Why Armoring B-17s’ Bullet-Riddled Areas Was a Mistake

November 21, 2026

New York, 1943. In the midst of World War II, the American army faced a tragic bleeding wound: too many B-17 bombers did not return from missions, shot down by German anti-aircraft defenses. To curb the carnage, the general staff gathered its best engineers with a simple directive: “Strengthen the aircraft.” They meticulously analyzed the fuselages of the planes that managed to come back to base. The finding was unequivocal: the wings, the tail, and the fuselage were peppered with bullet holes. The military logic seemed irrefutable: these exposed areas must be armored. It is the very essence of common sense. Yet a Hungarian mathematician named Abraham Wald rose up and said exactly the opposite. “If you armor these places, you will kill even more pilots.” He had identified one of the most dangerous traps of the human mind.

The Illusion of Visible Impacts

The generals’ error did not stem from their eyes but from their data. They looked at detailed records of the hits observed on the planes that survived. The concentration of fire on the wings and the tail was undeniable.

But Abraham Wald, a member of the highly secretive Statistical Research Group (SRG), asked the uncomfortable question: “Where are the hits on the engines and the cockpit?” The military’s response was an awkward silence. There were almost none on the aircraft that returned. For the army, that signified that the engines were not often hit. For Wald, it meant that the planes struck in the engines never returned.

Therein lies the genius of deduction: the most important data were those that were missing. The holes in the fuselage represented “survivable” damage. An airplane could return with a shredded wing. Conversely, a hit to the engine would cause it to explode in flight or crash on enemy territory. It disappeared from the statistics.

Wald thus recommended armor for the spots where there had been no hits on the planes that returned. The army followed his advice. The survival rate rose immediately.

The Silent Cemetery of Data

This phenomenon today bears a name: the “survivor bias.” It is a logical error that tends to overvalue the elements that have passed a selection process, while ignoring those that failed, simply because they are invisible.

This tale from 1943 is the perfect parable of our modern view of success. We are constantly bombarded with examples of extraordinary achievements that distort our judgment. When we hear “Steve Jobs and Mark Zuckerberg dropped out of college and became billionaires,” we see the planes that returned. We do not see the thousands of students who left college to end up in debt and anonymity. They constitute the “silent cemetery” of data. They do not make the covers of magazines, they do not write biographies. They are the shot-down planes.

The Architecture of Error

This bias seeps in everywhere. In architecture, we admire the solidity of Roman buildings and tell ourselves, “we can’t build like that anymore.” That is false. We only see the 1% of Roman structures sturdy enough not to crumble for 2000 years. The 99% of ancient architectures that were mediocre have long collapsed into dust.

The same goes for health. The smoking uncle who lives to 95 is a statistical anomaly, a plane that somehow landed. Using him as proof that “tobacco does not kill” is a mathematical insult to the millions of smokers who died prematurely and are no longer here to testify.

The lesson of Abraham Wald is more relevant than ever: to understand reality, don’t look only at what is visible. Seek the void. Always ask: “Who is missing from the equation?” It is often in the silence of the absent that truth hides.

Article prepared with the help of AI · Our method

Sindre Halvorsen

I write about space exploration, frontier science and the technologies that are quietly shaping the future. From Norway, I follow the missions, discoveries and ideas that connect life on Earth with what lies beyond it. My goal is to make complex subjects clear, useful and worth paying attention to.