The Birthday Paradox: 23 people, even odds
Put 23 strangers in a room and it is a coin flip whether two of them share a birthday. Your intuition says 180. Your intuition is counting the wrong thing.
Filed by The Archivist 1 min read
Intuition test — answer before you read on
How many people are needed before the odds of a shared birthday pass 50%?
Correct answer: A
23 people. There are 253 possible pairs among them, and that is what the probability actually depends on.
Put 23 strangers in a room and it is a coin flip whether two of them share a birthday. Your intuition says 180. Your intuition is counting the wrong thing.
What everyone sees
23 people, 365 days. Obviously the odds must be tiny.
What is actually happening
The question is not whether someone shares your birthday. It is whether any two people in the room share one. With 23 people there are 253 possible pairs, and each pair is another chance to match. Work it as the complement — multiply 364/365 × 363/365 and onwards — and the probability that everyone is unique drops below 50% at exactly 23. At 70 people it is 99.9%.
Why it stays hidden
Because intuition anchors to the wrong reference point: yourself. Once you swap “me and them” for “every pair in the room”, the arithmetic stops being surprising. Cryptographers exploit this daily — it is why hash collisions arrive far sooner than the key length suggests.
You are not counting people. You are counting pairs — and pairs grow much faster than people.
You are not counting people. You are counting pairs — and pairs grow much faster than people.
Collect this card
You are not counting people. You are counting pairs — and pairs grow much faster than people.
0 / 10,000 collected
Cross-references
Related files
Filed near this one in the index.
-
No visual on fileStatistical Illusions Entry #0437
The reason a study that replicates is worth two that do not
A single study is a claim. A replication is a test of that claim. Two unreplicated studies are two untested claims, not twice the evidence.
AdeptThe hidden part #0437Novelty fills journals. Replication fills knowledge. The incentive structure rewards the wrong one.
Statistics Open file -
No visual on fileStatistical Illusions Entry #0434
Why a percentage change needs its starting point
A fifty per cent increase from two is one. A fifty per cent increase from two million is one million. The percentage is identical; the meaning is not.
AdeptThe hidden part #0434A percentage without a base is not a statistic. It is a frame looking for a denominator the reader will supply.
Statistics Open file -
No visual on fileScarcity & Queues Entry #0462
The reason an invitation raises acceptance more than an offer
An offer says the item is available. An invitation says the recipient was selected. The difference is not in the item but in the identity it assigns.
AdeptThe hidden part #0462An offer presents a product. An invitation presents an identity. People accept identities faster than products.
Scarcity Open file -
No visual on fileScarcity & Queues Entry #0458
The reason an expiring discount beats a larger permanent one
A permanent reduction can be acted on at any time, which means it can be postponed indefinitely. A deadline converts an intention into a dated task.
AdeptThe hidden part #0458An open offer can be postponed forever. The deadline is not pressure; it is the thing that gives the intention a date.
Scarcity Open file