- Randomness plays a much bigger role in life than we’d like to admit.
- We cannot predict the future based on the past because the future is different.
- Silent evidence prevents us from having an accurate vision of history.
What Fooled by Randomness Talks About
Fooled by Randomness is a book written by Nassim Taleb. It explains that randomness, because it is unseen, plays a much bigger role in life than we are ready to accept. The book explores many situations where randomness should be taken into account more often and explains different principles related to that topic.
Taleb used to work as a trader in New York. Then he retired and wrote books.
This one is the first installment of The Incerto, Taleb’s series discussing randomness and asymmetry in life.
The Incerto is made out of:
The main thesis of the book is that we often underestimate the role of chance and randomness in life.
Taleb analyzes several situations where randomness intervenes without our noticing.
This book had an unexpected impact on me.
It killed my impostor syndrome. I learned that we are not as responsible for where we are in life as we think we are. Sometimes, randomness enables you to enjoy perks you think you don’t deserve to enjoy. That doesn’t mean it is wrong. Such is life.
Things in life aren’t as optimized as we think. Randomness can positively impact your life as much as it can negatively impact it.
Just like that, I stopped feeling bad for reaching positions I originally didn’t believe I deserve to reach. If it was random, why should I feel bad about it?
Fooled by Randomness is a great book, but the author unnecessarily lengthens the book at times (for his own enjoyment, sure, but it annoys me).
As a result, I’ll give this book a 9/10.
Short Summary of Fooled by Randomness
Life is mostly random.
We can’t perceive this randomness because our brains are not wired to.
We approach and interpret life using heuristics, a series of mental shortcuts that help us make sense of the world in a logical manner. However, the way we perceive the world isn’t at all how the world is. We are biased.
Let’s take narratives as an example.
Narratives are compelling stories we tell ourselves. A narrative links a series of random actions to each other to make them appear as a logical sequence of events.
Eg: when I tell myself my own life story, I say that I studied for one year but didn’t like it, so I took a gap year. I make it sound like the gap year was a logical thing to do, while it was in fact completely random.
Part of the job of building narratives is creating “past predictions.”
Making past predictions is the action of looking at historical events and making these events appear as a logical sequence, leading up to the situation we have today. Past predictions make history sound like all that happened in the past was “bound to happen”.
Eg: When we look at it today, many say that 9/11 was perfectly predictable. No, it wasn’t. It only appears predictable after it happened, due to past predictions.
So, what do we use past predictions and narratives for?
We use past predictions and narratives as sources of information to predict the future. The belief we have is that the future will resemble the past, so if we look at the past, we will be able to catch a glimpse of the future.
Unfortunately, this is another fallacy. The future is random and unpredictable.
It’s so unpredictable that the overwhelming majority of predictions we make end up being wrong. This is the first problem. The second problem is that we have a very narrow understanding of history.
History as we know it is not perfect. It’s an account of some of the events that happened in the past. It is missing all of the events that didn’t happen (in an alternative universe, the Titanic didn’t sink) as well as the events we don’t know about.
The past, as a result, is only a little more well-known than the future and the present.
The present isn’t well-known at all.
Few are aware that they’re living history when they’re actually living history. These moments are always realized ulteriorly (eg: no one knew at the beginning of WWI, that it was the beginning of WWI).
Nobody can predict the future, as we said.
People that do and that, somehow, end up being right every time (eg: “legendary investors”) are called lucky fools.
Lucky fools exist because enough people tried to do what they did so that at least one would succeed.
Eg: if 10 000 people try to become day traders, at least one will succeed out of mere luck.
Out of these 10 000 people, the successful ones, the “survivors” are hailed as “legendary investors” by the press, while their success was likely due to mere luck instead of skill. One person, eventually, wins the lottery because there are enough players for that to happen.
The rest of day traders (or lottery losers), those who failed, are called silent evidence.
They’re evidence that predicting the future isn’t possible. But they’re not taken into account in the press, of course. Failure is invisible.
Business books are always about people that succeed, never about those that lose (except for What I learned Losing a Million Dollars). Business books don’t take silent evidence into account.
Silent evidence is another reason why we can’t trust history. Since we can’t trust history, it would be foolish to believe we can predict the future based on the past.
And even if we had a perfect picture of the past, we still wouldn’t be able to use it to predict the future since the past does not resemble the future.
There will always be events happening in the future that have never happened before – hence impossible to predict.
These events, when unpredictable and with high impact, are called Black Swans.
Table of Content
Click to expand/collapse
- Chapter 1: If you’re so Rich, Why Aren’t You So Smart?
- Chapter 2: A Bizarre Accounting Method
- Chapter 3: A Mathematical Meditation on History
- Chapter 4: Randomness, Nonsense, And The Scientific Intellectual
- Chapter 5: Survival Of The Least Fit—Can Evolution Be Fooled By Randomness?
