Against Certainty

A founder takes the stage and explains why they won: vision, focus, the nerve to ignore everyone who said no. It is a good story, and the room takes notes. What the story leaves out is not a lie — it is just everything that did not fit on the slide. The timing. The one customer who happened to say yes. The competitor who moved too slowly. And the dozens of people who did the very same thing, with the very same nerve, and are not on any stage at all.

Success is real. It is also weak evidence about why it happened. The win you can see. The reasons you mostly infer afterward, from a sample of exactly one — and the mind doing the inferring has a strong, well-documented preference for one particular answer: me. This is the genius trap: not that successful people have no skill, but that success is a terrible witness to how much of it there was.

Start with the thing you cannot see.

The cemetery you don't see

Nassim Taleb has a name for the person who mistakes a lucky outcome for a skilled one: the lucky fool, "a person who benefited from a disproportionate share of luck but attributes his success to some other, generally very precise, reason." The problem is structural. Gather enough people taking enough risk and a few will win big on chance alone. "If there are enough players," he writes, "say thousands of 25-year-old players, we can expect to see a handful of (extremely rich) survivors — and a very large cemetery."

We only ever interview the survivors. The cemetery does not give talks. The cleanest illustration is a wartime one. In World War II, the statistician Abraham Wald was asked where to add armor to bombers, based on the damage patterns of the planes that came back. The intuitive answer was: wherever the bullet holes cluster. Wald's answer was the opposite — armor the places with no holes, because the planes hit there were the ones that never returned. The data was made entirely of survivors, and the survivors were quietly hiding the lesson.

The same distortion runs quietly through every track record you admire. About half of new U.S. businesses are gone within five years — the founders who didn't make it are not writing memoirs. In investing, the bias is baked into the numbers themselves: once you count the funds that quietly closed, one careful study found a hedge-fund index's reported returns were overstated by roughly 3.7 percentage points a year. Nobody is lying. The winners are simply the only ones still in the room to be asked.

The better you get, the more luck decides

There is a stranger problem, and it sharpens exactly as skill improves. The investor Michael Mauboussin calls it the paradox of skill. As everyone in a field gets better and more alike, the spread in skill narrows — and when the players are nearly equal, the small random differences, the luck, decide more of the result, not less.

Greater skill doesn't decrease the dependence on luck. It increases it.

Baseball is his example. In 1941, Ted Williams hit .406; no one has hit .400 over a full season since. Not because hitters got worse — they got better. As the whole league's skill converged, the spread in batting averages shrank (the standard deviation fell from .0326 in the 1940s to .0274 in the 2000s), and the towering outlier seasons simply stopped happening. "Hitting .380 in 2011," Mauboussin notes, "is the equivalent to the .406 that Ted Williams hit" seventy years earlier. The top of a skilled, crowded field is where luck does its loudest work — which means the more competitive your win, the more humble its explanation should be.

Can you lose on purpose?

So how do you tell, from the inside, whether a result was skill or luck? Not from one result — but there are two tells. The first is Mauboussin's quick test: ask whether you could lose on purpose. In chess or tennis you can throw a game deliberately; that is a sign skill is present. At roulette you can no more lose on purpose than win on purpose — it is pure luck. The second tell is repeatability. Skill persists across tries. Luck does not.

Daniel Kahneman once got to measure this directly. A wealth-management firm let him check whether its advisers' results held up from year to year — the fingerprint of real skill. He ranked twenty-five advisers across eight years and correlated each year with the next. The average correlation was .01. Effectively zero. "The results resembled what you would expect from a dice-rolling contest, not a game of skill," he wrote. The firm was "rewarding luck as if it were skill" — and paying large bonuses for it. The illusion of skill, he added, "is deeply ingrained in the culture of the industry."

Why winners are so sure

Now add the part that happens inside your own head. People reliably credit their successes to themselves and their failures to circumstance. It is one of the most replicated findings in social psychology: a meta-analysis pooling 266 studies put the effect at d ≈ 0.96, which is large. And it is strongest — this is the part worth sitting with — in precisely the people most likely to be reading a success story: the non-depressed, the Western, the already-succeeding.

It rides on a broader overconfidence. Ask people to rate themselves against the average and the arithmetic breaks: in one classic study, 93% of American drivers rated themselves more skillful than the median driver — which cannot be true, since half of everyone is, by definition, below the median. Success pours fuel on this. Win once, and the mind hands you a clean, flattering story about why, and you will badly want to keep it.

It would be easy here to reach for the Dunning-Kruger effect — the idea that the least able are the most overconfident. We won't lean on it, because the famous curve is partly a statistical artifact: plot self-rating-minus-score against score and you get a downward line out of pure noise. The robust, honest claims are the two above — most people rate themselves above average, and confidence routinely outruns accuracy. A site called Against Certainty owes its own evidence the same scrutiny it asks of yours.

Resulting, and the better question

The poker player Annie Duke has a word for the core mistake: resulting — judging the quality of a decision by the quality of its outcome. A good decision can lose; a bad one can win; over a single try the link between the two is loose. Pete Carroll's goal-line pass call at the end of Super Bowl XLIX was branded the worst call in football history — but only because it was intercepted. Of the sixty-six passes thrown from the opponent's one-yard line that NFL season, none had been picked off. The call was defensible. The dice were not.

So the better question is never "did it work?" It is "would the same decision still have been good if it had failed?" If only the win makes the call defensible, then it was never the call that was good. It was the world, voting in your favor, on a ballot you didn't control.

How to read your own track record

None of this is an argument that you were lucky and not good. That would just be resulting in reverse — reading skill off a single outcome, then erasing it off the same one. The move is the one this whole site keeps returning to: hold the belief as strongly as the evidence allows, and no stronger. Credit yourself for the part you could repeat, controlled, and would still defend if it had lost. Hold the rest — the timing, the open doors, the silent cemetery — at arm's length.

We built a small thing to make this visible: the Cemetery. A thousand people make the same bet; you set how much they actually differ in skill, and watch a lucky handful reach the stage and narrate their genius while the identical failures stay silent. Run another season and this year's geniuses revert to the crowd. Its companion piece, Defensible If It Loses, runs the same logic outward — on the decisions of the people we are quick to crown geniuses or fools.