Skip to main content
note
This page was translated from the German original, partly by machine. Some passages may read awkwardly or contain inaccuracies. When in doubt, please read the original.

Survivorship bias

In brief

Success is visible, failure stays invisible.

💬Quote

There are some who are in darkness
And the others are in light.
And you see the ones in brightness
Those in darkness drop from sight.

Bertolt Brecht, The Threepenny Opera, Song of the Insufficiency of Human Endeavour, 1928.

Definition​

Survivorship bias describes the tendency to draw false conclusions from the properties or strategies of the successful, because the numerous failed cases remain invisible and are not taken into account.

EN: Survivorship Bias

Survivorship bias is closely connected with several other biases and is reinforced by them:

  • Selection bias: A distorted sample due to selecting only visible/successful cases.
  • Publication bias: Studies with positive results are more likely to be published than negative or null results.
  • Base rate fallacy: Base probabilities are ignored when only the "winners" are considered.
  • Availability heuristic: Success stories are present and memorable — failures disappear from view.
  • Hindsight bias: Retrospective rationalization of success ("It was obvious all along that it would work").

Examples​

The great swimmers​

We look at Olympic champions and derive "recipes for success" from their training, diet or physique. In doing so we ignore thousands of similarly talented swimmers who were injured, gave up or, despite identical routines, never made the breakthrough. A false pattern is constructed from the survivors.

Startup founders who "figured it all out"​

Successful founders share playbooks in podcasts and books. You hear only the voices of the survivors, not those of the countless founders with the same strategy who failed. This produces spurious causality ("X leads to success") instead of a sober consideration of probabilities, market conditions and chance.

Effects​

  • Overestimation of "recipes for success" and underestimation of chance, context and base rates
  • Bad decisions due to distorted evidence (investment, product, career)
  • The forming of myths and false role models; risky imitation without risk assessment

Countermeasures​

  • Systematically capture the non-success cases too (failure data sets, postmortems)
  • Look at base rates and complete samples; explicitly examine the selection process
  • Take replication and publication bias into account; include negative/null results
  • Check success narratives for causality vs. correlation (alternative explanations, confounders)

Sources​

  • Wikipedia: Survivorship bias
  • Kahneman, D. (2011): Thinking, Fast and Slow — sections on base rates and selection biases.
  • Taleb, N. N. (2001): Fooled by Randomness — on chance, success and misinterpretations.

Artikel auf Deutsch | Article en français