I'm not sure why so many readers enjoyed this book. First of all, if you act how the book tells you to act, you are going to be a jerk. Second, I possess only the most basic familiarity with Bayesian statistics, economics and heuristics and I found this book to not only oversimplified but patently wrong in many places. Amazingly, this book falls victim to many (if not most) of the fallacies of which it attempts to disabuse the reader. Three examples:
1. The chapter explaining Base Rate Bias (which says we systematically fail to account for the base rate of an event's occurrence) uses the example of Mike, a fan of Mozart. Is Mike more likely to be a truck driver or an English professor? If you said "professor" you're wrong because you fell victim to the "base rate bias." Hahahaha. Isn't irrationality funny? There are 100 times more truck drivers than English professors so it's statistically more likely that Mike is a truck driver, right? WRONG. We actually don't know the answer. This example succumbs to the very bias it ostensibly reveals. Mike-is-a-truck-driver makes sense as an answer only if the incidence of Mozart-liking in English Professors is less than 100 times greater than the incidence of Mozart-liking in truck drivers. In other words, we cannot say whether Mike is more likely to be a truck driver unless we know the BASE RATE of Mozart-liking. If the incidence of Mozart-liking in English professors is 75% but only .001% in truck drivers, it's more likely that Mike is an English professor even if there are 100 times as many truck drivers as English professors. Yet, the author sticks by his "rational" conclusion that Mike is more likely to be a truck driver.
2. Dobelli stretches his "visionary" thinking to unfathomable depths of stupidity when discussing how humans (don't) understand probabilities. In an example, he claims that Water Treatment 1 is better than Water Treatment 2. Why? Because WT1 reduced the risk of death from 5% to 2%, a 3% drop. Shoddy WT2, on the other hand, only reduced the risk of death by 1%, from 1% to 0%. I'm not making this up. Dobelli claims WT1 is 3 times better than WT2! If that's the case, why aren't we all excited about hypothetical WT3 which reduces the risk from 100% to 70%?!? It's 10 times better than WT1 and 30 times better than WT2! We are not excited about WT3 because no one would choose a 70% (or even a 2%) chance of death over a 0% chance of death. (Oh wait, I shouldn't say no one because another chapter told me I'm supposed to expect improbable events - you know those events that, by definition, happen rarely.) Yet, Dobelli uses the rate of reduction as the primary measure rather than the obvious and preferable total probability of death. That's stupid and displays unclear thinking. Astonishingly, later in that same water treatment chapter, he says the only time it's worth considering small probability events is when their occurrence would be catastrophic. You mean "catastrophic" like death from drinking unsafe water?!? Apparently not.
3. The chapter on Groupthink references a study that used a methodology clearly falling under the Survivorship Bias (Chapter 1). The study looked at all the major groupthink errors, found common characteristics and assumes those common characteristics caused the failures. That is almost the exact definition given in Chapter 1 of the Survivorship Bias. I feel like I'm taking crazy pills here. I can't be the only one who sees this.
If you are serious about learning about our biases and fallacies, read Fooled by Randomness and Black Swan by Nassim Taleb, The Undercover Economist by Steven Landsburg, and blogs like Overcoming Bias and Less Wrong. Then go on iTunesU and find a basic probability and statistics course. You will be much better off than wasting your time "learning" how to spot your own irrationality from this book.