Paulos did a fine job with his previous book, Innumeracy. However, this book falls short. I was hoping to see concrete examples of news stories "gone mathematically bad," and then have Paulos show what was wrong with those stories. However, the examples used were more ambiguous -- i.e., general stories about certain topics. It was even more frustrating given that his chapter titles appeared to promise some concrete news story, but then failed to deliver. For example, "Afta Nafta, Lafta" had nothing to do with NAFTA or free trade. While it did provide some less-than-illuminating discussion about how details tend to come after the headline and first paragraph, I failed to see how any of this related to a substantive topic. Also, the preceeding chapter about economic forecasts came to the remarkably banal conclusion that reality can be more complex than simple models, illustrated by a transformationof the Laffer Curve into a squiggly line. OK, fine, but how about an admission that Laffer's analysis came with the standard "ceteris paribus" clause that all social scientists should be familiar with. And how about citing empirical investigations that test Laffer's predictions while controlling for confounding factors?
There are some good lessons scattered throughout this book, but as a previous reviewer said, it does tend to be scattered and lacking any common thread. Joel Best's "Damned Lies and Statistics" is a much better read.