Survivorship Bias

Definition

Core Statement

Survivorship Bias is the logical error of focusing on the people or things that "survived" some process and ignoring those that did not because of lack of visibility. This leads to false conclusions because the sample is not representative of the whole population.

Purpose

Correct decision making: Avoid optimizing for the wrong traits.
Investment analysis: Mutual fund performance looks better than it is because failed funds vanish.
Startup/Success advice: "I dropped out of college and succeeded" ignores the millions who dropped out and failed.

The Classic Example: WWII Planes

Abraham Wald & The Bullet Holes

Scenario: The military analyzed planes returning from battle.
Data: Most bullet holes were found on the Wings and Tail.
Military's Plan: "Put more armor on the Wings and Tail!"

Wald's Insight:

The planes you are looking at Returned. They survived.
This means bullet holes in wings/tail are survivable.
The planes that were hit in the Engine or Cockpit... never came back.

Conclusion: Armor the area where there are no bullet holes (the Engine). The missing data tells the story.

Common Scenarios

Context	Bias	Resulting Fallacy
Finance	Analyzing indices (S&P 500)	"Stocks always go up" (You ignored Enron, Lehman Bros which were delisted).
History	"They don't make them like they used to"	You only see the old buildings that survived. The cheap ones collapsed long ago.
Business	"Do what Bill Gates did"	Ignores the luck factor and the silent graveyard of failed startups.

How to Detect & Mitigation

Checklist

Where are the failures? Am I seeing the full dataset or just the winners?
Is the data censored? (e.g., Customer surveys only reach current customers, not those who churned in anger).
Look for the invisible: Ask "Who is missing from this room?"

Selection Bias - The broader category.
Missing Data - Techniques to handle mechanism of missingness.
Sampling Bias