In the consumer world, we have learned to exploit the collective wisdom of a community as part of our on-line buying experience. Can this "wisdom of crowds" apply to business communities, asks M Institute co-founder Jyoti Banerjee?
Tomorrow we will go to the Christmas fair at my daughter's school and attempt to guess how many jelly-beans are in a glass bowl. It's what you do at school fairs. What is certain is that no one will guess the right number. What is also likely is that almost all the answers given by the individuals entering the contest will be far off the mark. Yet, if it were possible to average the answers across all participants, the resulting mean will be about as good an answer as we can expect all day. One or two of the girls may improve on the mean score but they will represent a tiny minority.
How can I be so confident of this result? Just over a century ago , Francis Galton, polymath inventor of the idea of a statistical mean, found that the sort of aggregation used by measures of central tendency (such as the mean, median, etc) results in the removal of errors that individual datapoints could suffer from. At a country fair near Plymouth, he found a large group of people guessing the dressed weight of a slaughtered bull. The average across 787 entries turned out to be 1,197 pounds, which turned out to be within a pound of the actual weight of the dressed bull. Given that the crowd had a few smart people, a few informed people, and a bunch of no-hopers, what right did the average answer across the crowd have any right to be anything but dead wrong?
