Pun Salad

The good folks at the UNH Library ordered this at my suggestion. It is an overview of how statistical analysis and modeling of (typically) large datasets predicts and explains events in a myriad of settings. Ayres, a law prof at Yale, draws examples from the operating room, car dealerships, market research, moviemaking, wine criticism, sports, and many, many more. The underlying theme is that people who use statistics, regressions, and neural networks are ever more likely to outperform their counterparts who forego these tools and rely on more old-fashioned forms of expertise.

The book is heavily marred by its style, unfortunately. The constant repetition of the "Super Crunch*" tag wears very thin by approximately page 20. The Super Crunchers are coming to Super Crunch your data in the Super Crunching revolution, etc., etc. It sounds like a marketing campaign for a new cereal.

This irritation is augmented by the book's frothy style, which may work OK in a magazine column (Ayres writes them for Forbes), but … well, take this sentence from page 129: "More and more information is digitalized in binary bytes." Oh, really? Binary bytes, you say? As compared to those other kind of bytes they used to use? And, since "digitalized" kind of means that the information is being represented in binary, isn't that redundant anyway? And you say more and more information is going that way? Gee, and here I thought it was less and less. Silly me. Thanks for clearing that up.

Argh. Tyler Cowen called this book "highly readable", but it obviously doesn't work for me.

A possibly more substantive criticism of the book is the glaring omission of the stock market. If anyone has a lot of data, a lot of computer power, and a lot of incentive, it's stock-pickers. Where's the evidence that "Super Crunching" enables them to uncannily out-predict the seat-of-the-pants experts?

That said, however, there are a number of good stories and compelling examples in the book. You can read the first chapter at the New York Times. A generally positive review by David Leonhardt is here; he discovers, however, another of Ayres' unfortunate tendencies: to—um—reproduce others' words in his book without quotation marks.