"Chaos in Hollywood: Can science explain why a movie is a hit or a flop?," byJohn Cassidy. The New Yorker, 31 March 1997, pages 36-44.
Making movies is a risky business. There are plenty of examples showing thatthe right star, a catchy concept, or a brilliant script provide littleinsurance against financial disaster. Is there any way to predict which movieswill be profitable? In this article, the author describes the ideas ofeconomists Art De Vany and David Walls, whose research into which films profitand which ones bust convinced them that chaos theory describes the industrypretty well. An unpredictable "information cascade"---in which potentialviewers size up a movie based on information they get from friends, moviecritics, and other sources---cause viewers to gravitate to a handful of moviesout of the hundreds released each year. The conclusion is that, like long-termweather forecasting, predicting which movies will be successful has a built-incomplexity that makes it impossible to do with any accuracy. In interviewswith Hollywood film executives, the author of the article found that theydidn't give much credence to the economists' views. While the executivesagreed it is difficult to predict which movies would be hits, they believedthat certain factors, such as a big star or making a movie into an "event," arethe keys to success.
--- Allyn Jackson