Is it Luck or Skill? | Feb '26

It has been 4 months since my last entry. My instincts would tempt me to invoke the outlier — that is, I was suddenly preoccupied with a deluge of unpredictable priorities, and hence absolved of blame — but I suppose it was just my unyielding tendency to procrastinate. Yet in hindsight, this very habit has become a lesson, one that I have reached after ingesting a few books in the epistemology realm. Thinkers like Nassim Nicholas Taleb, Howard Marks, and Veritasium producer Derek Muller, have shaped and continue to inspire my growing fascination with how we know what we know.

Psychological fallibilities exert a profound yet overwhelmingly underestimated role on decision making. Its effects are subtle. We are all susceptible, and overcoming them entirely may be impossible. What matters then, is acknowledging their implications on our decisions. In this piece, I’ll attempt to synthesise some key takeaways from these great educators, and explore, perhaps provocatively, whether free will can truly exist when we are bound by our fallible biology. Hopefully, you’ll be able to make the connection to investment.

The Lucky Fool: Survivorship Bias

Lucky fools do not have the slightest suspicion that they may be lucky fools — that’s by definition. But it’s a phenomenon that pervades life. Is it fair to judge a doctor’s success by how many ill patients he has saved over his career? By many measures, that would be a fair conclusion. After all, though there does exist randomness in a patient’s recovery (imagine a sudden deterioration or recovery), his well being is more often than not tied to the doctor’s presence and skills. Put it this way, it’s reasonable to say that one would rather have a heart attack in the hospital than at the bottom of the Grand Canyon. But most of life exists outside the predictability of the hospital. Is it fair then to judge an investor’s success by their personal wealth, or a gambler’s success on their cumulative winnings? Very likely less so. But speak to a card counter cashing in his chips or listen to a Wall Street investor getting off the post covid rally on CNBC and he’ll tell you he knew about it all along, and drown you with the exact strategy he used to get to that point. Over time, we do believe them. Alas, we saw it with our own eyes. However, it is what we didn’t see that matters — the invisible graveyard of millions of bankrupt clones and gamblers (less that bit of luck) who used the same strategy and succumbed to the pressures of a result-seeking society. It’s not exactly convenient to interview the dead… even if there are so many of them.

Why our brains lie to us, and why we believe it: Narrative fallacy, induction and Noise vs Signal

Given an explanation of the lucky fool, it’s easy for you in retrospect to spot that lucky fool. You’ve fallen for hindsight bias, but let’s just stick to the perspective of that lucky fool for now. The lucky fool didn’t make a deliberate choice to fit explanations to his success. He hates the idea that he was successful because he got lucky (we all do), so he weaves a story of “vision” and “grit” around them. But this induction is dangerously fallible, especially if he expects (and most certainly does) this “vision” and “grit” to bring him his next pot of gold. Taleb illustrates this eloquently. Imagine yourself as a turkey. You live a comfortable life. You’re fed well and in a timely manner. From this observation you’d expect to live a long, perhaps gluttonous life, only to be proven wrong on the day before thanksgiving when your head is unexpectedly disassembled from your body. Many investors are lucky fools precisely because they have been fed by a bull market for months. They mistake the butcher’s kindness for their own “skill” at being a turkey.

Our brains did not evolve to trade stocks, calculate life expectancy, or manage global supply chains. They evolved for a very specific, high stakes environment: the African Savannah. Imagine you’re an early human and you hear a rustle in the tall grass. You have two choices:

1. Assume it’s a predator: You run away. If you’re wrong, you just lost a little energy. If you’re right, you live.

2. Assume it’s just the wind: you stay put. If you’re right, you saved energy. If you’re wrong, you are removed from the gene pool.

Evolution favoured the 1st scenario, biologically wiring us to see patterns even where they don’t exist, because in the wild, paranoia was a survival advantage. Rather instinctively then, we retrospectively impose an explanation on our success when there isn’t one, making the slaughter on thanksgiving catch us completely off guard.

Modern society reinforces these fallacies, with the proliferation of information creating the illusion of us knowing more than we actually do. It’s the classic signal vs noise debate. The more information you give someone, the more hypotheses they will formulate along the way, and the worse off they will be. The divergence between what we actually know and what we think we know increases exponentially with new information. The problem is that our ideas are sticky. Once we produce a theory, we are unlikely to change our minds — so those who delay developing theories are often better off dealing with uncertainty. It’s no wonder why economists seem to know exactly when the world will end (a recession) after 10 days of crunching numbers through complex models, even when it’s certainly unknowable.

The Environment of Risk: the game itself is rigged against our intuition

Investing isn’t a casino or a textbook. The “unknown unknowns” are what breaks the system, not the “risks” we have factored into our model. On the surface a casino seems like the perfect example to illustrate risk management, given gambling itself is a game of probability. In a casino, one would think, the risks include lucky gamblers blowing up the house with a series of large wins or cheaters taking away money through devious methods. Consequently, the casino has high-tech surveillance systems to track cheaters, card counters and other people who try to derive an advantage over them. In effect, a casino is a closed system where risks are expected, systematically managed and eliminated to a reasonable extent. According to Taleb, this makes gambling a sterilised and domesticated uncertainty. In the casino you know the rules, you can calculate the odds and the type of uncertainty we encounter. Hence these risks don’t constitute the uncertainty that casinos face. Those that do, however, have nothing to do with what can be anticipated knowing that the business is a casino. Some of the largest losses incurred or narrowly avoided by the casino fell completely outside their sophisticated models. A casino lost around $100 million when an irreplaceable performer in their main show was maimed by a tiger. (It had even considered insuring the attack against audience members but never thought to insure the performer himself) A disgruntled contractor was hurt during the construction of hotel annex, and was so offended by the settlement offered that he made an attempt to dynamite the casino. These off-model hits by black swans swamp the on-model risks by a factor of close to 1000 to 1.

