A review of Cognition And Chance: The Psychology Of Probabilistic Reasoning by Raymond S. Nickerson.

This book was not perfect, but I did find it very useful and interesting. Specifically I was interested in the first half which talked about the philosophical nature of issues surrounding chance and randomness. The author seemed to make a reasonably thorough survey of the paradoxes and dilemmas that arise from the various (he enumerates them) systems of thought regarding probability and chance.

I learned a lot about interesting issues in probability. I also was able to feel like I completely understood Bayes' Theorem for the first time as opposed to just going through the mechanics of it. The author strips away all of the ambiguous bits and provides some examples which are so clear cut that the core essence of Bayesian thinking is laid bare. If you need to learn Bayes Theorem, just reading chapter 4 could be worth the trip to the library.

The last 5 chapters of the book were less interesting to me. They were filled with interesting stuff to be sure, but I felt that the treatment of this material was half-hearted and better found elsewhere. It’s mostly lots of Kahneman and Tversky (and their imitators) style behavioral economics experimental results. That’s interesting, but the Wikipedia page on Cognitive Biases is more than sufficient for the interested reader. Slogging through an unnecessary 150 pages was also more than sufficient. The writing is good and if you had no exposure to this kind of stuff, it would be as interesting as it ever is.

My biggest complaint about the book is that it was long and the chapters were also very long. I feel like it should have been two books (one on probability and one on cognitive biases). Instead of each chapter being around 50 pages, that could have been better organized. For example chapter 9 contained a section on "Conditional Probabilities" which was surprising since there was a whole long chapter (4) on Bayesian thinking. Chapter 9 even had Bayes' Theorem again.

I must say that it was well written. It was clear and grammatically easy to read. It was filled with useful citations and quotations. Many of the topics in the early chapters were quite fascinating. If you’re interested in where modern technical thinking’s ability to deal with uncertainty breaks down, the first 7 chapters of this book are fantastic. If you’re interested in how the human brain naturally deals (or doesn’t deal) with uncertainty, the second half may be of interest. In all the first half was worth 5 stars to me.