What it is
Studying only the cases that made it through, so the failures never enter the sample.
In the wild
Every successful founder dropped out, so dropping out helps.
How to answer it
Ask where the failures went, and count them too.
Evidence and sampling
Survivorship bias
Also called Selection on the outcome
In depth
Origin and naming
The classic illustration is Abraham Wald at the Statistical Research Group during the Second World War: asked where to add armour on bombers, he studied the holes on planes that came back and armoured the spots that showed none — those were the planes that never returned. The term 'survivorship bias' now covers any sample that filtered out its own failures.
More places it shows up
A profile of unicorn founders notes they all worked eighty-hour weeks — with no column for the founders who did the same and folded.
A music magazine asks arena headliners how they made it and takes 'just keep playing anywhere' as career advice.
A property seminar interviews landlords who bought cheap in 2009 and generalizes their timing into a law of markets.
A mutual fund's brochure touts its ten-year returns, calculated over the funds that survived the decade; the shuttered ones are off the books.
How to spot it
- Advice comes only from people whose strategy visibly worked — losers were not invited to the panel.
- The dataset 'begins' at a point when failures were already filtered out; nobody quotes the full intake.
- Ask 'who tried this and failed?' and the room has no names to give.