Survivorship Bias

Examples

The classic case comes from World War II. Analysts studied returning bombers to decide where to add armor, and the bullet holes clustered on the wings and fuselage. The statistician Abraham Wald pointed out the flaw: these were the planes that made it home. Holes in the engines were missing from the data because those planes were missing — at the bottom of the sea. The armor belonged where the surviving planes were clean.

The modern version is everywhere:

Post: “College is a scam. Gates and Zuckerberg dropped out and became billionaires.” Reply: “How many dropouts didn’t? They don’t get documentaries.”

Or in everyday health talk:

Uncle: “My grandfather smoked a pack a day and lived to 95.”

Grandfathers who smoked and died at 60 aren’t at the dinner table telling their side — a one-case cousin of the anecdotal evidence fallacy.

Why it happens

You can only learn from the cases you can see, and the world filters which cases reach you. Failure is quiet: failed startups don’t give keynotes, deleted apps don’t get reviews, bankrupt funds vanish from the performance tables. Nobody has to lie for the visible sample to be wildly unrepresentative — the selection does the distorting on its own, which is why this bias survives in fields, like finance, that measure it formally.

Add our habit of studying winners in order to copy them, and the trap closes: the harder you look at survivors, the more confident and the more wrong you can become.

How to counter it

• Ask for the denominator: “Out of how many who tried?”
• Hunt the missing cases — the dropouts, the closed shops, the quit users. What did they do? If it’s the same things the winners did, the “secret of success” explains nothing.
• Beware advice reverse-engineered from winners alone, including your own success stories.
• When someone cites a survivor, be charitable: the example is real; the sample is what’s broken. Say so in those words.