The Illusion of the Recipe: A Critique of Replicable Success

We do not learn from success. We compose it.





Individual success is not equal to its own replicability. This claim sounds almost trivial when stated plainly, yet the entire culture of entrepreneurship, self-help, and personal branding is built on the quiet denial of it. We live inside a civilization of imitation. We are told, implicitly and explicitly, that if we study the habits of the successful closely enough, adopt their schedules, mimic their risk tolerance, repeat their slogans, we will inherit their outcomes. This essay argues that this belief is not merely optimistic but structurally confused, and that the confusion has a precise philosophical anatomy worth dissecting.

Consider a thought experiment. Imagine ten thousand people flipping a coin, each calling heads or tails at random, ten times in a row. Probability guarantees that a small number of them, perhaps nine or ten people, will call every single flip correctly. If we gathered those few and asked them to explain their method, they would not shrug and say luck. Human beings do not tolerate meaninglessness well. They would offer an account: a particular way of holding the coin, a breathing rhythm before each toss, a personal philosophy of confidence under uncertainty. And the account would be coherent. It would even be sincere. The tellers would believe it themselves. But coherence is not causation, and sincerity is not evidence. What we would be witnessing is not a method that produces success, but success that has retroactively produced a method.

This is the structure we must learn to see whenever we encounter a successful individual explaining themselves. The explanation is real as a psychological event and worthless as a causal law.

The first and most important philosophical distinction to install here is the one between a necessary condition and a sufficient condition. A necessary condition is something that must be present for an outcome to occur, without guaranteeing that the outcome will occur. A sufficient condition guarantees the outcome whenever it is present. Waking at five in the morning, dropping out of university, working through the night, taking on irrational risk, none of these are sufficient conditions for building a great company. At most, in specific cases, they might have been necessary background conditions, present but not causally decisive. The entire genre of imitation-based self-improvement rests on a silent and unexamined conversion of necessary into sufficient. It takes a condition that, at best, needed to be there, and treats it as a condition that, by itself, produces the result. This is not a minor logical slip. It is the central mechanism by which admiration curdles into superstition.

The second structure worth naming carefully is what statisticians call survivorship bias, though the philosophical content of the idea is older and deeper than its statistical packaging suggests. During the Second World War, engineers studying returning bomber aircraft wanted to know where to add armor. They examined the planes that came back and found bullet holes concentrated on the wings and tail. The instinct was to reinforce those areas. A statistician named Abraham Wald pointed out the error: the aircraft in front of them were the ones that had survived being hit in those places. The planes hit in the engine and cockpit were not available for study, because they had not returned. The data was not missing at random. It was missing precisely where the truth was.

Every celebrated founder, every quoted visionary, is a returning aircraft. We study their wings because their wings are what we can see. We do not study the engines and cockpits of the thousands who attempted the identical gesture, the same risk, the same audacity, and were shot down entirely, leaving no memoir, no interview, no biography. The comparison we are implicitly asked to make, between the one who succeeded and the version of ourselves who might succeed by copying them, is statistically incoherent, because it excludes by construction the very population that would tell us whether the behavior in question helps or harms. We are not comparing success to failure. We are comparing success to silence.

There is a third layer, subtler than the first two, which concerns not statistics but narrative psychology. Nassim Nicholas Taleb has written extensively on what he calls the narrative fallacy, the tendency of the human mind to construct a simplified story out of a sequence of facts and to mistake the elegance of that story for an understanding of its causes. The danger of narrative is that it is built backward. We know the ending before we assemble the plot, and every fact gets selected and weighted according to how well it serves a conclusion we already possess. A founder's early failure becomes, in retrospect, a necessary trial. A stubborn refusal to listen to advisors becomes, in retrospect, visionary independence. Had the company failed, the very same failure would be filed as an inability to learn, and the very same stubbornness would be filed as arrogance. The facts do not change. Only the story changes, because the story is a servant of the outcome, not a cause of it.

Here is the sentence I want to leave at the center of this essay, the one worth remembering after the argument fades: 
We do not learn from success. We compose it.
This is the deepest reason why success resists replication. It is not simply that we lack information about the failures, though we do. It is that even our information about the successes is not really information. It is retrospective literature dressed as method.

