) or play money value as is the case with the Yahoo Tech Buzz market. The notion is that the "market" only works if the participants have something at stake (presumably something scarce and desirable, like real money). In the case of play money markets, this might be reputation or desire to win or "to be right" or "not be wrong". I wonder if the structure of "market" is necessary for such a mechanism to work -- could there be a predictive commons? Surowiecki argues that for crowds to be "smart", they need to be diverse and the members independent, otherwise you get herd behavior. The market by definition assumes the separation between participants on the basis of conflicting (self) interest. But the positing of a "self" is a philosophical assertion, and "self-interest" a political position. Perhaps some sort of collective-interest expressed through the individual, the struggle of internal contradiction as opposed to "self-interest". Hmmm.

Research indicates that these markets can be more accurate predictors than a room of experts. Marty's blog calls for one for environmental issues. An open source version anyone.

jd