Super Crunchers: Why Thinking-By-Numbers is the New Way to be SmartIan AyresBantam2007
At a little more than 200 pages, 'Super Crunchers' is an entertaining and informative book for generally interested readers with or without technical (i.e. on statistics or computer technology) background. The author, Ian Ayres, is a law and management professor at Yale Law and Management School. In the book, Ayres tries to show that, in many cases, predictions are better carried out by computer-based algorithms (which he calls 'Super Crunchers') than by human experts. To make his point, Ayres presents a multitude of vignettes. He starts with Orley Oshenfelter, a quantitative economist who empirically came up with a simple regression formula to predict wine quality based on rainfall and temperature data. Even though wine critics rejected his formula, it made better predictions than the critics. The next example concerns Bill James, who developed a simple formula to predict the runs created by baseball players. The formula was intended to do the job of scouts, looking for big league baseball players. Most examples in the book apply the statistical techniques of regression, randomized trials, and neuronal networks. Some of the examples covered concern firms that help people find matching partners through regression analysis. More interesting are the cases of firms like Offermatica, which found an original approach to randomly test web pages. Questionable is the case of Epagogyx, a firm that uses neuronal algorithms to predict the box office receipts of films based on the movies' scripts. Even though the author is overly enthusiastic about the future of computer based decision making, he presents interesting counterexamples from his own academical experience. There, he was involved in providing statistical evidence to support claims of racial and gender discrimination.
The book beautifully concludes with a valuable and intuitive explanation of the uses of the statistical concepts of mean and standard deviation that even the author's eight years old daughter understands, even though - according to the author - she is smart but no genius.