NOVA | scienceNOW | One-Minute Experiment

In 1906, 800 people attending a fair outside Plymouth, England took part in a competition to guess the weight of an ox. The crowd included butchers and farmers as well as many casual participants who knew little about cattle. Organizers offered prizes for the best estimates, but each contestant had to pay a small fee to participate—two factors that encouraged the group to take their guessing seriously. Most people misjudged the ox's weight by at least 30 pounds, but the crowd as a whole proved remarkably accurate: Its averaged guess came to 1,197 pounds, just one pound less than the animal's actual weight.

It was no fluke. This example and others like it illustrate what author James Surowiecki calls the "wisdom of crowds," in which a group's averaged prediction is almost always more accurate than one individual's guess, even if the members of the group haven't discussed their answers with one another. But how does "group intelligence" come into play if the group members aren't allowed to work together on a problem? In general, each crowd member has unique life experiences and background knowledge that can serve as context in making predictions. Because together the group has more information to work with, it can make better choices.—Rima Chaddha

One-Minute Experiment

Here's an experiment you can try for yourself. You'll need a pen or pencil, a notebook, a calculator, a stopwatch—and a small group of people:

Label a page in your notebook "Time Guessed." Then number each line until you get to 30.
Begin your experiment somewhere with lots of people—school, work, or family gatherings, say.
Start asking people if they want to take part in your experiment. For those who agree, tell them to guess when they think one minute has passed. Start your stopwatch as they begin to count the seconds in their mind and then stop it when they say "Done."
As you collect people's guesses, write their estimates in seconds on each line of the page.
After you've gathered guesses from 20 people, add them up and divide by 20 to get the average of their guesses. The average guess will show you how close to the one-minute mark the group was as a whole.
Try asking an additional 10 people and repeating step 5. Since you now have 30 guesses, divide your answer by 30 instead of 20. As more people guess, the average guess will probably get even closer to 60 seconds.

How did your experiment go? Let us know on our NOVA scienceNOW discussion board. Or, if you'd rather take part in a "wisdom-of-crowds" experiment yourself, try Counting Cabs.