In 1906, the British statistician and polymath Francis Galton attended a country fair at which the attendees were invited to estimate the weight of an ox. Out of curiosity, Galton borrowed the cards on which the guesses were written, took them home, and ran the numbers. To his surprise, the average of the 787 entries turned out to be almost exactly the precise weight of the animal. It was more accurate than even the individual estimates of butchers and farmers, who presumably had an eye for such things.
Galton published his findings in Nature, thus establishing the study of collective intelligence: the principle that groups of people can be smarter than even the smartest, most-expert individuals among them. Since then, research in multiple fields has consistently demonstrated that aggregating a variety of judgments frequently and sometimes dramatically increases accuracy, as individual errors cancel themselves out. Nearly 100 years after Galton attended the fair, a best-selling book popularized his idea as “the wisdom of crowds.” Crowdsourced judgments have been used to improve medical diagnoses, scientific research, and economic forecasts.
Of course, crowds are far from infallible. Wise probably wouldn’t be the word we’d use to describe the throng of amateur investors who recently inflated the price of GameStop shares. Stock-market bubbles contravene a crucial condition of a smart crowd: that there should be a diversity of judgments. When a group herds toward one answer, it effectively becomes like one very fallible individual, rather than a sum of different guesses made independently. The wisest crowds are composed of individuals who disagree with one another.