Survivorship Bias: The WWII Planes That Came Back
Featuring Abraham Wald
During World War II, US analysts studying bullet holes on returning bombers recommended adding armor exactly where the planes were hit most. The logic was airtight, the data was real, and the statistician Abraham Wald told them they had it precisely backwards. The flaw wasn't in the numbers. It was in which planes were sitting in the hangar to be counted, and which ones never made it back to be counted at all.
For founders and operators, this is survivorship bias in its cleanest possible form, and it quietly distorts decisions everywhere: studying winning companies for success patterns while the firms that did the same things and failed stay invisible, reading testimonials while churned customers go silent, hiring for traits your best people share without knowing if those traits also predict failure. The case sharpens the habit of asking who is not in your sample before you trust what the sample seems to say. Wald's reasoning, and the conclusion it forced, is what the app has you reconstruct rather than be handed.
Frequently asked questions
What is survivorship bias and the WWII planes example?
Survivorship bias is the error of drawing conclusions only from the cases that survived to be examined, ignoring the ones that did not. In World War II, US analysts studying bullet holes on returning bombers recommended adding armor where the planes were hit most. The flaw was that they were only counting planes that made it back.
What did Abraham Wald conclude about armoring the bombers?
Statistician Abraham Wald told the analysts they had it precisely backwards: armor should go where the returning planes were not hit. Planes hit in those places never made it back to be counted, so the absence of holes there marked the fatal spots. The flaw was not in the numbers but in which planes were available to study.
Why is survivorship bias so dangerous in business decisions?
Survivorship bias quietly distorts decisions because the failures stay invisible. Studying winning companies for success patterns ignores firms that did the same things and failed, reading testimonials ignores churned customers who went silent, and hiring for traits your best people share ignores whether those traits also predict failure. You trust a sample without asking who is missing from it.
What can founders learn from survivorship bias?
The lesson is to ask who is not in your sample before you trust what the sample seems to say, because the cases that failed and disappeared often carry the real signal. Winners, testimonials, and your best hires all hide the silent failures. CaseBook turns this into a move you apply to your own company, with an AI coach that reads your answer.