The premise behind most “secret of success” advice is that a single trait, talent, effort, grit, discipline, explains why some people rise and others do not. A 2018 paper in Advances in Complex Systems, by the physicists Alessandro Pluchino, Alessio Biondo, and Andrea Rapisarda, built a simple model to test whether that premise actually holds up against a basic statistical puzzle: talent in a population is normally distributed, clustered around an average with a predictable bell curve, while wealth, a common stand-in for success, follows a highly skewed power-law distribution, with a small number of people vastly outperforming everyone else. If talent alone explained success, the two distributions should look more alike than they do.
Correction (September 16, 2026): The earlier title and summary incorrectly described the simulated workers as equally talented and the winner as 650 times luckier than everyone else. Talent was normally distributed, and the roughly 650-fold figure described success relative to the modeled average.
The model simulated 1,000 agents, each assigned a fixed level of talent drawn from a normal distribution, then tracked over a working life exposed to random lucky and unlucky events. Across 100 runs of the simulation, the single most successful agent recorded was not the most talented. It was an agent with talent almost exactly at the population average, whose final success came out roughly 650 times higher than the average performer, entirely because that agent happened to encounter more lucky events along the way. The researchers describe talent as a necessary condition for reaching the highest levels of success, but not remotely a sufficient one on its own.
What this simulation is, and what it is not
This is one modeling paper, not an empirical measurement of any real career or real population, and its authors built it specifically to illustrate a statistical mechanism rather than to track actual people through actual working lives. The result should be read as a demonstration that a talent-only explanation cannot mathematically account for how skewed real-world success distributions look, not as direct proof that any specific successful person’s story is “mostly luck.” But the mismatch it identifies, a normally distributed input producing a power-law distributed outcome, is a real and well-documented feature of income and wealth data, and the simulation offers one clean, quantified explanation for how that mismatch could arise even in a world where talent is genuinely rewarded.
A second popular “secret” that has not held up well under scrutiny
Talent is not the only individual trait that has been sold as the secret ingredient. Grit, defined as sustained passion and perseverance toward long-term goals, became one of the most widely promoted answers to the success question over the past decade. A 2017 meta-analysis in the Journal of Personality and Social Psychology, by Marcus Credé, Michael Tynan, and Peter Harms, pooled 584 effect sizes from 88 independent samples covering more than 66,000 people and found that grit correlates so strongly with the existing personality trait of conscientiousness, at a level the researchers describe as close to complete overlap, that it functions largely as conscientiousness under a new name rather than as a distinct predictor of success in its own right.
This is one meta-analysis, not the final word on every claim made about grit, and the researchers did find that one specific facet of grit, perseverance of effort, predicted outcomes like academic performance even after accounting for conscientiousness. But the overall picture complicates the popularized version of the concept considerably: grit as a freestanding, learnable secret to success has less independent scientific backing than its enormous popularity would suggest, because most of what it measures was already captured by an existing, well-established personality trait.
Why both findings point in the same direction
Read together, the talent-versus-luck model and the grit meta-analysis complicate the “secret of success” premise from two different angles. The Pluchino model shows that even granting talent a real and necessary role, statistically it cannot explain the shape of real success distributions without randomness doing a large share of the work. The grit research shows that a trait marketed as a distinct, learnable secret is substantially the same thing psychologists already had a name for, and that its added predictive value beyond that existing trait is real but considerably smaller than the popular framing implies. Neither finding says individual qualities do not matter. Both say that a single, discoverable “secret,” whether framed as talent, grit, or any other single trait, is a poor fit for what the data on success actually looks like.
What this suggests, without overcorrecting into fatalism
None of this research supports concluding that effort, skill, or persistence are irrelevant, or that success is purely random and therefore not worth pursuing deliberately. The Pluchino model explicitly treats talent as necessary, just not sufficient, and the grit research still found a real, if modest, independent contribution from perseverance of effort. What both lines of research undercut is the specific promise embedded in “the secret of being successful”: that one identifiable ingredient, once found and applied consistently, reliably produces outsized outcomes. The more accurate, less marketable picture involves a real floor set by ability and effort, with a much larger share of the variance in outcomes at the very top explained by exposure to opportunity and chance than either popular self-help framing or many successful people’s own retrospective accounts tend to acknowledge.
What this does not prove
The Pluchino model is a simplified simulation built to illustrate a statistical principle, not a direct measurement of any real labor market, industry, or country, and its specific numeric results, like the 650-fold advantage of the luckiest average-talent agent, are properties of this particular model rather than an empirically measured real-world figure. The Credé meta-analysis, while large, drew primarily on academic and workplace samples, and it is not established how its findings extend to more unusual or extreme forms of success, such as founding a company or achieving fame, that were not the meta-analysis’s focus. Neither study offers a replacement “real secret” to substitute for the ones it complicates. What they jointly support is a narrower and more defensible claim: extraordinary success is very unlikely without real ability and sustained effort, but the leap from ordinary success to extraordinary success appears to depend heavily on factors well outside any single trait a person can simply decide to cultivate.