Survivorship Bias

Also known as: Survival bias

Survivorship bias is a logical error where we focus on the "survivors" of a process (those who succeeded) and ignore the "failures" (those who didn't), leading to false conclusions about what causes success. We see the winners, but the losers are invisible.

Statistical Biases

2 min read

experimental Evidence


Survivorship Bias

The Psychology Behind It

History is written by the victors. In data, history is written by the survivors. When we look for patterns, we naturally look at the visible data. But often, the most important data is the data that is missing.

The most famous example comes from WWII. The military looked at planes returning from battle to see where they had been shot, planning to reinforce those areas. Statistician Abraham Wald stopped them. He pointed out that they were only looking at the planes that survived. The planes that were hit in the engines or cockpit didn't come back. Therefore, the bullet holes on the returning planes showed where a plane could be hit and still survive. They needed to reinforce the areas with no bullet holes.

Real-World Examples

Business Advice

We study successful companies (Apple, Google) and try to copy their habits. "Steve Jobs dropped out of college, so I should too!" We ignore the thousands of dropouts who started failed companies. Their stories are not on the magazine covers.

Architecture

"They don't build them like they used to." We look at old buildings and think they were built better. In reality, the badly built old buildings collapsed or were demolished long ago. Only the strong ones survived.

Music

"Music from the 70s was so much better." We only listen to the hits that stood the test of time. We have forgotten the 99% of terrible songs that were on the radio back then.

Consequences

Survivorship bias can lead to:

  • False Causality: We attribute success to random traits (e.g., "waking up at 4 AM") rather than the actual factors.
  • Overconfidence: We underestimate the difficulty of success because we don't see the failures.
  • Risky Behavior: We take risks (like skipping insurance) because "I've never had an accident before" (survivor logic).

How to Mitigate It

Look for the graveyard.

  1. Ask "What is missing?": When presented with data, ask what data was excluded or destroyed.
  2. Study Failure: Don't just read biographies of billionaires. Read post-mortems of failed startups. They often did the exact same things as the winners, but had bad timing or bad luck.
  3. The Base Rate: Always ask for the success rate of the entire pool, not just the characteristics of the winners.

Conclusion

Survivorship bias is the invisible silent killer of truth. It tricks us into believing that success is easy and predictable. By remembering the silent majority of failures, we can see the world—and the odds—more clearly.

Mitigation Strategies

Invert the Problem: Instead of asking 'What do successful people do?', ask 'Do unsuccessful people also do this?' If yes, it's not the cause of success.

Effectiveness: high

Difficulty: moderate

Data Auditing: Check if the dataset includes 'dead' or 'inactive' entries. If not, it is biased.

Effectiveness: high

Difficulty: moderate

Potential Decision Harms

Aspiring founders take reckless risks because they only see the 'unicorns' and not the graveyard of failures.

major Severity

Doctors may overestimate the survival rate of a disease if they only see patients who live long enough to reach the specialist hospital.

critical Severity

Key Research Studies

A method of estimating plane vulnerability based on damage of survivors

Wald, A. (1943) Statistical Research Group, Columbia University

The foundational text on survivorship bias, demonstrating how to infer missing data from surviving data.


Related Biases

Explore these related cognitive biases to deepen your understanding

Neglect of Probability

2 min read

Neglect of probability is the tendency to completely disregard probability when making a decision under uncertainty.

Statistical Biases

/ Probability blindness

Ludic Fallacy

2 min read

The ludic fallacy is the misuse of games to model real-life situations.

Statistical Biases

/ Gaming fallacy

Sampling Bias

2 min read

Sampling bias is a bias in which a sample is collected in such a way that some members of the intended population have a lower or higher sampling probability than others.

Statistical Biases

/ Ascertainment bias

Selection Bias

2 min read

Selection bias is the bias introduced by the selection of individuals, groups or data for analysis in such a way that proper randomization is not achieved.

Statistical Biases

/ Sampling bias (related)

Texas Sharpshooter Fallacy

2 min read

The Texas sharpshooter fallacy is an informal fallacy which is committed when differences in data are ignored, but similarities are overemphasized. From this reasoning, a false conclusion is inferred.

Statistical Biases

/ Clustering illusion (related)

Pareidolia

2 min read

Pareidolia is a specific form of apophenia involving the perception of images or sounds in random stimuli, such as seeing faces in clouds.

Statistical Biases

/ Face pareidolia