- Chapter 6: Skewness And Asymmetry
- Chapter 7: The Problem of Induction
- Chapter 8: Too Many Millionaires Next Door
- Chapter 9: It Is Easier to Buy and Sell Than Fry an Egg
- Chapter 10: Loser Takes All—on The Nonlinearities Of Life
- Chapter 11: Randomness And Our Mind: We Are Probability Blind
- Chapter 12: Gamblers’ Ticks And Pigeons In A Box
- Chapter 13: Carneades Comes To Rome: On Probability And Skepticism
- Chapter 14: Bacchus Abandons Antony
Summary of Fooled by Randomness Written by Nassim Taleb
Probability is not a mere computation of odds on the dice or more complicated variants; it is the acceptance of the lack of certainty in our knowledge and the development of methods for dealing with our ignorance.
The lucky fool is someone that attributes his success to skills instead of luck, when luck played the most part.
In general, we widely underestimate the role of luck in life. We see patterns where there are none and try to explain how events “logically lined up” while in fact, they happened randomly.
The book will further talk about situations in which people mistake luck for skills. The reason for this mistake is that humans are not wired to deal with probabilities.
Part I: Solon’s Warning
When Solon, a wise Greek legislator, visited Croesus, a very rich man, he warned him not to take his wealth for granted as life can change quickly. You can be rich today, and lose everything tomorrow.
-> what comes with luck, can be taken away by luck.
-> the black swan problem: it does not matter how often you succeed if one failure makes you lose it all.
Chapter 1: If you’re so Rich, Why Aren’t You So Smart?
Lucky fools don’t realize they are lucky. In fact, they’re persuaded their success is due to their action and knowledge.
When someone experiences success, their brain rewards them with serotonin. They become confident and seek to reiterate their success. The more they win, the more they want it. It’s a virtuous circle.
When the first failure hits, they start doubting. Then they fail a second time. And the virtuous circle becomes a vicious one.
People cannot conceal their emotions, and emotions play a huge part in how we act.
People that become leaders of their organization are not more skilled than others. They just feel and show emotions differently – they have charisma – which helps them become leaders.
Your Dentist Is Rich
People don’t understand probability.
Imagine a dentist doing his job for 30 years, living in an upper-middle-class house.
Now imagine a janitor winning the lottery and moving in the dentist’s neighborhood, in an upper-middle-class house.
If we look at these two people’s lives in parallel universes, which one is more likely to always obtain the same result?
The dentist. The janitor wins the lottery in one universe out of one million.
In essence, this is what probability is. How likely will one event repeat in a parallel universe?
Chapter 2: A Bizarre Accounting Method
You can’t judge a performance according to the results it gets. You need to judge it according to the cost of the alternatives.
Looking at what might have been is called looking at alternative histories.
Imagine you play Russian roulette for $10 million. 5 times out of 6, you win. 1/6, you die.
Imagine you win and become rich. You write a blog post about how you made $10 million playing Russian roulette, and suddenly, everyone starts to play Russian roulette to become rich.
Would it be smart? No.
Probability is about assessing what may happen in other universes. In the Russian roulette case, there are 6 outcomes (because there is one bullet in a barrel can have 6 bullets in total.)
- No bullet – you survive
- No bullet – you survive
- No bullet – you survive
- No bullet – you survive
- No bullet – you survive
- Bullet – you die
Six outcomes, but one will happen in reality.
Now, imagine thousands of people play Russian roulette for $10 million once a year, for 20 years. Statistically, a few of them will make it to the end.
The media will speak of their incredible wealth. They will be famous, and heroes.
But no one will speak of all the other people that died playing.
This metaphor should help us understand how we should consider wealthy people.
We look at someone’s wealth without looking at how the wealth was obtained, which gives a false impression.
$10 million earned from Russian roulette ≠ $10 million earned being a dentist.
-> certainty is events that are likely to reproduce across different alternative histories.
-> uncertainty is the opposite.
Life is more vicious than Russian roulette, for three reasons.
- It delivers the killing bullet so infrequently that you forget it may happen. This is the black swan problem. The one event that provokes complete ruin – or death.
- “Life’s barrel” is unobservable. We can’t possibly account for every outcome, every event that “might have happened”. We can’t possibly think about all the alternative histories.
- You can’t warn someone of a loss that didn’t happen. Eg: you can’t tell someone that won at Russian roulette that what they did was dangerous if they don’t know they were playing Russian roulette in the first place. Applied to business, you can’t warn people (that are unaware of the situation they are in) that they could have lost it all.
Alternative histories don’t exist to make prophecies or help you “win”. They exist to find out the black swan, and see how likely the result you obtained was in the first place.
In general, we are not wired to estimate risk and probabilities.
People will more readily sign up for insurance against a specific event less likely to happen, than against an abstract event more likely to happen.
Risk detection and risk avoidance do not come from the rational part of the brain, but the emotional one.
