Psychology
Survivorship Bias in Historical and Business Analysis
Quick fact
During World War II, statistician Abraham Wald advised the US Air Force to reinforce the places on returning planes that had NO bullet holes, because the planes that had been hit there never made it back—a classic example of survivorship bias.
Why this is interesting
Why do we believe that startup success is more common than it really is? And why do we think historical heroes were more exceptional than they were?
Read the full explanation
Understanding Survivorship Bias in Historical and Business Analysis
Survivorship bias is a mental shortcut that happens when we only look at the winners—the people, companies, or ideas that made it—and ignore the countless failures that didn't. This makes success seem more likely than it actually is. Think of it like looking at a pond with only ducks on the surface and concluding that all fish are ducks—you never see the fish below. In history, we study the few who achieved greatness (like Napoleon or Einstein) but rarely the millions who tried and failed. In business, we read about Apple and Google but not the thousands of tech startups that went bankrupt. The bias happens because the failures are often invisible: they don't write books, they don't get press coverage, and their data is often lost. Our brains are wired to notice what's present, not what's absent, so we overestimate the odds of success and misjudge the traits that lead to it.
A deeper explanation
The mechanism behind survivorship bias is a type of selection bias. When we analyze a set of outcomes, we often only have access to the data from the 'survivors'—those that remain after a selection process (e.g., companies that are still public, people who are still alive, products that are still on shelves). The data from the 'non-survivors' is missing, either because it was never collected, was destroyed, or simply isn't as visible. This missing data creates a distorted view of reality. For example, in historical analysis, the archives are filled with records of successful leaders and victorious battles, but accounts of defeats and failed movements are often sparse or forgotten. In business, when we look at successful companies, we see their decisions and strategies, but we might attribute their success to those decisions without realizing that many failing companies made the same decisions. The root cause is that the selection process itself (survival) is correlated with the outcome we're studying, so we systematically exclude relevant data. This leads to overestimating success rates, misattributing causes, and making poor decisions—like investing in a startup because it resembles a unicorn, without considering the thousands of similar startups that failed.