Dance with Chance


by Making Luck Work for You by Spyros Makridakis, Robin Hogarth and Anil Gaba                               

As futurists we are deeply interested in uncertainty and how people make decisions in the face of uncertainty. This is because the future is inherently uncertain and yet we still have to act and make decisions in the real world. The truth is that many people choose irrationally when faced with making decisions. The book opens with a story that indicates that up to 5000 Americans may have died in three years after the September 11 terrorist attack because they make decisions to travel by car instead of flying. In addition to the deaths statisticians have calculated that a further 45,000 serious injuries and 325,000 less serious injuries would have been avoided if people’s travel patterns had remained the same. They came to this conclusion by analysing the patterns travel and injury in the period before and after the terrorist incident. Travelling by plane is safer than driving by a significant factor and yet people made the decision to drive. It is easy to see how people would have made that decision in the weeks after the September 11 incident but less obvious a year or two years later. The conclusion of the authors is that people's decision to drive was largely influenced by an illusion of control. When we get behind the wheel we believe we have a measure of control where is the reality is this is simply not true.

It is one of the authors key points that most control is an illusion and that paradoxically we can gain more influence on what happens to us once we accept that we have very little control and change decision-making to reflect this acceptance. The authors go on to use an example of a survey they conducted which asked people what they really want in their life, and then afterwards asked them what level of control they believe they have over what they want. People generally wanted similar things and recorded quite high levels of confidence in the level of control they have over those issues, while the evidence the authors present indicates the opposite.

The authors use a story to illustrate different kinds of uncertainty. The story is about a man who records his travel time to work every day in order to determine what time he should leave for work. That man then goes on holidays and is killed by a coconut falling on his head. They call the first level of uncertainty “subway risk” and the second “coconut risk”. Subway risk or uncertainty follows a well-known pattern of normal distribution. That is, there is an average travel time and a significant number of journeys fall within a certain time of the average time. This allows us to plan a departure time to work and be reasonably confident that we will arrive on time nearly all the time. The travel time that falls outside of this pattern is unusual and will be caused by events such as train crashes or severe weather that can be easily explained to the boss. In foresight and statistical terms these are low likelihood events that have a low impact.

Coconut risk or uncertainty is completely different. It is illustrated by the pattern of earthquakes. The number of earthquakes and their severity follows a fairly consistent pattern every year. Therefore we can predict the number of severe earthquakes that will occur in the world with reasonable accuracy. However we have no idea at all where those earthquakes will occur other than they are highly likely to occur in high earthquake zones. In foresight and statistical terms these are low likelihood and high impact events with high levels of uncertainty.

When thinking about the future it is important to have a good understanding of these kinds of uncertainty. Subway uncertainty can be used in planning and thinking about the future in a very defined way. Coconut uncertainty can also be used but the strategies that relate to it are different. A good example is house or life insurance. Having covered these issues of uncertainty and control the authors go on to look at specific areas. The first of these is health. Interestingly they quote a study on the first page of the health chapter which shows that identical twins die on average 10 years apart, a figure which surprised me and gives some indication that while genetics is obviously important, both chance and your own lifestyle play a large part in when you die. The authors look at a number of studies on various aspects of health such as cholesterol, vitamins, and weight. There is a lot of conflicting evidence around these areas and their conclusion is that the best thing you can do on average is to avoid the medical system. This is a very difficult thing for people to do because people do not live as averages and the fear that something that they have might be missed and will kill them drives them to take the risks involved in accessing the medical system. Again an illusion of control increases risk. The authors use cancer screening and the possibility that cancer screening in some instances finds problems that otherwise would not impinge on people’s health. There has been some debate on this in medical circles around this issue relating to screening for cancer in the last couple of weeks and if you go to our Tumblelog at and put cancer in to the search box you will find a New York Times article on the subject of screening and a reply from the American Cancer Society on the subject in two separate posts.

The authors also look at the subject of money. Their analysis matches other work that we have seen and read in that the performance of fund managers and managers of companies is by a large degree dictated by luck. On page 60 they show a table that calculates that $10,000 invested in 1802 would have been worth $8.4 billion in 2006 if you had invested in the stock market and just under $11 million if you had invested in long term government bonds. The key point they make is that the difference in return over that period is 2.4% between the two investments to create that huge discrepancy, so you really have to ask yourself is paying a significant fee to a fund manager rather than investing in an index fund really worth it. They present a lot of evidence to support this conclusion and it is really worth buying the book and just reading this information.

In Chapter 9 they examine thinking about the future and quote a study that asked students to write about the future, one group from their current state imagining their future, and one from a future state describing the history of how they got there. The results were that the descriptions of the imaginary past were a lot more detailed and rich than their descriptions of an imaginary future. There are issues with this in how people think because humans are very good at providing plausible logical stories of how an event happened once it has happened but pretty poor about doing so before an event happens. What looks simple and logical in retrospect is rarely so in prospect. The authors go on to describe how commonly simple forecasting methods produce results that are generally as good as the complex ones if not better, for a lot less cost.

Towards the end of the book the authors move towards action – so what should you do? Their recommendations are:

  1. Accept you don’t have control.
  2. Assess the level of uncertainty as best you can. (although this has problems – see our review next month on The Plight of the Fortune Tellers).
  3. Augment the range of uncertainty because humans classically underestimate the range of uncertainty we face.

They then go on to describe two types of decision making – repetitive or unique and four ways of making those decisions:

  1. Blinking – as in Malcolm Gladwell’s book Blink where gut instinct is recommended.
  2. Thinking – deeper, rational analysis.
  3. Sminking – using simple models (therefore the SM) to make decisions for you in repetitive decisions. For example they describe simple rules for assessing risk to grant people credit cards where a few criteria capture the vast majority of uncertainty and risk and significantly reduce the costs for making decision. In part they recommend some of this for recruitment and I think there is something in this. Lots and lots of money are spent on recruitment but there is evidence that the selection process and outcomes are not necessarily correlated.
  4. Using Experts – we have written before about experts and their limitation when thinking about the future but we and the authors recommend using experts in defined ways.

In the end they recommend a combination of blinking to make an initial choice and then assessing your choice by thinking (the authors contend this is how Grand Masters decide on chess moves).

At the end they write a chapter on happiness which seems to be tacked onto the book because they are interested in the subject. I think they should have left it out.

The book presents a strong case with compelling evidence and it is about a subject that I believe everyone should know more about – how uncertainty, risk and probability really work. Understanding those issues equips you far better to make decisions and think about the future.

The book is weak on the emotional basis of how we make decisions. It mentions emotions and their role in a couple of pages and is somewhat dismissive. A book cannot cover everything but I think they should have put a little more in about this subject and discussed balance in decision making more than they did. The last chapter on happiness is an indulgence that was not needed.

Overall Recommendation
I would recommend that you read the book, at least for its detailed and compelling evidence on investment returns. If you understand that better you will make better investment decisions, and if you understand that section you will gain a lot of understanding about general risk and uncertainty that should be useful to you.