Jan

12

Thinking: Fast and Slow By Daniel Kahneman

Reviewed by Russell Sears

This book is a excellent example of why a single scale rating is not sufficient. If I were to rate this book solely on the criteria of "Must Read" , I would have to give it a 10. But if you are to rate a book on its pleasure to read or even its completeness, this title would not be as high.

There are three reasons why I believe you "must" read a book. The first reason is to learn something new. The second reason is it will challenge you; it will either change the way you conduct business and your life, or it will help you understand why what you do is working and strengthen your commitment.  And the third reason is its influence on others. Even if it is wrong, or you do not believe it, it has already been influential or will change the world in which you live. Daniel Kahneman has changed the business world. This book would have to be given a 10 on all 3 of these criteria.

As an investment professional I have some expertise and experience with many of his topics. He gives much attention to biases, business analysis, other investment professionals' behavior. As an actuary, I also have experience with statistical analysis, research, the scientific method and regression analysis which are the "hero" of this book. I also have had many similar experiences as Daniel  Kahneman, so I know how people react when the "thinking" worlds collide. This book tell you the psychological processes that are occurring when things like "expertise" conflict with hard numbers. This is were I have learned much from this book. Both about myself and about why others do the things they do.

First Daniel Kahneman has a Nobel Prize. He has already had a tremendous influence on how people do things. He has brought new insight into economics. Many of the stories reflect how he helped people change their lives or business. Further, many of his stories are about other influential colleagues and his collaboration and conflicts with them. It tells tales from how to improve a military to how a checklist decreased infant mortality. It shows that understanding, the way we think, can be improved. Because our thinking has holes in it, that we are just discovering. It is not always optimal. It is clear that he has helped change the way the world conducts business.

This book will make you think: about thinking like you never have before, unless you are already at the edge of psychological "thinking" research. It tells of "fast thinking" which is largely instinct/intuition. It also tells of "slow thinking" which is more analysis. But psychologists are beginning to understand how these two "systems" interact. The interaction, the how and when each system relates was fascinating. Sometimes one system has complete control and other times the other does. However, often the control is regulated and switched between them. One system will ask the other to influence it to help make the decision. It would seem, the way I read the book, which system we  believe was in charge is often an illusion we wanted to believe. These ideas have revolutionized much of decision making.

While showing some admiration for how marvelous and powerful these systems can be, the book largely is about the ways these systems cause problems and errors. It should change the way you conduct business. It tells of common mistakes people have in thinking. It tells of shortcuts people take that do not work and of ways people over think when they should make things simpler and finally the way people, even scientists, fool themselves into believing things they should not. 

And even when it does not change your methods, it should enhance how you think about business in two ways. Either challenge you to solidify the reasons why you do the things the way you do, Or help you understand why what you are doing works. It often calls for numbers to be put on the table. That calls for you to use the scientific method to validate your methods and develop better methods, methods that can be validated. Once developed you can determine when these methods should be ignored. Often ignoring the model is much rarer than people would think. People like to tinker, throw in their opinion and add their "expertise". However, If your methods are scientifically based, rarely does this improve things. The exception he calls the "broken leg rule", that it is ok to ignore the numbers modeled if the candidate has a totally unexpected event. For example the model compares quarterbacks can be ignored if one quarterback has "broken his leg" the day before.

I have made a list of how this book has changed my way of thinking and how I will conduct business and make decisions.

I cannot stress enough how much I think leaders should read this book.

However, this book can be a difficult read. In the intro, Daniel Kahneman, says that most of the opposition to his book on thinking is because it highlights the biases in thinking, and does not give enough respect to how amazing thinking is. He implies that people have a bias against changing the status quo. People have built up businesses on these illusions and they do not want them to change. 

However, this book is hard to read and itself seems to over generalize and not always give the reader the scientific rigor he calls for from others. Being a Nobel prize winner, I would like to give him the benefit of the doubt, that much of these problems are to make the book convincing and readable to the lay person as well as the scientist. If you use regression analysis, much of the book will bore you, because it tries to simplify and explain it in words and stories how this method works. Likewise if you are not familiar and do not use these methods, it will probably be a difficult read.

