# Survivorship Bias: Definition, Examples & How to Avoid It

> Survivorship bias means drawing lessons only from the winners you can see. Learn the Abraham Wald bomber story, everyday examples and how to avoid the trap.

Canonical: https://mentarium.io/biases/survivorship-bias.md

Cognitive bias

6 min read By [Nemo](https://mentarium.io/about) Updated October 30, 2026

> **Survivorship bias** is the error of drawing conclusions from the people, companies or things that made it through some selection process while ignoring the ones that did not. Because the failures are invisible, the survivors look like the whole story, and their traits get mistaken for the causes of success.

- **Category**: Cognitive bias (evidence and sampling)
- **Origin**: Abraham Wald, World War II aircraft analysis
- **Best used for**: Judging success stories, strategies and advice
- **Related**: Selection bias, availability heuristic, hindsight bias

## What is survivorship bias?

Survivorship bias is learning from the winners while forgetting that the losers existed. When you only see what made it through a filter, the filter itself disappears from view, and you end up crediting the survivors’ traits for an outcome that may owe a lot to luck or to the filter.

The classic signs:

- You are studying a group defined by success (billionaires, unicorn startups, best-selling authors).
- The failures are hard or impossible to observe (they closed, quit, died or were deleted).
- The lesson is “do what they did.”

If all three are true, you are probably looking at a biased sample. College dropouts who founded famous companies are visible; the far larger number of dropouts who did not are not interviewed on podcasts.

This page covers the bias: the error itself and why it is so easy to make. For the habit that prevents it (always ask to see the graveyard), see the companion mental model, [survivorship](https://mentarium.io/mental-models/survivorship), which is the model that fixes this bias.

## Where it comes from

The most famous illustration involves Abraham Wald, a mathematician who worked with the Statistical Research Group at Columbia University during World War II. The group analyzed military problems, including how to make bombers more likely to return from missions.

The popular telling goes like this: the military looked at returning planes, saw bullet holes clustered in certain areas, and proposed adding armor there. Wald pointed out that the sample only included planes that came back. Planes hit in other areas, such as the engines, mostly did not return. So the places without holes on surviving planes were the places where hits were deadly, and those were the places to armor.

Two honesty notes. First, Wald’s actual work was a set of technical memos on estimating how vulnerable different parts of an aircraft were, using data from returning planes; the tidy dialogue in the popular story is a simplification. Second, the widely shared picture of a plane covered in red dots is a later illustration, not a wartime document. The core insight, though, is real and is the heart of the bias: the missing data is often the most important data.

The term “survivorship bias” is now standard in statistics and finance, where it describes, for example, performance studies that only include funds still operating today.

## How to beat survivorship bias

1. **Ask where the graveyard is.** For any success story, ask: “Who else tried this, and what happened to them?” If you cannot answer, you are missing the denominator.
2. **Find the base rate.** How many people or companies start out on this path, and what share end up where the survivors are? The [base rates](https://mentarium.io/mental-models/base-rates) model turns a vague “it worked for them” into a number.
3. **Compare winners with losers, not winners with nothing.** A trait only explains success if it is more common among survivors than among failures. If failed startups also had passionate founders who worked long hours, those traits do not explain much.
4. **Check how the sample was built.** Was the data filtered by time, by success or by visibility? Databases, rankings and “top 100” lists are survivor lists by design.
5. **Discount advice from survivors.** Successful people are honest about what they did, but they usually cannot know which parts mattered and which parts were luck. Treat their advice as one data point.

 The Mentarium Field Guide

60 mental models, organized by the decision in front of you

Each model on one or two pages: when to use it, the questions to ask, an example and the trap. Plus the 25 biases that break decisions and 12 ready Model Stacks.

 [See Field Guide · $14](https://mentarium.io/shop/field-guide)

## 3 real-world examples

### Business

Say you read ten case studies of companies that bet everything on one bold product and won. The lesson seems to be “bet big.” But the companies that bet big and lost rarely get case studies; many of them no longer exist. Without them, you cannot tell whether bold bets are a good strategy or just a high-variance one that produces a few loud winners and many quiet failures.

### Personal life

Say a friend tells you their grandfather smoked every day and lived to 95. That is a true story, and it is survivorship bias in action. The smokers who died young are not around to tell their stories. The data on large groups of smokers, which includes those missing people, tells a very different story.

### Investing

Say a list of mutual funds shows strong average returns over the past 20 years. Funds that performed badly over that period were often shut down or merged into others, so they may not appear in the current list. The surviving funds’ average overstates what a typical investor would have earned picking funds 20 years ago.

## Common mistakes

- **Treating “the winners did X” as evidence that X works.** It only shows that X is compatible with winning. You need the losers to know whether it helps.
- **Thinking only data people suffer from it.** It shapes everyday beliefs: “music was better back then” (the forgettable songs were forgotten), “they built things to last” (the flimsy ones fell down).
- **Assuming the failures are similar to the survivors.** Sometimes the failures did something very different. You have to look, not assume.
- **Dismissing all success stories.** The fix is not cynicism. Survivors can teach a lot once you also study who did not make it.
- **Forgetting it inside your own data.** Customer surveys only reach customers who stayed. The ones who left, and their reasons, are missing unless you go and ask them.

