The most important facts about this book are (a) it is not for beginners, or even for those who apply game theory but do not care about theory in itself; and (b) it is very interesting and entertaining for game theorists and those who like game theory as an intellectual exercise. The range of authors interviewed is very wide, so there is much to learn for most readers, who will be acquainted only with a subset of the areas discussed. There is perhaps too much emphasis on economics and not enough on biology (and none on any other behavioral discipline), given that game theory has many more solid explanatory and empirical successes in animal behavior theory and epidemiology than in economics, where only auctions, and perhaps industrial organization, are solid successes.
Of the five questions asked each of these game theorists, the only generally interesting question is an inquiry into what the interviewee thinks are really interesting successes in his own work and others. The answers often amount to a short review and bibliography on the person's work, which is very valuable.
I found the views of these game theorists (most of whom I know) to be generally mature and insightful. The only wrong-headed and sophomoric comments I encountered were Robert Aumann's critique of experimental economics which were incorrect, uninsightful, and not worthy of this great man. The idea of theorists criticizing empirical researchers would be unheard of in a truly scientific field, which economics is not, precisely because it refuses to take theory as a handmaiden to the facts. What Aumann does not appreciate is that game theory has revolutionized the methodology of data gathering in the behavioral sciences, as witnessed by the Nobel prizes of Daniel Kahneman, Vernon Smith, and others. Without game theory, the behavioral economics revolution could not have happened, and microeconomic theory would still be the wasteland of arid theorizing that students have had to suffer through since the mid-Twentieth century.
Let me give some specifics from Aumann's interview. "The conclusions of behavioral economics are based to a large extent on questionnaires and polls," says Aumann. This is just completely false. I can't think of a single result in the experimental literature that depends in any way on polls, and questionnaires are used only for debriefing, demographic information, and the like. "Another problem is that the conclusions are based partly on experiments with money. Though there is some incentive given, this is usually small and people don't really pay much attention to it," says Aumann. This is quite misleading. Subjects in most experiments are motivated to participate solely on the basis of financial reward, and the amounts involved are comparable to wages that they could earn elsewhere. Field studies often use incentives that are extremely high compared to the daily wage, and there is no data indicating that subjects "don't pay much attention" to the monetary payoffs to the games they play. Aumann's views on this matter are shockingly uninformed. Aumann also says "I agree with the behavioral economists that people don't think about the decisions they make." First, this is a gross distortion of what behavioral economists say, to the point of being a malicious parody. Of course people think about the decisions they make, and humans are very, very fine decision-makers. Aumann's alternative account of decision-making is plausible, but not at all in conflict with the experimental literature. I am deeply distressed by Aumann's remarks. They indicate a serious area of misunderstanding in game theory. What is surprising, perhaps, is that there is so much unity in the views of game theorists concerning what we have learned and what remains to be accomplished in the future.