Survivorship Bias Lesson – How Only Studying the Winners Can Mess Up Your Foresight
In 1943, statistician Abraham Wald and the Statistical Research Group at Columbia University were asked where to add armor to American bombers coming home from raids over Europe. The planes returned peppered with bullet holes across the wings and fuselage, and the obvious move was to reinforce exactly those spots. Wald argued the opposite: armor the places with no holes. The damage on the survivors marked where a bomber could be hit and still make it back. The fatal hits were on the planes that never returned — and those were never in the data.
That’s survivorship bias: drawing conclusions from the winners alone — the survivors, the success stories, the planes that landed — while the ones that failed sit silently outside the sample. “Just save up a down payment.” “Just finish the degree.” “Here’s what successful founders do.” Each is advice distilled from the people it worked for, with everyone it didn’t nowhere in the room. There’s an ego version, too: if the system worked for me, it must be sound. The discipline is to hunt for the missing data on purpose — to ask who isn’t in the sample, and why. So whose absence are you failing to account for?
Eric Garland is an intelligence analyst personally sanctioned by the Russian Federation in May 2022 for his work on Russian intelligence operations. He is the author of the 2016 “Game Theory” thread, named by The New York Times as one of the 25 most important tweets in history.
🔔 Subscribe for daily geopolitical intelligence analysis.
📝 Substack: ericgarland.substack.com
🎤 Book Eric to speak: Leading Authorities — https://www.leadingauthorities.com/speakers/eric-garland
SOURCES
Abraham Wald — “A Method of Estimating Plane Vulnerability Based on Damage of Survivors Returned,” Statistical Research Group, Columbia University, 1943 (reprinted, CRC, 1980)
Competitive Futures — “Cognitive Bias for Intelligence Analysts,” Survivorship Bias (No. 31)
