Truly a fascinating book! De Mesquita has used the game theory of John Nash to develop a series of models that help him to predict future events. The predictions are not simply binary (yes/no) prognostications--they are in-depth analyses that describe what will happen, and why. The author claims a 90% success rate. The last few chapters include a set of detailed predictions made by HIS STUDENTS using his models. Some of the predictions (most notably, Pakistan) are starting to come to pass, now. Others will be proven--or disproven--within the near future. Of course, this book is sort of an advertisement for the author's consulting company. Nevertheless, I've been recommending this book to everybody.

A really remarkable book. The author is the pre-eminent scholar on applying algorithmic techniques to predicting macro scale problems/solutions in economics and politics. It has interesting and wide-scale implications. It shows just how predictable and primitive human incentives are.

Entertaining applications of game theory (w/o much math, mainly computer simulation results) to politics, lawsuits, and buying car. May be worth it for the car buying advice alone.

A popularized (no equations) book on using game theory to predict networked political problems which often turn out to be resolved or described by a mathematical model of the political actors' self-interest. Worth reading to realize that such models are usually intuitive.