Reckoning With Risk, by Gerd Gigerenzer

The test for breast cancer is extremely reliable. It correctly detects breast cancer in 90% of cases when the cancer does exist, and only mistakenly reports it in 9% of cases when the cancer doesn't exist. The incidence of breast cancer in women is 1 in 100. Suppose you (or, for men, a woman close to you) take a test for breast cancer, and unfortunately it returns a positive result (i.e. it detects the cancer). What is the probability that you do have breast cancer? Would you be surprised to know it's just 10%? Not 90%, 99% or some other high number?

Another example: DNA testing on a murder weapon matches your DNA, and a forensic expert says there's only a 1 in 100,000 chance of that happening. Are you doomed? Would you be surprised to know that in a city of, say, 2 million people, this means you're 95% likely to be NOT guilty, based on that DNA evidence alone?

Do these examples surprise and confuse you? If so, take heart: They surprise and confuse most people - laypeople and experts (doctors and lawyers) alike. Unfortunately, this can have disastrous - sometimes tragic - consequences in law, medicine and other fields.

This is the topic of Gerd Gigerenzer's excellent book about working with risk and uncertainty. Read it and you might be horrified at some of the horrible mistakes being made by experts giving advice. At least you'll be in a better position to question them and become better informed.

Is this the best book ever written about dealing with uncertainty? I'm not sure. But it's certainly well worth the read.