There is a fourth layer, and it is the hardest to see because it does not depend on dishonesty, selective memory, or the desire to look good. It depends on a limit built into cognition itself. When a mind is absorbed in genuine problem-solving, most of what is happening is not language. The nervous system is busy encoding, weighting, and re-weighting an enormous number of contingencies into something closer to a skill than a sentence. Michael Polanyi named this condition with the phrase that has since become inescapable in the philosophy of knowledge: we know more than we can tell. Gilbert Ryle drew the same line from a different angle, distinguishing knowing-how from knowing-that, and observed that a person can possess the first in abundance while being nearly incapable of producing the second in any usable form. Hubert Dreyfus, studying how novices become experts, found the same pattern in every domain he examined, from chess to piloting: expertise, once achieved, stops running through explicit rules altogether and starts running through a fast, embodied, largely inarticulate pattern recognition that resists translation back into instructions, because it was never assembled out of instructions in the first place.

This is a different problem from the narrative fallacy, and it is worth being precise about the difference, because the two are easy to collapse into one another. The narrative fallacy describes a distortion: a story gets bent, after the fact, to fit a conclusion it did not really produce. What is being described here is not distortion but untranslatability. Even a founder with no interest in flattering themselves, a founder answering with total sincerity and total humility, would still produce an account that leaves out almost everything that mattered, because the part that mattered was never stored in a form that language can retrieve. The general principle they offer, work hard, trust your instinct, stay obsessed, is not a lie. It is a true but almost empty description of a process whose actual content was procedural, contextual, and largely below the threshold of conscious articulation. We can understand the principle. We cannot, from the principle alone, reconstruct the ten thousand small adjustments the nervous system made in response to conditions that no sentence captured, because no sentence was generated at the time they were made. The recipe is not hidden out of vanity or forgetfulness. It is hidden because it was never written, not even in the head of the person who executed it.

A fifth and final layer requires us to leave probability and narrative behind and enter something closer to historical philosophy. Success does not occur in a vacuum. It occurs inside an irreducible lattice of timing, geography, capital availability, regulatory gaps, cultural readiness, and interpersonal networks that existed in that configuration exactly once. To imitate the visible action of a historical agent while ignoring the invisible scaffolding that supported it is to mistake the foreground for the whole painting. A young entrepreneur in a garage in the nineteen seventies operated inside a specific configuration of cheap real estate, a nascent industry with almost no incumbents, capital that was naive rather than sophisticated, and a labor market willing to accept enormous personal sacrifice for equity that had not yet been proven to be worth anything. None of these conditions can be summoned by an act of will in a different decade. The gesture can be repeated. The world in which the gesture was meaningful cannot.

This is not historical determinism in any crude sense. It is not a claim that individuals have no agency or that choices do not matter. It is a claim about the ontology of an event, that an event is not a discrete, portable unit that can be extracted from its context and reinserted elsewhere while retaining its properties. A decision to take a particular risk, at a particular institutional moment, is not the same object as an identical decision made forty years later inside a saturated, regulated, and already-mapped industry. The action looks the same from the outside. Its causal weight is entirely different, because causal weight is not a property of the action alone. It is a property of the action in relation to everything around it.

With these five structures in view, the necessary and sufficient confusion, survivorship bias, the narrative fallacy, the opacity of tacit knowledge, and the historicity of the event, we are in a position to close off, one by one, the justifications that inevitably arise whenever this argument is presented. This closing off is not a rhetorical flourish. It is the actual philosophical labor the essay must perform, because an argument that persuades in the abstract but leaves its objections unaddressed persuades no one who is actually invested in the belief being criticized.

The first objection is the simplest: but this person really did succeed using this method. This is true, and it is also irrelevant to the claim being made. The truth of a single historical case does not establish a general law. This is the ancient problem of induction, given its sharpest formulation by David Hume: no finite number of observed instances, however consistent, logically entails a universal rule about future or different instances. One true case tells us that the method was compatible with success in that instance. It tells us nothing about whether the method is a reliable cause of success in general, especially when we have already established, through survivorship bias, that we are only looking at the compatible cases.

