Book Review: Fooled By Randomness

Fooled by Randomness from Nassim Nicholas Taleb explores the hidden role of chance in life and the fallibility of human knowledge. It explains how luck, uncertainty, probability, human error, risk, and decision-making work together to influence our actions, set against the backdrop of business and investing.

{{divider}}

Chance is liberating: it respects failure, and limits the merit that devolves from success - David Walsh

{{divider}}

Balance is key in life, however as a society we have a propensity to focus heavily on just one side of the spectrum, limiting our understanding of the bigger picture. One example of this is how we often mistake luck and randomness for skill and determinism. Dentistry and plumbing both require skill, however in an environment such as the stock market, while educated estimations are a necessity, on the whole the premise is essentially gambling.

{{divider}}

What’s interesting about this book is the reminder that all theories are only true until proven false. Essentially, things constantly change and the next observation may prove us wrong. The basis of all empirical science is a process called induction - we infer things about the nature of the world based on our observations. Our rational, empirical science ultimately comes back to the subjective nature of our perception, interpretation and context.

{{divider}}

Therefore when trying to prove a scientific theory, scientists focus on trying to prove it wrong, and the evolution of science and our understanding of the world constantly updates and progresses in this fashion. When it comes to business and investing, we therefore cannot assume that the past is a relevant sample of what the future holds, and the only option is constant change through adaptation.

{{divider}}

For example, if stock prices always rose on Monday, rational investors would buy all stock on Sundays, thus changing the market dynamic and eliminating the effect. Although our brain is hardwired to feel otherwise, life is unfair and not always linear. An example of this is the layout of the common keyboard, also known as the QWERTY layout. It’s not actually the most optimal solution, and simply arose from the initial need to avoid jamming old-fashioned type writers. However this solution has prevailed thanks to a ‘path-dependent outcome’, if this were not now the dominant form in the market and we were to start again from scratch, we would probably design a different layout.

{{divider}}

Sub-optimal products such as this come to dominate the market because they have passed ‘the tipping point’ that creates a positive feedback loop where new customers jump on the bandwagon because they want to be part of the trend. Such non-linear events as this are not rational or predictable. This is why guerrilla tactics can have much larger effects. Our nature is to assume that incrementally adding grains of sand to a sandcastle will have minimal effects, however when such an incremental change reaches the tipping point, it can ripple out into huge effects, such as a toppling of the entire system.

{{divider}}

Without Rhyme Or Reason

{{divider}}

Our fallible human reasoning is context dependent and mostly based on simple heuristics. Our hominid brains are already ill-equipped to deal with today’s high information environment. Despite what we may believe, as opposed to being a sophisticated AI machine, it is more like a patch-work of rules and shortcuts called heuristics.

{{divider}}

These heuristics have evolved to help us make quick decisions so that we can act, instead of contemplating and calculating for eternity.  Unfortunately the price we pay for this is that our reasoning becomes irrational and marred by bias. For example, we may disproportionately attribute our successes to ability, and our failures to bad luck - this is known as attribution bias.

{{divider}}

Our thinking is also path dependent, meaning that the route with which we arrive at a given situation dictates how we think about it, even though there may not necessarily be any causal link between seemingly associated events. It also dictates how we feel about given outcomes because of comparison bias - if you won 5 million today and lost 4 million tomorrow, you would be much unhappier than if you’d simply won 1 million tomorrow, even though the end result is the same.

{{divider}}

Path dependency is why people cling to their existing opinions; scientists and politicians get attached to the ideas they advocate, and refuse to change their minds, even in the face of contradictory information. This also happens because humans have a propensity to attach their identity to their ideals, so counter-points can feel like a personal attack.

{{divider}}

From an evolutionary standpoint it makes sense to feel attached to the things we have invested a lot of time in, but this effort is counterproductive when it inhibits our ability to change our minds and contradict ourselves at will. This is why the most certain people are often those who know the least, whereas those that are constantly questioning themselves have probably explored the situation more accurately - the more you know, the more you know you don’t know.

{{divider}}

Intuitive Reasoning

{{divider}}

We can’t rely too heavily on just our rational reasoning, or our intuitive emotions alone, and the best decision often involves an alignment of the two. Emotions are irrational precisely to stop us from temporising and deliberating over a situation for too long. However neurobiologists are finding that we physiologically feel our emotions first, and then attempt to rationalise them later. This means that our emotions have a stronger influence over our thinking, rather than the other way around.

{{divider}}

Acknowledging this may be key to making more intelligent decisions, but it’s not that simple, depending on how equally adept we are at talking ourselves into and out of certain situations, thereby making us feel differently about them. When Ulysses sailed his ship past the deadly yet seductive sirens, he had his men pour wax in their ears, so that they would not be distracted by their song.

{{divider}}

Pattern And Prediction

{{divider}}

In retrospect, we always find patterns, causes and explanations in past events, but these mostly do not help us when it comes to predicting the future. Learning from history does not come naturally to humans, and in retrospect, earlier events always seem more predictable than they really were at the time - this is known as hindsight bias.

{{divider}}

If you sufficiently analysed any amount of past data, some inevitable amount of pattern would emerge from it - not necessarily due to causal relationship, but because of our ability to find pattern in randomness. This is why you hear claims of people being able to find predictions of past world events from random sources such as statistical irregularities in the Bible. Like our ancestors who divined the future by examining bird entrails, we tend to naturally find patterns in causal relationships when there may not be any.

{{divider}}

History Doesn’t Repeat Itself, But It Often Rhymes

{{divider}}

This is not to say that data analysis is completely useless - and natural repeating patterns do exist in life and our environments, but we need to be more aware of our inherent bias when interpreting select sets of results. As COVID has demonstrated, we are also inherently poor at understanding the potential impact of rare events. When hedge funds report losses, they often refer to large and ‘unexpected events’, factors that their risk management models were completely unprepared for. These statements ignore the fact that ‘things which have never happened before’ actually happen all the time, and are always unexpected. In fact, Black Swan events such as COVID are becoming increasingly common.

{{divider}}

An unexpected event can dwarf expected events in magnitude, which is why they often show up as outliers in data, and are ignored when conducting risk analysis. In a similar fashion, early climate researchers removed the largest temperature spikes in their data, because they thought them unlikely to occur. But in fact, these spikes added disproportionately to climate change, an outcome the resultant climate model failed to anticipate.

{{divider}}

Levity and Stoicism

{{divider}}

This brings our focus back to the process, rather than the outcome. The end result is often more out of our hands than we would like to imagine, and the best we can do is to bring our attention to enjoying the journey and not the destination. We can use humour to enjoy harmless randomness, and stoicism to endure the more difficult kind. As the Yiddish say, ‘if I must eat pork, it had better be the best kind’.

{{divider}}

Despite our best efforts, we may all fall victim to randomness of the more difficult kind, such as an unexpected diagnosis. And in such an event, a common initial reaction is for our minds to search for a root cause, however irrational. Stoicism encourages us to accept what we cannot change in finding a way to move forwards, without necessarily interpreting it as something doggedly deterministic.

{{divider}}

Nassim Nicholas Taleb is a Lebanese-American essayist, scholar, mathematical statistician, and former option trader and risk analyst, whose work concerns problems of randomness, probability, and uncertainty.

{{divider}}

‍