During World War II, the United States faced a practical problem: where should they reinforce the armor on their planes? The answer seemed obvious. Just look at the fighters returning from battle, mark where the bullet holes were — wings, fuselage, tail — and armor those areas. That's when a 26-year-old mathematician, Abraham Wald, turned the logic on its head.
Wald said: reinforce exactly where there are no holes. The reasoning is unsettling precisely because it's so simple. If a plane made it back to base riddled with holes in the wings, it's because taking a hit to the wing isn't fatal. The planes hit in the critical spots — engine, cockpit — simply never came back to tell the tale. The missing data mattered more than the visible data.
What survivorship bias is
This mistake has a name: survivorship bias. It happens every time we draw conclusions by looking only at those who "survived" a process, ignoring every case that vanished along the way. And it's nothing new. Two thousand years before Wald, the philosopher Diagoras had already given the same answer. He was shown paintings of sailors who prayed during a storm and escaped shipwreck, as proof that faith saves. He asked: and where are the paintings of those who prayed and drowned?
The mistake you make every day
The trap is in almost every business decision. Everyone cites Zuckerberg, Gates and Jobs — they dropped out of college and became billionaires — as if that were the recipe. But no one mentions that Jeff Bezos graduated in engineering and computer science, Tim Cook in industrial engineering, and Elon Musk holds two degrees. And, above all, no one sees the millions who dropped out and went broke. You see the winner on display in the shop window and never the graveyard of similar attempts that failed. So you copy half the game and call it strategy.
The same goes for the competitor's "success story," the campaign "everyone is running," and the advice of the entrepreneur who made it. None of them show the baseline: how many did the exact same thing and never made it to the stage.
How to see the whole game again
The way out isn't to doubt everything — it's to deliberately look for the data that's missing. Before an important decision, it's worth asking: who tried this and failed? Where are the paintings of those who drowned? One practical approach is to build a simple matrix with four cases: who did it and succeeded, who did it and failed, who didn't do it and succeeded, who didn't do it and failed. Usually only the first box is lit up; it's the other three that reveal whether something truly works or was just luck.
What this has to do with your business
In the day-to-day of a company, survivorship bias lives in the data you look at. It's natural to analyze only the customers who closed, the posts that went viral, the ads that converted — and to ignore the quotes that went nowhere, the visitors who left the site in silence, the leads who never replied. That's exactly where — in what didn't come back — the most valuable information usually hides. Measuring the whole business, not just the part that survived, is the first step to making decisions with the full game in view. And that kind of data reading is what technology, well applied, makes possible.
Post inspired by an edition of the Email do Rony newsletter, by Rony Meisler (founder of Reserva). Worth reading at the source: businessofbrandspost.substack.com.