## Survivorship bias vs hindsight bias

|  | Survivorship bias | Hindsight bias |
|---|---|---|
| What it is | Learning only from those who made it through a filter | Believing, after an outcome, that you saw it coming |
| What goes missing | The failures | Your earlier uncertainty |
| Typical phrase | “Every successful founder I know did this.” | “It was obvious that would happen.” |
| Typical result | Copying strategies that mostly fail | Overconfidence in your ability to predict |
| Main fix | Ask for the graveyard and the base rate | Write predictions down before outcomes arrive |
| Mentarium page | This page | [Hindsight bias](https://mentarium.io/biases/hindsight-bias) |

They often team up. Survivorship bias hides the failed attempts, and hindsight bias makes the winners’ path look inevitable. Together they produce the most common business story of all: a hero who saw what others missed.

## The Mentarium take

Survivorship bias is the reason so much advice is confident and so little of it is useful. The people with the biggest platforms are, by definition, survivors, and they describe the path they took as if it were a map. It is not a map. It is one route that happened to reach the summit, with no record of the routes that ended at a cliff.

Our rule: **a success story without a failure count is an anecdote.** That does not make it worthless. It makes it a hypothesis that needs the missing half of the data.

On the Mentarium **Bias Radar**, survivorship bias is one of the seven that most often distort business decisions. Strategy, hiring and product choices all get justified with “this is what the winners do.” When you hit one of those choices, use the Map step of **The Mentarium Decision Loop** to list the evidence for each option and tag every item as either “includes failures” or “survivors only.” Survivor-only evidence still counts, but it should never be the only thing holding a decision up. And because confident survivor stories feed [overconfidence bias](https://mentarium.io/biases/overconfidence-bias), the Stress-test step should ask one more question: “If I fail at this, will anyone ever hear about it?” If the answer is no, the stories you have been hearing are filtered.

**Fixed by** [Survivorship](https://mentarium.io/mental-models/survivorship)

## FAQ

### What is a simple example of survivorship bias?

Noticing that old buildings in a city are beautiful and concluding that buildings used to be built better. The ugly and badly built old buildings were mostly torn down, so the ones still standing are a filtered sample.

### What is the Abraham Wald airplane story?

During World War II, the statistician Abraham Wald worked on how to protect bombers. The popular version says the military wanted to armor the areas where returning planes had bullet holes, and Wald pointed out that the planes hit elsewhere had not returned, so the undamaged areas needed protection. The real analysis was more technical, but the core insight is the same.

### How is survivorship bias different from selection bias?

Survivorship bias is a type of selection bias. Selection bias is any error caused by a non-random sample. Survivorship bias is the specific case where the sample is filtered by who or what survived.

### Why is survivorship bias a problem in investing?

Funds and companies that perform badly are often closed, merged or dropped from indexes and databases. If you only study the ones that remain, past returns look better than they really were for investors at the time.

### How do you avoid survivorship bias?

Ask where the failures are. Before learning from a success, find out how many tried the same thing and what happened to them. If you cannot see the failures, treat the lesson as a hypothesis, not a rule.

## Related models & biases

[Mental model Survivorship Survivorship, as a mental model, is the habit of asking what is missing from the evidence before you learn from it: the failures, dropouts and dead projects that never made it into view. Its rule is 'always ask for the graveyard', and its classic example is Abraham Wald's WWII analysis of where to armor returning aircraft.](https://mentarium.io/mental-models/survivorship) [Mental model Base Rates and Base Rate Neglect A base rate is how often something happens in a relevant population, before you look at any specific case. Base rate neglect (the base rate fallacy) is ignoring that number in favor of vivid case details. Kahneman and Tversky showed it with the taxicab problem, where most people overweight a witness and ignore how few cabs are Blue.](https://mentarium.io/mental-models/base-rates) [Cognitive bias Hindsight Bias Hindsight bias is the tendency, once you know how something turned out, to believe the outcome was more predictable than it really was. It quietly rewrites memory of your earlier uncertainty, producing the 'I knew it all along' feeling and making past decisions look either obviously smart or obviously foolish.](https://mentarium.io/biases/hindsight-bias) [Cognitive bias Availability Heuristic The availability heuristic is a mental shortcut in which you judge how common or likely something is by how easily examples come to mind. Recent, vivid, emotional or heavily covered events are easy to recall, so they feel more frequent than they really are, while quiet, common risks feel rarer.](https://mentarium.io/biases/availability-heuristic) [Cognitive bias Overconfidence Bias Overconfidence bias is the tendency to be more sure of your judgments, abilities or predictions than the evidence justifies. Researchers split it into three forms: overestimating your own performance, overplacing yourself relative to others, and overprecision, being too certain your beliefs are accurate. It is one of the most studied and most costly biases.](https://mentarium.io/biases/overconfidence-bias)

 Mentarium Decision Coach · free

Use this with your AI

Connect the free Decision Coach to Claude or ChatGPT and it walks you through the Decision Loop, picks the right Model Stack and runs the Bias Radar on your reasoning.

 [Connect it free](https://mentarium.io/coach)

 Mentarium Decision Coach · free

Use this with your AI

 [Connect it free](https://mentarium.io/coach)