Having to travel regularly jobwise means spending way too much time waiting on airports. This can be used usefully working, or reading, of course. Some airports offer some additional relief by having halfway decent bookstores (Schiphol has definitely not improved in the past years, ever since Libris shut down there, but Trondheim has improved recently) where some time can be spent browsing the shelves in search of interesting books. This recently led to the discovery of this book (for the record: at Oslo, Gardermoen) that attracted my attention because uncertainty and randomness are subjects that I’ve been highly interested in during the past years.

I must confess that I had never before (consciously) heard about Mlodinow (which is pronounced as ‘ma-la-DI-nov’, as he explains in “Subliminal”), but he has worked at the Max Planck Institute and the California Institute of Technology and written books with Deepak Chopra (nothing that would register with me), Stephen Hawking (heard of him, yup) and even written for television shows MacGyver and Star Trek which earns him a lot of respect for sure!

The back cover promises that “Mlodinow reveals the psychological illusions that prevent us understanding everything from stock-picking to wine-tasting, winning the lottery to road safety, and reveals the truth about the success of sporting heroes and film stars, and even how to make sense of a blood test”. Well, yes, the book does a bit of that too, but I would say that the main part of the book is devoted to a discussion of probability and the development of probability through the ages. A typical example of how selective PR guys who write the backs of books tend to read or write…

The title of the book (that was published in 2008 and the paperback in 2009) is obviously taken from the random Brownian motion of particles, one of the many subjects that will pass in the course of 219 pages. People who have read a number of popular scientific books on psychology and mathematics like Kahneman’s “Thinking Fast And Slow”, Gigerenzer and Paulos’ “Innumeracy” or Bill Bryson’s great “A Short History of Nearly Everything” will recognize a lot, but I found that Mlodinow’s fresh (and sometimes humorous) style of writing and the use of partly different examples made it a fine read anyway.

The prologue and first chapter introduce us to the matter and illustrate with some examples that our thought processes are often flawed when it comes to understanding situations or outcomes where chance is involved. Random events often look like non-random events and it’s important not to confuse them. Our minds have difficulty handling uncertainty, yet we have to deal with situations where we have incomplete information all the time. Often things go quite well, but sometimes we’re led to wrong decisions with terrible outcomes or to wrong conclusions because we all too often confuse skill with luck and correlation with causation. Especially the latter is a problem because our desire for certainty has our mind looking for patterns and causes for events and then settling for a convenient one - such as the classic example that sparked Kahneman’s interest where air force instructors thought that yelling at cadets improved their results after an error - which in fact was an expected result of randomness through the mechanism of regression to the mean.

Mlodinow goes all the way back to the beginnings of mathematics with the ancient Greek who didn’t have any probability at all - it was the more practically oriented Romans that saw the value of this. After this there would be a serious hiatus until gamblers (and philosophers and mathematicians) from the 16th Century on would pick things up. And so we follow people like Cardano, Galileo, Pascal, Fermat, Bernoulli (several of those even), Laplace, Gauss and Bayes through concepts like sample space, the laws of large and small numbers (the latter the sarcastic moniker given to the misguided attempt to apply the law of large numbers when numbers aren’t large at all), the frequency interpretation of randomness and the subjective interpretation of randomness, the gambler’s fallacy and prosecutor’s fallacy, false positives, the law of errors and normal distributions, confidence levels, variance and standard deviation.

There are many important lessons to be learned like that it’s dangerous to judge ability by short-time results (which usually is a result of the misguided application of the law of small numbers), or assume that success is automatically derived from skill or genius. Mlodinow also stresses the fundamental difference between statistics and probability and he warns against the fact that numbers always seem to carry the weight of authority (e.g. a grade at school) even if this is not necessarily justified or supported by evidence.

From chapter 8 onward things are a bit less focused on probability and we get more into the territory that the text on the back cover promised us (true enough, there was lot of it hidden in the previous discussion too). First we look at order in chaos. Randomness is associated with disorder, but by and large things often are not that disorderly or show at least surprising regularities (one interesting illustration is that Apple had to make the shuffle function on its iPod actually less random in order that people would perceive it as random!). But as the author points out: though there are orderly patterns in random variation, patterns are not always meaningful. So it’s important to discover the meaning in patterns AND to discover when there is no meaning to them. 

The entire ninth chapter is dedicated to this important subject because it’s human nature to look for patterns and assign meaning to them (even if it isn’t there). A lot of it is rooted in our desire for control and often we somehow settle for an illusion of control by behaving as if chance events are actually subject to control - because if people would accept that things happen randomly that would mean that they are NOT in control, a situation that may be unacceptable to some. This is one of the reasons for the widespread phenomenon of confirmation bias. The author therefore concludes the chapter with three (not numbered) steps:

1) realize that chance events produce patterns,

2) learn to question our perceptions and theories, and

3) spend as much time looking for evidence that we are wrong as we spend for reasons that we are correct.

The final chapter has some interesting links to safety. Firstly it discusses Lorenz’s finding that small differences can lead to massive changes, the so-called butterfly effect which easily can be linked to Black Swans (something that isn’t mentioned at all - quite logical regarding the year of publishing). Secondly Perrow’s Normal Accident Theory is quoted, and interestingly also turned around: if the theory explains why things sometimes go wrong, it should also be able to explain why inevitably things sometimes go right (a line of thinking that wasn’t that widespread in 2008, I believe). The page before this, Mlodinow stressed that we should rather focus on the ability to react to events rather than relying on the ability to predict them, on qualities like flexibility, confidence, courage and perseverance (or, to say it in safety-terms: resilience). 

In conclusion the ‘normal accident theory of life’ shows not that the connection between actions and rewards is random, but that random influences are as important as our qualities and actions. So ability alone doesn’t guarantee success, but if we are able we can enhance our chances by the number of chances taken and opportunities seized. Mlodinow says that it’s important to plan, but with our eyes open and understanding how things really work, and not just how we expect or would like them to work.

As a whole a fine and entertaining introduction to probability, randomness and some mistakes that we should watch out for!

Find out more about Leonard's work on his website:

I’ve read the Penguin paperback, ISBN 978-0-141-02647-3