Investing: The Last Liberal Art (Columbia Business School Publishing)

by Robert G. Hagstrom

and singing that the scene became unruly. Now Arthur was confronted with a problem: How could he decide which nights to go to El Farol and which nights to stay home? The chore of having to decide led him to formulate a mathematical theory he named the El Farol Problem. It has, he says, all the characteristics of a complex adaptive system. Suppose, says Arthur, there are one hundred people in Santa Fe who are interested in going to El Farol to listen to Irish music, but none of them wants to go if the bar is going to be crowded. Now also suppose the bar published its weekly attendance for the past ten weeks. With this information, the music lovers will build models to predict how many people will show up next Thursday. Some may figure that it will be approximately the same number of people as last week. Others will take an average of the last few weeks. A few will attempt to correlate attendance data to the weather or to other activities for the same audience. There will be endless ways to build models to predict how many people will go to the bar. Now let’s say that every lover of Irish music decides that the comfort level in the small bar is sixty people. All one hundred people will decide, using whatever predictor has been the most accurate over the last few weeks, when the limit is going to be reached. Because each person has a different predictor, on any given Thursday some people will turn up at El Farol and others will stay home because their model has predicted more...