Luck vs. Skill: The Survivors You See

With enough random players, someone always gets a hot streak — and you only ever hear from the winners.

Finance, Trading & Markets · Lesson 22 · 9 min read

A stock-picker has beaten the market five years in a row, and now he’s selling a course on his “system.” Five straight years sounds like undeniable skill — proof he’s cracked the code. But here’s the question almost nobody asks: out of thousands of pickers, how many would rack up a five-year streak by pure luck — and would you ever even hear from the ones who didn’t? Hold that. The answer reframes nearly every “genius” track record you’ll ever see.

You only see the survivors

Survivorship bias is the error of judging a situation only by the visible winners, because the losers have quietly disappeared from view. We hear about the lottery winner, the lucky stock-picker, the startup that made it — almost never about the vastly larger crowd who tried the same thing and failed.

Because the failures are invisible, success looks far more common, more achievable, and more skill-driven than it really is. The data you naturally see is pre-filtered to show you only the survivors.

Randomness manufactures “geniuses”

With enough players, pure chance guarantees some impressive streaks. Picture 3,200 people each guessing the market every year, a coin-flip 50/50. After one year, about half guess right; after five years, a small fraction will have been right all five years straight — by luck alone.

Those lucky survivors look like geniuses, write books, and sell courses. They genuinely believe in their “system,” and so do their followers — but they’re the coin-flip winners, not proof a winning system exists. This is why a hot streak, on its own, can’t distinguish luck from skill: somebody was always going to win the lottery of randomness.

Do the math on the “5-years-in-a-row genius.” If each year is a 50/50 coin flip, the chance of being right five years straight is (1/2)⁵ = 1/32. Out of 3,200 random pickers, you’d expect about 3,200 ÷ 32 = 100 of them to achieve a perfect five-year streak by chance alone. A hundred “geniuses,” zero skill — and those 100 are exactly the ones you’ll hear from.

Telling luck from skill

You can’t banish luck, but you can ask better questions before crediting skill:

• Sample size — one big win or a short streak is weak evidence; randomness produces those easily.
• Repeatability — does the performance persist over long periods and many independent tries?
• A plausible mechanism — is there a real, explainable reason it should work, or just a story fitted after the fact?

Beware the gambler’s fallacy too — a coin isn’t “due” for tails after a run of heads; independent odds don’t remember the past. The honest stance is humility: in a world this noisy, much of what looks like skill is luck wearing a confident smile. (Education to think clearly, not advice on picking anything.)

Sizing up the picker selling his “system”:
1. His pitch: five years of beating the market proves skill.
2. Reframe with survivorship bias: how many pickers started, and how many failed and vanished?
3. Run the numbers: with ~3,200 random pickers, about 100 would get a five-year streak by luck alone — and those 100 are precisely the ones writing courses, while the ~3,100 who failed are silent.
4. Apply the tests: tiny effective sample (one streak), no proof of repeatability, and a “system” story likely fitted after the fact. Verdict: the streak is fully consistent with luck. Maybe he’s skilled — but the track record alone can’t show it.