Clicked Gallery

What is survivorship bias?

Highlighted from a real textbook passage. Explained by Clicked.

Used in a sentence

University Course Reader · STEM

Critics argued the fund industry's performance claims suffered from survivorship bias, since failed funds vanish from the averages.

The reader highlighted one term — in an article or a paper PDF. Clicked made the statistics term “survivorship bias” easy to understand:

Explained in three depths

Same facts, different vibe — Slang mode 😎

The Clicked way

●○○

Overview

Survivorship bias is drawing conclusions from whoever made it through a filter, while the ones who didn't are nowhere in your data. Study thriving companies, long-lived smokers, or sturdy old buildings and you are studying survivors: the failures closed, fell silent, or crumbled long ago. The missing cases would change the answer, but they never show up to be counted.
●○○

Overview

Survivorship bias is when the evidence looks great because the counterexamples are gone. Your feed is a gallery of people the filter let through: founders who made it, traders up 400%, someone's grandpa who smoked daily and hit 95. The ones the same choices wrecked aren't posting. That is not a track record. That is what is left after the losses deleted themselves. 😎

A quick take — often all you need.

●●○

Detail

Survivorship bias happens when the data that reaches you has already been filtered by success, so the failures are invisible and everything you compute is tilted. The famous case comes from the Second World War, when analysts mapped bullet holes on returning bombers and proposed armor where the holes clustered. The statistician Abraham Wald turned the logic around: these planes made it home, so their holes marked survivable damage. Planes hit in the engine never made it back, and the armor belonged where the survivors looked clean. That inversion is the whole skill, because the modern versions are everywhere. Advice from billionaire dropouts reaches you because they succeeded; the dropouts who failed write no books. Old furniture seems better made because the flimsy pieces became firewood decades ago. Fund companies quietly close their losing funds, so the surviving lineup boasts an average no investor actually earned. Nobody has to lie for any of this. An honest look at whoever is left still produces the wrong answer, because the leaving was not random. It is a form of selection bias, and the antidote is one question asked early: who is missing from this data, and what happened to them?
●●○

Detail

Survivorship bias is why "it worked for them" is the weakest evidence on earth, and once you see the shape of it, it is everywhere. Every hustle guru teaching their method is someone the method didn't ruin; the people it ruined are busy being quiet. "They don't build things like they used to" is praise aimed at the old houses still standing, and only the sturdy ones are still standing to collect it. Grandpa's cigarettes prove nothing except that you only get stories from grandpas who are still here to tell them. Finance runs a professional version: close the embarrassing funds, and the family photo of the remaining ones averages out heroic. The trap is that no lying is required, which is what makes it so hard to catch. Everyone in the sample is telling the truth. The sample itself is the lie. So build the reflex: when the evidence is a lineup of winners, ask what happened to everyone who started. If nobody can answer that, the impressive part of the story is the part that walked away. 😎

Want more? One click digs deeper.

●●●

Analogy

Survivorship bias is judging your graduating class by who attends the twenty-year reunion. The room is full of steady careers and happy updates, and the natural conclusion is that the class did remarkably well. But a reunion is a filter. Classmates in rough shape mostly stayed home, and the ones who moved without a trace never got an invitation at all. Nobody deceived you. The room is honestly reporting on the room. It just isn't the class; it is the slice of the class that a good stretch of life delivers to a banquet hall, and any conclusion drawn inside it inherits the filter at the door.
Deepest level reached.
●●●

Analogy

Survivorship bias is the in-app poll of a mediocre game. "97% of players love it!" Technically true, because the poll runs inside the app, and everyone who hated the game uninstalled it months ago. The haters cannot be surveyed; the survey lives in the one place they no longer go. The remaining players answer honestly, five stars all around, and the number is real, gorgeous, and useless. Same math outside a cinema: poll the crowd at the closing credits and the movie scores wonderfully, because everyone who hated it walked out an hour ago. The walkouts would love to weigh in. The poll never reaches them. 😎
Deepest level reached.

Unfamiliar concept? A real-world example makes it click — fresh analogies on tap.

AI explanations may contain errors · Not professional advice

Formal definition — The same term, explained the usual way

Survivorship bias is a form of selection bias in which analysis is restricted to entities that passed a survival or persistence filter, while those eliminated along the way are absent from the dataset, biasing estimates of performance, durability, or risk upward. Documented instances include mutual fund databases that exclude closed funds, clinical studies affected by participant attrition, and inferences from historical artifacts, which necessarily overrepresent durable items. Mitigations include reconstructing the full initial cohort, tracking dropouts explicitly, and intention-to-treat analysis in trials.

Want Clicked to explain terms like “survivorship bias” directly in your browser — including on PDFs?

Add to Chrome — Free

50 free Explanations · No credit card required