Behavioral Economics has been an expanding section of bookstores for a few years now, and a lot of the books coming out are thought-provoking and intelligent. This one makes the others look better by comparison.
I have global complaints, like the plodding pace of the writing, the confusing way in which some of the experiments are presented, the odd withholding of information (at one point the author declines to explain the difference between two different auction styles, citing the complexity, but they must have been able to explain each style to the participants in order for the experiment to work), and such. Even if the experiments themselves were done well, this would be a major reason to avoid this book, even for someone doing a thorough reading of the lay-literature of Behavioral Economics.
Then there's some nitpickier complaints, like how the author feels compelled to mention the horrible injury, and arduous healing process, that he suffered years before. While i do understand that this was a major life event for him, and in fact got him started in the field, it's unclear what it adds to the book to mention it every chapter. He's also compelled to mention his other book repeatedly, just in case we missed the fact that this is his second book, even though this one doesn't build on the last one directly.
The major failing of this book, though, is the experiments themselves. And, for reasons of sensationalism and piling on the bandwagon of complaining about the 2008 financial crisis, one of the most flawed experiments leads the book. In short, in an attempt to show that paying financial experts huge bonuses actually harms their efficacy, they show that paying random passers-by huge bonuses for tasks totally unrelated to their actual profession harms their efficacy. There are several marked differences between the experimental subjects and the bankers to whom the author would like to generalize the results (no, i'm not arguing that bankers are 'special' as the author claims his banker friends often do, just that they're businesspeople and random people on the street in India probably aren't, if they're wandering down the street in the middle of the day and have the time to stop and do the experiment). There are several marked differences in the way the bonuses were offered, whether they were expected, how the bonuses related to the tasks, etc. Basically, either the author totally failed to convey how the experiment was actually related to his conclusion, or he took an experiment only tangentially related to the topic and claimed it was related in order to lead his book with a chapter about the 2008 financial crisis.
Other experiments, while less flawed, are just pointless. For example, he showed that if someone's rude to you, you're more likely to take an opportunity to punish them than you would be if they weren't rude to you. (Shocking, i know, but he has the data to back up the claim.)
Another great sin is that the author only rarely follows through on the promise of the subtitle. There are dribs and drabs of comments on how being irrational can be beneficial, but more often it's simply observations that humans aren't, in fact, machines of pure reason that always seek the greatest gain for the least effort. After a few requisite digs at how Rational Economists totally fail to actually predict human behavior, this 'oh, and check that out, people aren't perfectly rational!' mantra gets redundant.
If you like books on Behavioral Economics, read Freakonomics and SuperFreakonomics. They are better written, with better data and better analysis. And if you've already read those two, stay well clear of this one, as you'll find it vapid, slow, frustrating, and poorly written when compared to those.