Back to Paradoxes

Survivorship Bias

The bullets that killed planes didn't leave holes on returning aircraft. Abraham Wald saw what others missed.

🛩️ The WWII Armor Problem

It's 1943. Allied bombers are being shot down over Germany at alarming rates. The Navy needs to add armor—but armor is heavy. Where should it go?

YOUR MISSION: Click on the plane to add armor where you think it's needed most. Look at where the returning planes have been hit (red dots) and protect those areas!
0
Fuselage Armor
0
Wing Armor
0
Engine Armor
0
Cockpit Armor
💡 Wald's Counterintuitive Insight

📜 The Brilliant Mathematician

In 1943, the Statistical Research Group (SRG)—a classified program at Columbia University—assembled America's brightest statisticians to help win the war. Among them was Abraham Wald, a Hungarian-Jewish mathematician who had fled Nazi persecution.

The military presented a problem: returning bombers showed damage concentrated on the fuselage and wings. The engines? Almost untouched. The obvious solution: reinforce the fuselage and wings where the damage was.

"You're looking at the wrong data. These planes came back. You need to armor the places where the returning planes are NOT hit."
— Abraham Wald's insight

The Missing Data

Wald recognized what everyone else missed: the data was only from survivors. Planes hit in the engines, fuel systems, and cockpit didn't return—they were lying at the bottom of the English Channel or in German fields.

The "clean" areas on returning planes weren't safe—they were fatal. A plane can fly home with holes in its fuselage, but a single bullet in the engine means it never returns.

The military listened. They armored the engines, fuel systems, and cockpits. Bomber survival rates increased dramatically, and Wald's insight continued influencing aircraft design through the Vietnam War.

🚀 The Startup Graveyard

We celebrate Zuckerberg, Musk, and Bezos. We study their habits, read their biographies, and copy their strategies. But for every unicorn, there are thousands of identical failures we never hear about.

Below are 100 startups. Only 10 survived. When we only see the survivors, we develop dangerously wrong ideas about what leads to success.

10
Survivors (visible)
??
Failed (hidden)
??%
Actual Fail Rate

📊 The Mutual Fund Illusion

Mutual fund advertisements boast impressive returns. But they only show funds that still exist. Failed funds are quietly dissolved and vanish from the records.

Survivors Only: 12% avg
Including Failed: 6% avg

Research by Morningstar found that including defunct funds reduces 10-year returns by approximately 1% annually. Over decades, this compounds into massive wealth differences. That "market-beating" fund manager? Their competitors who failed worse are invisible.

🔍 Survivorship Bias Is Everywhere

🎵
Music Was Better Before
We only remember hits from the 70s. The thousands of terrible songs didn't survive. Today's bad music will be forgotten too.
🏛️
Ancient Architecture
"They don't build 'em like they used to!" We only see buildings that lasted 2000 years—not the millions that crumbled.
🎓
College Dropouts
Gates, Zuckerberg, and Jobs dropped out and became billionaires. Millions of dropouts working minimum wage don't get TED talks.
🐱
High-Rise Syndrome
Cats falling from higher floors have lower injury rates—because we only count cats that survived to reach the vet.
📚
Self-Help Books
Bestselling authors share "secrets" to success. The thousands who followed identical advice and failed don't write books.
🏥
Medical Studies
Early trial dropouts often had bad reactions. If only survivors complete studies, treatments look safer than they are.

🛡️ Armor Against the Bias

1. Ask "Where Are the Missing?"

For every success story, ask: "How many tried the same thing and failed?" If you can't find failures, your data is incomplete.

2. Seek Disconfirming Evidence

Don't just study successful people or companies. Study failures intentionally. What did they do wrong? What did they do the same as successes?

3. Demand Base Rates

Before copying a strategy, ask: "What percentage of people using this strategy succeed?" A 10% success rate means 90% failure rate.

4. Be Suspicious of "Secrets"

If success had a simple formula, everyone would follow it. The very existence of a "secret" suggests survivorship bias is hiding the failures.

"The dead don't write histories. The failed don't give TED talks. The bankrupt don't write business books. What you see is systematically different from what exists."
— The Core Lesson

🏆 Abraham Wald's Legacy

Wald's insight saved countless lives during WWII, but his impact extends far beyond. His work on sequential analysis revolutionized quality control in manufacturing. His statistical methods influenced everything from clinical trials to economic policy.

Tragically, Wald died in 1950 when his plane crashed during a trip to India—an ironic end for the man who taught us how planes should be armored. He was 48 years old.

His lesson endures: the data you can see is shaped by what's missing. The holes in returning bombers told a story—but it was the planes that never returned that held the truth.

Next time you see a successful person, a thriving company, or an impressive statistic, remember Wald's question: What's not here, and why?