How we think about risk: Ergodicity and the facade of averages

We like to think of risk in the form of expected value. Often, that’s because we don’t have skin in the game. When you’re not the one bearing the downside, averaging outcomes feels intellectually sufficient. More likely, it’s because expected value gives us the comforting illusion that a messy, complex reality can be reduced to a single clean number. This tendency to simplify blinds us to ergodicity: the assumption that the average outcome of a group is the same as the average outcome of one person over time. In some cases that is true: a coin flip, a dice roll, Brownian motion…

Yet far more instances in life are not ergodic. Picture a group of 100 different people playing one round of Russian Roulette for $1 million each. Statistically, about 83 people will walk away with $1 million, and 17 will be dead. If you’d look at the average wealth of the group, it’s a massive success, since the average person made roughly $830,000. Now, imagine you played Russian Roulette 100 times in a row. Do you end up with $83 million? No. You end up dead. Your probability of survival is effectively 0. The expected value of the group is irrelevant to you because you cannot average your way out of being dead.

That is exactly the world of leveraged investments. A trader might use a high-risk strategy that works 99% of the time. For five years, he looks like a genius. He has “skill”. But if he keeps playing, that 1% risk of total loss will lead him to default eventually. In this non-ergodic system, to succeed, you must first survive.

Synthesising the “Free Will”: The Problem of Silent Heroes

If we accept that we are turkeys lulled by a friendly butcher, and that our skill is often just the residue of a non-ergodic lucky streak, we must confront the problem of silent heroes. The catastrophe that did not happen generates no headline, no bonus, no applause. A legislator who strengthens cockpit doors before 9/11 cannot point to the lives saved. The success is invisible. We are biologically and culturally blind to the catastrophe that didn't happen.

It reveals a fundamental flaw in how we attribute merit. We reward the visible recovery — firefighters rescuing victims, legislators passing laws, military commanders on their Iraqian holiday — but ignore the invisible prevention. In the hospital, the doctor’s skill is visible because he has skin in the game. The feedback is immediate, physical, and non-negotiable. But in the randomness of our broader society, the “experts” on television screens have no skin in the game. They can be wrong indefinitely and still be invited back to speak, their “luck” explained away by narrative fallacy.

We arrive at a provocative crossroad. If our successes are often masqueraded luck, and our greatest acts of skill, prevention, are invisible even to ourselves, where does Free Will actually reside? I would argue that Free Will exists only as a deliberate, painful awareness of our own fallibility. I urge everyone to embrace the humility of not knowing something, to resist the temptation to over-credit themselves for success, and to learn to take quiet satisfaction in “non-events. Skill tends to reveal itself as the humility to build a system that is robust enough to survive being wrong.

In a more practical sense, our agency lies not in forecasting the next shock or quantifying uncertainty. It lies in structuring our exposure such that no single shock can permanently remove us from the game. For the casino, that would mean maintaining operating reserves, purchasing catastrophic insurance, and deferring managers’ compensation so that decision-makers personally bear the downside.

For investing (and life more broadly) it means avoiding leverage that can trigger forced liquidation, diversifying across uncorrelated risks, keeping liquidity buffers, and favouring strategies with limited downside and open-ended upside. It means optimising for survival rather than maximum expected return.

Nevertheless, social incentives create a persistent paradox. In environments compromised by cognitive biases, restraint will be viewed as underperformance. The investor who forgoes marginal gains to preserve optionality appears timid in calm markets. He is ignored and invisible. The investor who “takes risks” is showered with praise, capital and opportunity. Yet, it's almost certainly too late when his luck runs out in the face of the improbable. In a non-ergodic world, our agency is a deliberate choice to sacrifice short term gratifications for long term survivability, to remain in the game long enough for favourable randomness to matter. And to that end, humility is the first step.

Some Closure:

To bring this back to my own four-month silence, perhaps my procrastination was not my lack of discipline, but a subconscious pause for me to digest these existential ideas. I may have inadvertently played the role of the silent hero in my own learning by avoiding the catastrophe of writing a piece that lacked the depth these four months provided. But I could just as easily have made that up… I don’t know!

Anyways, I highly encourage everyone to read Taleb’s Incerto collection of books. It has contributed invaluably to my own intellectual development, and I hope it will for yours as well.

Disclaimer: if you read closely, this entire piece might have read like an embodiment of the narrative fallacy. Why should you trust my “pattern recognition”, my “synthesis”, my “narrative”? You shouldn’t. That’s precisely the point. I present these arguments (many inherited from Taleb) not as universal truths. I do not know whether they are correct, but I do know that we are systematically biased, and that our confidence routinely exceeds our understanding. The rest must be empirically tested against reality… that’s where you come in.