# Survivorship: The Mental Model of Always Asking for the Graveyard

> Survivorship as a mental model: before learning from winners, ask for the graveyard. Abraham Wald's WWII armor analysis, steps and real-world examples.

Canonical: https://mentarium.io/mental-models/survivorship.md

Mental model

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

> **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.

- **Category**: Mental model (statistics and critical thinking)
- **Origin**: Abraham Wald, Statistical Research Group, WWII
- **Best used for**: Judging advice, success stories and any filtered data
- **Related**: Base rates, inversion, survivorship bias

## Ask yourself

- Who started this but isn't in the sample I'm looking at?
- What filter decided which cases I get to see?
- Did the failures do the same things as the winners?
- Where would I go to find the graveyard?

## What is survivorship as a mental model?

Survivorship is the habit of asking what the evidence leaves out before you learn from it. Its rule: always ask for the graveyard. If you only study the planes that came back, the companies that grew or the people who got rich, you are looking at a filtered sample, and the filter may be the whole story.

This page is about the method. The error it prevents, [survivorship bias](https://mentarium.io/biases/survivorship-bias), has its own page. Here the focus is what to do: how to find the missing cases, and how to change your decision once you do.

The model rests on three questions:

- **What was the filter?** Something decided which cases you can see: survival, success, publication, popularity.
- **What does the graveyard look like?** Who tried the same thing and dropped out?
- **Do winners and losers differ on the trait I care about?** If both groups did it, it doesn’t explain winning.

## Where it comes from

The best-known example comes from World War II. Abraham Wald, a mathematician and statistician, worked with the Statistical Research Group at Columbia University, which analyzed military problems for the United States. One question was where to add armor to bombers, since armor is heavy and can’t go everywhere.

The data came from aircraft that returned from missions, with damage spread across the wings and fuselage. The obvious answer was to armor where the holes were. Wald’s insight was the reverse: these were the planes that survived being hit there. Hits in the areas with few holes, such as the engines, were the ones that tended to bring planes down, which is why those planes weren’t in the sample. Armor belonged where the returning planes were clean.

The popular retelling compresses Wald’s work, which was a set of technical memos on estimating vulnerability from survivors’ damage. The logic is faithful, though: the missing data was the most important data.

Long before Wald, Cicero told a story about Diagoras, who was shown painted tablets of sailors who had prayed and survived a shipwreck, as proof that prayer works. He asked where the tablets were of those who prayed and drowned. Nassim Nicholas Taleb retells it in _The Black Swan_ (2007), calling the failures history doesn’t record “silent evidence”.

## How to use survivorship

1. **Name the sample and the filter.** Write down what you are learning from and how it reached you: “founders interviewed on podcasts”, “funds still open today”, “returning aircraft”.
2. **Estimate the graveyard.** How many started? How many dropped out? Use [base rates](https://mentarium.io/mental-models/base-rates): what share of new restaurants, startups or funds usually survive?
3. **Go find some of the dead.** Look at shut-down companies, closed funds, abandoned projects, people who tried and quit. Even a handful changes the picture.
4. **Compare the trait across both groups.** If winners and losers both worked 80-hour weeks or both dropped out of college, that trait doesn’t explain success.
5. **Use [inversion](https://mentarium.io/mental-models/inversion).** Ask what killed the failures, then check whether your plan avoids those causes. The graveyard teaches more about risk than the winners teach about method.
6. **Adjust your odds, not just your story.** Use [probabilistic thinking](https://mentarium.io/mental-models/probabilistic-thinking): if the base rate of success is low, your plan needs a margin for being in the majority.

A quick version for everyday use: whenever you hear “everyone who did X succeeded”, reply in your head with “how many did X?” If nobody knows, the claim is a story, not evidence. The same goes for reviews, testimonials and your own memories of what worked. The cases that went badly tend to leave the record quietly, through closed accounts, deleted posts and things people prefer not to mention.

 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 study ten famous founders who dropped out of college and conclude that dropping out helps. Asking for the graveyard means asking how many dropouts started companies that failed, and how many graduates built large companies. Once you include them, the pattern usually weakens or disappears. What the winners share with the losers can’t explain the difference. Look instead for what the winners did that the failures didn’t.

### Personal life

Imagine a friend says old furniture was built better, pointing to a 100-year-old oak table. The poorly made furniture from that era mostly broke and was thrown away long ago. What survives is the sturdy share. The model doesn’t say old furniture was worse; it says the surviving pieces are not a fair sample of what people used to buy.

### Investing

Fund performance tables often show only funds that still exist. Funds that did badly tend to be closed or merged, and they drop out of the averages, which makes the category look better than investors’ real experience was. When you compare a strategy’s track record, ask whether closed funds are included. If they aren’t, the reported average is likely flattering.

