Read these books.

2009-01-11 11:04

Here’s two books that you should read, even though you may have never heard of them.

Book number one: Darrell Huff’s How to Lie With Statistics. This is a widely-recommended classic, and it has survived the test of time. One particularly statistics-blind idiot I once had the misfortune to suggest this book too said that, being right, he had no need to lie. This rather misses the point; the goal is not to learn to lie, but to learn to detect it when other people lie. As Huff points out, “The crooks already know these tricks; honest men must learn them in self-defense.”

Book number two: You’re much less likely to heard of Warren Weaver’s Lady Luck, but honestly I think it may be a better book. (It may also be harder to find in print.) Lady Luck is an excellent introduction to the theory of probability in general. A topic you may not think you need to know much about, but in fact you use probability every day — or make a lot of very stupid decisions.

Statistics and probability are both fields of inquiry that many people find upsetting and uncomfortable. We like to imagine that, whether through science or religion, we can obtain certainty. We want to know what is true. However, this is not to be. What we get is instead the opportunity to have confidence of what is likely to be true. Will smoking give you cancer? Maybe. It’s not certain, by any means. But are you more likely to get cancer if you smoke than if you don’t? Yes.

Being able to think coherently about probability is not a necessity, but it is a great advantage in making informed decisions likely to lead to success. It won’t make you rich; it will increase the probability that you become richer than you would have been, though (assuming that’s even what you want…)

Peter Seebach



  1. Just got me Warren Weaver’s Lady Luck. The beginning is fun (with Lewis Carroll), it looks rather serieus further up.

    — Dirk · 2009-10-24 13:45 · #