Gary Smith is my kind of curmudgeon. This book is basically a long set of essays on how people screw up their stats. Sometimes this is particular complaints about particular (often famous) scholars. Sometimes it’s discussion of common conceptual errors. I especially enjoy how pretty much nobody is spared, from politicians to famous scholars, to beloved intellectuals like Steve Levitt of “Freakonomics” fame.
If you’re into turning a skeptical eye on everything, or if you want to learn how to think more clearly about science, the news, politics, economics, and anywhere else people are likely to abuse stats, this is a a great book for you.