Chapter 3: A Mathematical Meditation on History
The author developed a range of tools that help him think about risk.
Let’s have a look at them:
Alternative sample paths: these are the alternative histories we talked about. They are not alternatives in terms of outcome (the result), but in terms of path (what you do to reach the outcome).
If you buy a stock today, and decide to sell it in one year, this tool is interested in all of the prices this stock will have during the year, and that may influence your behavior.
Random sample path: succession of virtual historical events, starting at a given date and ending at another, subjected to some varying level of uncertainty.
Stochastic processes refer to the dynamics of events unfolding over the course of time.
Monte Carlo generator: a simulation tool that simulates anything. We can use it to determine how many times such or such outcome would happen. It’s a tool that enables you to “learn from the future” by looking at the outcome of different paths.
It’s not natural for people to learn from history. They always think “this time, it’s different”, or “this won’t happen to me”.
This is why the economy booms then busts. People don’t learn. The cycle repeats itself.
The denigration of history and its lessons may be called some form of historical determinism. Historical determinism asserts that whatever happened was “bound to happen”. Since people indeed, don’t learn from history, we may assume that history is forced to repeat itself.
When we look at the past, it seems easy in hindsight to predict the events that happened. A lot of people today said “we should have expected and prevented 9/11”.
This is the hindsight bias.
They look at the event with the information they learned after the event happened. As a result, predicting it seems rather easy.
-> a mistake is not something to be determined after the fact, but in the light of the information until that point.
This means that you should spot a mistake before you make it – there is no point otherwise.
So, why do people keep on looking at history as if it could have been predicted, while it couldn’t?
Because our minds are not designed to understand how the world works, but to find solutions to a problem quickly.
Now, another bias. People that find themselves good at predicting the past, will also find themselves good at predicting the future (while they can’t).
Hence, we never learn.
Unlike hard sciences (physics, chemistry), history cannot lend itself to experimentation. Overall, in the medium to long term, history will deliver most of the possible scenarios.
The idea is ergodicity. It means that under similar conditions, long sample paths would resemble each other (the equivalent is “history is a continuous cycle”).
In the long term, everyone becomes the average of all of his alternative histories (luck does no longer plays, crushed by time).
Distilled Thinking on Your Palmpilot
What’s the difference between noise and information? Noise has more randomness.
Same thing when we look at the difference between journalism and history.
Journalists should look at the world like historians. Instead of saying “the market is up”, they should say “the market is up but it’s not super relevant as it is coming mainly from noise”. Meaning: the market is up, but it was unpredictable (random).
Mathematically, progress means that new information is more valuable than old information. When in doubt about the value of new info, it is always better to reject it.
Why? Let’s take tech as an example.
The mainstream narrative says that tech changed our lives. But it’s not obvious. When you look at the number of tech patents, only a few of them changed our lives.
But of course, we only see and count the ones that actually worked out (telephone, airplane, etc).
We don’t see all of the tech that we never used.
Because the tech that worked has had a positive impact, people say that that tech is de facto good.
However, the opposite may be true. The opportunity cost of missing the positive tech is small compared to all of the “bad” tech one has to go through to get there.
The same can be said of information. The amount of bad information (noise) one has to go through to find the good one isn’t worth it. Bad information is not only useless, it’s toxic.
It was demonstrated by Shiller. He wrote a paper where he said that markets swung too much in relation to the value they were supposed to indicate. They were too high or too low to pretend to be efficient.
If markets weren’t efficient, it meant that where they stood daily was irrelevant.
Philostratus In Monte Carlo: On The Difference Between Noise And Information
What’s the difference between noise and information?
Imagine a dentist that invests in the stock market in his free time. He is a good investor and earns yearly 15% with a 10% error rate (volatility).
That 15% is over one year. But a lot of things can happen in one year.
When you translate the probability that his portfolio is up broken down into periods of time, this is what happens.
Every second, the dentist has 1/2 chance to see his portfolio going negative.
But if he looks at his portfolio once a year, for twenty years, he will see green 19/20! Not bad.
What does it mean?
- Over the short term, the dentist is not looking at the returns of his portfolio, but at the variance.
- Our emotions don’t understand that.
-> this explains why people that frequently look at noise aka randomness (aka the news) burn out.
Chapter 4: Randomness, Nonsense, And The Scientific Intellectual
In Vienna in the 1930s, physicists decided to update their scientific methods.
They declared that a statement could fall into two categories.
- Deductive: 2 + 2 = 4.
- Inductive: verifiable somehow: Eg: It rains in Spain.
Inductive statements are often impossible to be derived any principles from.
These two methods split intellectuals into two groups: scientific intellectuals, and literary intellectuals.
How to recognize them?
If we ask the Monte Carlo generator to produce something, it cannot repeatedly produce scientific intellectual work, as these only happen in fixed, certain conditions. But it can at will reproduce literary intellectual work (work whose results are similar in different conditions).