This is the latest in a long line of excellent popular science books by William Poundstone. Its topic is "behavioural decision theory", the study of how individuals make systematically irrational decisions in choice situations, risky gambles (where the probabilities are known), and uncertain gambles (where probabilities are unknown).
A lot of this field has focused on the latter two segments: decision making under risk and uncertainty. This was in some ways due to a methodological quirk, as these two scenarios were much easier to test with experiments, questionnaires, and real-life data.
Poundstone covers both these areas, but he also looks at how some companies are beginning to exploit consumers' common irrationalities. For instance, a fundamental axiom of rational choice theory is the "independence of irrelevant alternatives". Say you are choosing between good A and good B. Adding a third good to the choice menu -- good C -- shouldn't affect your original choice between good A and good B. Say you initially chose good A from the set {A,B}; then, the independence of irrelevant alternatives states that you shouldn't then choose good B from the set {A,B,C} (although you could choose C).
This axiom frequently gets violated in the real world, and clever marketers are wising up to this. Say you are choosing from two vacuum cleaners: a very cheap cleaner (A), and a more expensive model (B). In this case many people would choose A. However, the company can alter your choice by adding a third vacuum cleaner: an incredibly expensive state of the art model costing thousands of pounds (C). Even if you will never buy this new vacuum cleaner (it is an irrelevant alternative), it can still alter your choice: by comparison it will make model B seem much less expensive than before, and will make A seem like a shoddy bargain basement model. The mere presence of C has shifted consumers into spending much more money on vacuum cleaners than they did before, even if nobody buys C.
This book is a lot of fun and well worth the read.