Should-Read: Noah Smith: The Wisdom and Madness of Crowds

Should-Read: Noah Smith: The Wisdom and Madness of Crowds: “Prelec… Seung, and… McCoy ask… what people think others will guess. If herd behavior is present, some people will know it, and will be contrarians…

…they’ll guess something different from what they think other people will say. Prelec et al. find that the forecasts that receive the most contrarian support–the guesses that people pick even though they think others will guess differently–tend to be the right ones. They find that these forecasts, which they label the “surprisingly popular” options, tend to outperform standard crowd averages in a number of applications, with error rates more than 20 percent lower…. This method obviously has some very important potential applications for finance…. In the search for the ultimate forecast, the wisdom of crowds might turn out to be very good, but not quite the best.