The second objection follows closely: but most successful people share this trait. Correlation across a visible sample is not causation, and the objection collapses the moment we ask the comparative question that is almost never asked, which is what percentage of failed attempts also shared the exact same trait. If early risers fail at the same or higher rate than late risers, waking early is not a differentiator, it is simply a common feature of an entire population of ambitious people, most of whom fail regardless. A trait shared by both the successful and the unsuccessful cannot be doing causal work in explaining the difference between them.

The third objection is more sophisticated: but the reasoning behind the method sounds logically sound. This objection mistakes coherence for validity. A story can be internally consistent, even persuasive, without corresponding to the actual causal structure of the world. Logical elegance is a property of narrative construction. It is not evidence. The most dangerous stories are precisely the ones that sound the most reasonable, because their reasonableness disarms scrutiny.

The fourth and final objection is usually delivered with a note of exasperation: if you do not imitate someone who already succeeded, how are you supposed to succeed at all? This objection performs a subtle substitution that must be named directly. The critique in this essay is not directed at ambition, risk-taking, or the will to attempt difficult things. It is directed specifically at the belief that copying the surface behavior of a successful individual reproduces the underlying causal conditions that supported them. Rejecting the recipe is not the same as rejecting the kitchen. One can commit fully to building something ambitious while refusing to believe that a specific sequence of borrowed habits is what will determine the outcome.

Before closing, one further complication deserves to be named, because without it the critique risks smuggling in the very assumption it set out to dismantle: the assumption that success is a single, continuous quantity, arrayed along one axis, on which some people simply sit higher than others. It is not. Success is always success relative to a chosen unit of measurement, and the unit is rarely made explicit, which is precisely what allows it to be swapped silently whenever convenient. Alasdair MacIntyre's distinction between the internal goods of a practice, the excellence a discipline produces on its own terms, and the external goods that surround it, the wealth, fame, and institutional power that may or may not follow, is useful here because it exposes how much work the unit of measurement is doing before any comparison even begins. Measured by market capitalization, media presence, or cultural gravity, an entrepreneur may appear to eclipse the scientists whose foundational work made his entire industry conceivable in the first place. Measured by depth of contribution to the structure of human understanding, the ranking does not merely shift, it inverts, and inverts without much controversy among those who work inside the relevant discipline. Neither ranking is wrong. Each is simply answering a different question while using the same word, success, to disguise the fact that the question changed. This is not a claim that all achievements are secretly equal, a leveling move that would be its own kind of dishonesty. It is a claim that the comparison our culture performs constantly and almost unconsciously, between the entrepreneur and the researcher, the founder and the artisan, the visible and the foundational, is not a comparison of magnitudes at all. It is a comparison across incommensurable scales, dressed up as a single leaderboard because a single leaderboard is easier to sell, easier to envy, and easier to imitate than a plural, honest account of what excellence looks like inside a hundred different practices that share no common metric.

It would be intellectually dishonest, and also less interesting, to end on pure negation. If the specific, imitated behavior is not what travels from one success to another, something less visible might still be worth cultivating, provided we are precise about what it is. What seems to generalize, cautiously and without guarantee, is not a behavior but a disposition: the capacity to tolerate prolonged ambiguity, the willingness to revise a belief quickly when reality contradicts it, a relationship to failure that treats it as information rather than identity. These are not formulas either, and they come with no promise attached. But they are, at least, honest about their own uncertainty, which the imitated habit never is. A habit borrowed from a biography presents itself with false confidence, as though causality had been settled. A disposition, honestly described, admits from the outset that it is necessary at most, and never sufficient.

This is perhaps the only responsible place for such an essay to arrive. Not at cynicism, which would counsel paralysis, and not at the comforting fiction that success is a system waiting to be decoded, but at a mature discomfort. We are creatures who prefer causal stories to statistical humility, and the market for those stories, in business literature, in interviews, in commencement speeches, will never disappear, because it answers a psychological need that has nothing to do with truth. The task is not to eliminate our appetite for these stories. It is to read them as literature rather than as instruction, to admire the survivor without mistaking the survivor's memory of the storm for a map of it.
___________________________________________________________________________________

Individual success, then, is not a method waiting to be extracted and applied elsewhere. It is a singular event, produced by the unrepeatable intersection of timing, context, chance, and a disposition that can be described but never manufactured on demand. What can be carried forward is not the recipe. It is, at most, the willingness to cook without one.

Post a Comment