Availability and survivorship
What comes to mind easily isn't what happens often, and the cases you can see are the ones that made it through. Learn to count instead of recall, and to go looking for who is missing from the picture.
The availability heuristic
Judging how likely something is by how easily an example comes to mind. Amos Tversky and Daniel Kahneman described this in the 1970s. It's a decent shortcut, because common things usually are easy to recall. It breaks whenever something is easy to recall for a different reason: it was recent, it was frightening, it happened to you, or it was repeated for a week.
After a few days of coverage about a rare illness, waiting rooms fill with worried people. Nothing about the illness changed. Only how easily it comes to mind.
WHAT STICKS
Five things your memory promotes
Recent: this week beats last year. Vivid: a face and a name beat a percentage. Frightening: fear writes in bigger letters. Repeated: five reports of one event feel like five events. Yours: what happened to you outweighs everything. Notice that 'common' isn't on the list. News is made of rare events, because common events aren't news.
'Nobody reads books any more' usually means 'nobody in the last three conversations I had'.
Check yourself
Hooman is deciding what to do with his savings. A relative lost money in a failed fund two years ago and still brings it up at every gathering, so Hooman keeps everything in cash at home. What is availability doing here?
- Nothing, because a real example from a relative beats a statistic
- One loud, repeated story is standing in for a rate
- It is making him too hopeful about funds
- Nothing, because with savings, being careful is always the safer error to make
Show the answer
One loud, repeated story is standing in for a rate
Yes. One vivid case has replaced the question 'how often does that happen?'. Losses that arrive slowly and silently, like money losing value or being stolen at home, never get retold at a gathering, so they carry no weight.
Feels common, is common
What feels common
Whatever you can picture fastest: the story a friend told on Friday, the headline, the one time it happened to you. Fast, emotional, and shaped mostly by what gets reported and repeated.
What is common
A count. How many out of how many, over what period. Dull, hard to picture, and usually findable if you ask. Lesson 4 gave this a name: the base rate. Availability is what fills the gap when you don't go and get one.
Check yourself
Elham has seen three videos about people cheated while renting a flat online. Her friend says this shows the risk is high. Elham answers that three videos show the scam exists and makes good video, but say almost nothing about how often it happens. Elham is right.
Show the answer
True
She is. Easy recall tells you something exists and is memorable; it can't tell you a rate. Frightening, repeatable stories are exactly the ones that get filmed and shared, which is why they feel common. To know how often, she needs a count, and then the specific checks that actually stop it: seeing the flat, meeting the owner, paying nothing in advance.
Survivorship bias
Drawing a conclusion from the cases that made it through, when the ones that didn't are missing from the picture by design. The failures don't hide on purpose. They simply stop being visible: shops close, channels stop posting, people who took the other road never get interviewed.
Every business book on the shelf was written by someone whose business worked. Founders who did the same things and closed didn't get a book deal, so you can't check whether the advice is what made the difference.
Check yourself
Which of these pictures is missing its failures, and which includes everyone?
- 'Everyone I know who moved abroad for work is doing well'
- A course that surveys all 300 people who enrolled, finishers or not
- 'Old furniture was better made, look how these pieces have lasted'
- A clinic reporting results for every patient who started the treatment
- An evening spent reading interviews with people who quit to travel
- A landlord's record of every tenant in the building over ten years
Show the answer
Missing its failures: 'Everyone I know who moved abroad for work is doing well', 'Old furniture was better made, look how these pieces have lasted', An evening spent reading interviews with people who quit to travel
Includes everyone: A course that surveys all 300 people who enrolled, finishers or not, A clinic reporting results for every patient who started the treatment, A landlord's record of every tenant in the building over ten years
WHERE THEY HIDE
Why the past looks better than it was
'They don't make things like they used to.' The forty-year-old fridge still humming in someone's kitchen is real. So were the thousands from the same year that broke and went for scrap, and nobody kept those. The same trick makes old streets look lovelier than new ones: the ugly old buildings were pulled down. You are looking at a filtered sample and calling it a golden age.
It works on people too. Advice from those who made it, a feed full of good news, a friend's 'everyone is buying right now'. Same shape every time: the ones it didn't work for left the room quietly.
Check yourself
- Saeed: 'Four people from my class left engineering for design and all four are happier. Design must be the better field.'
- Saeed: 'Four left for design and are happier. I also tracked down three who tried it, couldn't find work and went back. Now I want to know what those three did differently.'
What changed between the first thought and the second?
- He stopped trusting his classmates
- He collected a larger number of examples
- He found the cases that didn't make it, so his sample stopped being only winners
- He concluded that design is the worse field
Show the answer
He found the cases that didn't make it, so his sample stopped being only winners
Yes. Seven people instead of four isn't the point; the point is that three of them failed. A sample that can only contain successes can't tell you how likely success is.
Four questions that bring the missing back
- How many tried?
Behind every winner there is a denominator. Ten people you watched succeed, out of how many who started? If nobody can tell you, that itself is the answer.
- Where would the failures be now?
Closed shops, dormant channels, people back at their old job. Name the place, then find two of them and ask what happened.
- Would I have heard if it had gone badly?
If the answer is no, you're looking at survivors. Nobody posts the flat that turned out to be damp.
- Count, don't recall
When something feels common, ask for a count instead of reaching for an example. In your own life too: how many times in the last year, actually?
Check yourself
Match each belief to the cases that are missing from it
Show the answer
- 'Every founder I follow works eighty-hour weeks' → The ones who burned out stopped posting
- 'Appliances from that decade never break' → The ones that broke were scrapped years ago
- 'Crime is out of control, it's on the news nightly' → News reports rare events, not common ones
- 'This programme works, look at the before-and-after photos' → People it did nothing for never sent photos
Lesson recap
- Your mind judges how likely something is by how easily an example comes to mind, and easy recall mostly means recent, vivid, frightening, repeated or personal.
- News and feeds are made of rare events, so they are a bad guide to what is common.
- Survivorship bias is the other half: the cases you can see are the ones that made it, and the rest stopped being visible.
- Ask how many tried, where the failures are now, and whether you would have heard if it had gone badly.
- When something feels common, get a count instead of reaching for an example.