## Common mistakes

- **Copying winners’ habits.** Habits that the failures shared explain nothing.
- **Assuming the graveyard is small.** In most competitive fields, failures far outnumber survivors.
- **Stopping at “it’s survivorship bias”.** Naming the bias isn’t the point. Find the missing data and change the decision.
- **Overcorrecting into cynicism.** Survivors can still teach something, especially about what they avoided. The model asks for comparison, not dismissal.
- **Forgetting your own filters.** Your feed, your network and your memory are filtered samples too.
- **Trusting the sample you were handed.** Lists, rankings and case studies were selected by someone. Ask how before you use them.

## Survivorship vs survivorship bias

|  | Survivorship (the model) | Survivorship bias (the bias) |
|---|---|---|
| What it is | A method for finding missing data | The error of judging by survivors only |
| Core rule | Always ask for the graveyard | Learning from the visible winners |
| Classic example | Wald armoring where returning planes had no holes | Armoring where the holes were |
| Output | Adjusted odds, risks to avoid | Overconfident recipes for success |
| Tools | Base rates, inversion, comparison groups | None, it is the problem |
| Related | [Lindy effect](https://mentarium.io/mental-models/lindy-effect), which uses survival as a filter on purpose | Hindsight bias, halo effect |

## The Mentarium take

Most advice you meet is survivor testimony. Books, podcasts and posts come from people who made it, and they sincerely credit what they did. Their account may be true and still not useful to you, because the people who did the same things and failed aren’t on the stage. “Always ask for the graveyard” is the cheapest correction we know.

Survivorship bias is on the Bias Radar for business decisions, and it is easy to see why: strategy discussions run on case studies of winners. In the Stress-test step of The Mentarium Decision Loop (Frame → Map → Stress-test → Decide → Review), add one line to your notes: “Who tried this and failed, and what did they have in common?” If you can’t answer, you have found your next piece of research.

Note the contrast with the Lindy effect. Lindy uses survival as evidence on purpose: for ideas and technologies, lasting is the signal. Survivorship warns against using survival as a recipe. Both are true at once. Bet on what has lasted, and don’t assume you know why it lasted until you have seen what didn’t.

**Counters** [Survivorship Bias](https://mentarium.io/biases/survivorship-bias)

## FAQ

### What is the story of Abraham Wald and the bullet holes?

During World War II, the statistician Abraham Wald worked with the Statistical Research Group at Columbia University. Looking at damage on aircraft that returned from missions, he reasoned that the areas without holes mattered most, because planes hit there tended not to come back. The popular version is simplified, but the core logic is his.

### What does 'ask for the graveyard' mean?

It means that before drawing a lesson from success stories, you look for the people or projects that tried the same thing and failed. They are usually invisible because nobody writes about them. Without them, you can't tell whether a strategy works or just survived by luck.

### How is the survivorship model different from survivorship bias?

Survivorship bias is the error: drawing conclusions from the cases that made it through a filter. The survivorship mental model is the habit that fixes it, a set of questions and steps for finding the missing data before you decide. Our survivorship bias page explains the error itself.

### Who said 'silent evidence'?

Nassim Nicholas Taleb uses 'silent evidence' in The Black Swan (2007) for the failures history doesn't record. He retells Cicero's story of Diagoras, who, shown portraits of sailors who prayed and survived a shipwreck, asked where the portraits were of those who prayed and drowned.

### Where does survivorship show up in everyday life?

In business advice built from successful founders, in fund performance tables that drop closed funds, in old buildings that make the past seem better built, and in fitness or diet testimonials. Anywhere the failures quietly leave the sample.

## Related models & biases

[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) [Mental model Inversion Inversion is a mental model that approaches a problem from the opposite end: instead of asking how to succeed, you ask what would guarantee failure, then avoid those things. It also means testing a belief by asking what would have to be true for the opposite to be right. Charlie Munger made it famous in business.](https://mentarium.io/mental-models/inversion) [Mental model Probabilistic Thinking Probabilistic thinking is the habit of treating outcomes as likelihoods rather than certainties: estimating how likely each possibility is, updating those estimates as new evidence arrives, and judging decisions by the quality of the reasoning and the odds at the time, not only by how they turned out.](https://mentarium.io/mental-models/probabilistic-thinking) [Mental model The Lindy Effect The Lindy effect is the idea that for non-perishable things such as books, ideas, technologies and institutions, each extra period of survival implies a longer remaining life expectancy. Popularized by Nassim Nicholas Taleb, it is a mental model for betting on what has already lasted over what is new and untested.](https://mentarium.io/mental-models/lindy-effect) [Cognitive bias Survivorship Bias 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.](https://mentarium.io/biases/survivorship-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)