Esprit critique
← All biases
bias · Probabilities

Survivorship bias

Survivor bias, Biais du survivant

Drawing conclusions from a sample that contains only the cases that passed a selection, forgetting those that were eliminated.

Mechanism

Eliminated cases are by nature invisible: they do not testify, are not published, do not appear in rankings. The sample we observe is therefore not representative.

In practice

Example

People cite entrepreneurs who dropped out of school to succeed, without counting the many who made the same choice and failed.

Famous case

During the Second World War, rather than armouring the areas riddled with hits on aircraft returning from missions, Abraham Wald recommended armouring the areas with none: planes hit there did not come back. The same bias affects databases of investment funds, from which funds closed for poor performance disappear (Elton, Gruber and Blake, 1996).

Defending yourself

How to protect yourself

Ask yourself: what happened to the cases I don’t see? What filter produced the sample I am looking at?

Don’t confuse

Cherry-picking (fallacy)

In survivorship bias, it is reality that has filtered the data: only the survivors remain visible. In cherry-picking, it is the person arguing who chooses them.

Going further

Source

Abraham Wald, Statistical Research Group at Columbia University (1943).

Practise

In the app, each entry offers ways to practise recognising the concept. No sign-up: your progress stays on your device.

Open the interactive entry