Thread by @EugeneNg_VCap

Sharing my 20 learnings & reflections from reading Nassim Taleb’s “Fooled by Randomness” 👇🏻

1 | Randomness vs Non-Randomness.

Differentiating luck disguised and perceived as non-luck, and more specifically randomness disguised and perceived as non-randomness.

2 | More random and more frequent.

It is more random than we think, rather than its all random.

Similarly, black swans are more frequent than we think, rather than it being very rare.

3 | Understand Probability, not Compute it.

Probability is not a mere computation of odds, it is a way of thinking, the belief in the existence of an alternative outcome

Probability is not about what may happen, but what will happen. If it has happened before, only when then.

4 | Mathematical Modelling in our minds.

In real life, it is almost never one dimensional with one sole variable, but multi-dimensional and highly path dependent.

Because of limited 3D graphical representations, it is far better to model it mathematically in our minds.

5 | Application of probability in the real world.

Probabilities work best where rules are clearly and explicitly defined, are computable, and the risks consequently measured.

But that’s not the real world. In real life, it is not a deck of cards.

6 | Understanding deviation from the norm and the role of luck.

In real life, the larger the deviation from the norm, the broader the range of outcomes, the larger probability of results coming from variance/randomness/luck rather than skills.

All don’t, but most do.

7 | Probabilities can differ across time scales.

The scaling property of randomness is generally misunderstood.

15% p.a. return + 10% p.a. volatility

➡️ ~ 93% probability of success in any given year.

➡️ ~ 50.02% probability of success over any given second.

8 | Observing an evolving combination of variance and returns.

Over the short-term, one observes the variability of the portfolio, the variance, not the returns.

Over the long-term, one eventually observes the returns and less the variance.

9 | Frequency or probability in itself is irrelevant, the magnitude of the outcome (risk of blowup) matters much more.

The frequency or probability of the loss, in and by itself, is totally irrelevant; it needs to be judged in connection with the magnitude of the outcome.

10 | Not getting killed. Don’t be a dangerous idiot, who just don’t know it yet. Staying power is significantly underrated.

Don’t want to be LTCM or the trader that eventually gets wiped out (high probability of small gains vs low probability of large loss than bankrupts).

10 | cont’d

Don’t want to be playing a game like the Russian Roulette (revolver with 1 bullet in 6 chambers) and end up being dead.

Though we might see a handful of extremely rich survivors, the cemetery of dead is very very large.

11 | Important to Understand Skewness, particularly negative skewness

Maximizing the probability of winning does not lead to maximizing the expectation from the game when one’s strategy may include negative skewness, a small chance of large loss & a large chance of a small win.

11 | cont’d

If you engaged in a Russian roulette–type strategy with a low probability of large loss, one that bankrupts you every several years, you are likely to show up as the winner in almost all samples—except in the year when you are dead.

12 | Mathematics as a tool to think, reflect & meditate.

Mathematics is principally a tool to meditate, rather than to compute.

Not all things are random, some are much more random, some less random.

Some outcomes will have higher probabilities and others have lower.

13 | Induction without empirical observations can led to devastating conclusions.

No amount of observations of white swans can allow the inference that all swans are white.

But the observation of a single black swan is sufficient to refute the conclusion of all swans are white.

14 | Theories are never right (not yet).

Because we will never know if all the swans are white, until we see a black swan.

Theories that are known to be wrong, have been tested and adequately rejected.

14 | cont’d

Theories not yet known to be wrong, are exposed to be proved wrong.

If a theory cannot be verified or disproved, it can only be provisionally accepted, but still doesn’t mean it is right.

15 | Safer to use data to reject than to confirm hypotheses

Use data to disprove a proposition, never to prove one.

Use history to refute a conjecture, never to affirm one.

Quantitatively reject by finding counterexamples.

Yet cannot accept, as no counterexamples yet.

16 | Only qualified comments truly matter.

Unless the source of the statement has extremely high qualifications, the statement will be more revealing of the author than the information intended by him.

16 | Only qualified comments truly matter (cont’d)

Wittgenstein’s ruler: Unless you have confidence in the ruler’s reliability, if you use a ruler to measure a table you may also be using the table to measure the ruler.

17 | The Role of Randomness.

Understand the role of randomness.

Know that there will always be randomness, question of more or less.

Don’t attribute your successes to skills, but your failures to randomness.

18 | Differentiate between Skill and Randomness

Know the distinction between those skills that are visible (dentist) vs less visible (investor), where the latter belongs to a randomness-laden profession.

19 | Degree of Randomness matters, and over time as well.

The degree of randomness in such an activity and our ability to isolate the contribution of the individual determine the visibility of the skills content.

20 | Repetitiveness helps to confirm if skill is present.

Thinking ergodically & deeply about repetitiveness. It often reveals if skills exists in many long-term paths.

Context of history matters, understand the ins and outs of randomness, the invisible histories.