A review by benjamin_manning
The Evolution of Cooperation: Revised Edition by Robert Axelrod

5.0

Simply sensational - I'm obsessed. Finished this material in my game theory course a few weeks ago and the prof recommended this book. This book is a full dive into iterated prisoner's dilemmas (basically playing some non-zero sum game over and over) to see what the best behavior for people is over the long term in possibilities where they can cooperate versus try to take advantage of one another. Axelrod makes some pretty tough material completely successful to those without mathematical inclination and lays out some really clever real world applications to game theory.

A few things I learned:

1. In order to effectively cooperate, one must not be too complicated in what they're doing! I never though about this before, but it makes sense. For other people to want to continually engage with you you, they have to understand you.

2. Soldiers in the trenches of world war I used to set up unofficial treaties on the front lines, these came about almost entirely spontaneously and then continued for long periods of times since the same units were facing each other over and over; fascinating.

3. While I loved this book, I'll admit that I now notice (courtesy of Duncan Watts) that Axelrod offers a lot of VERY definitive explanations when they might not totally be warranted about why and when cooperation happens. This are the "everything is obvious" afterwards type understanding of things that are difficult to have causal attribution. It's REALLY easy to think one knows why something occurs, even with compelling evidence, yet totally be wrong and I think that some of Axelrod's explanations of real-world cooperation might suffer from this problem a bit.