Before the book got to this section however, many of his discoveries I found myself asking how solid was the evidence. For example when "system 2" (slow thinking) kicks in he states your heart rate WILL increase and pupils dilate. Simple enough, except are there exceptions to the rule, can people train themselves not to show this sign. Say perhaps a poker player. I do not know.

Many of the ideas on how these system process are stated as if they are clearly proven. While it would seem that the gray areas are not.

Further on some cases, when a few people's thinking do not show biases but most do the case of those that do is explored. While the case of those that reached the right conclusion is left as insignificant. For example a test were one person calls for help, and others have been put in isolation but perfectly capable of helping. Most do not help. Whereas if they knew nobody else was going to help they would have. Police have known this for a long time. No matter the statistics most people assume they would help if put into this situation. Social science students very familiar with statistics do not get this. Because their story of themselves disagree with the statistics. It is only after meeting these people in an interview and asked to predict if they would or would not help do they "get" what the statistics are saying. Only then do the students understand that it is difficult to predict what they would do in similar situation and catch the error when it happens. People like experiencing it for themselves is the conclusion. Not  via statistics. He implies only experience/ personal stories convince, statistics do not, often even amongst scientists. However, statistics do convince some people to change their minds, not just stories. As an actuary, often this is when money is on the line, statistics can change minds. Numbers have always spoken to me, a good statistical argument has brought much more conviction than 100 sermons (my Dad was a preacher). But it has often been a puzzle to me why this same conviction did not occur to others.

I am not arguing that my thinking does not need to be changed due to this research. I am however, left wondering how much better my thinking could be corrected if those people that did not display biases were studied. Is it really so simple as these are the people that have learned to modify their natural way of thinking. Or do these people think differently?

I would like to give a Nobel laureate the benefit of the doubt, so I figured that this was simply due to trying to convince even the non-scientist of the benefits of the scientific method. However what I found the most shocking error  was in his analysis of business world and investment professionals.

I would call this his repeated ignoring the "alpha" in regression statistics by focusing only on the correlation or the randomness of predictions. For example he tells the tale of visiting a investment firm which has 28 years of return on their professionals. The correlation from one year to the next turns out to have been near zero. His conclusion was that he had statistically proven that the professional had not added value and did not earn their bonuses. He had perhaps proven that their results still had random fluctuation and risks. However, he completely ignored did they add returns consistently, that is added alpha. For those that know nothing about linear regression, let me give an extreme example. Say return was 100% + a random variable  every year. The correlation between years is still near zero but the pro has increased your return every year

He may have done this calculation but not tell it because it did not fit the story at hand. The lesson he was trying to prove was experts opinion is no better than a random guess. But the correlation statistic did not prove this.

While perhaps there is some explanation for why he ignored this alpha but he also gave a similar view on CEO and their bonuses and salaries. He used Google founders being turned down in their early attempt to sale the company for a mere million as an example hindsight bias being luck. He told of tales of book and magazine articles of overperformance leading to guru status being signs of "reversion to the mean". While I would agree that such advice is largely to be ignored, much of their success that year is due to random luck, not superior wisdom to the other CEO's in the S&P 500. The next year they are not as likely to be as "lucky" and return to the average. Similarly for the Sports Illustrated magazine cover supposed "curse".

Even great players need some luck to be good enough to be on the magazine cover. It is not a curse, but reversion to the mean, they have no lock on luck. However, in both cases, they belong to a pretty elite group. The CEO's that run S&P covers and the major league baseball players. He did not argue that the ball player did not add value however. Yet, that is what was argued with the CEO's bonus. Despite the 100 year drift of public stock companies. I can only imagine how hard it is to "run" a modern S&P company and keep that drift in place.

A few great seasons are all we expect from an athlete. After all he will get older and be replaced by young better trained with all the accumulated wisdom wiser athletes. Likewise should we expect different from CEOs? Do both add value well beyond most people. I will let the statistic on the table decide.

For this less than scientific rigor and in parts boring writing I would have to say while a must read it is not always a joy to read and would give it a 5 or average on that scale. Great ideas, not always great execution.

After all applying his own logic on revision to the mean and performance to a Nobel laureate, and writing to the masses seems fair.


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search