In a room of just 23 random people, there's a better-than-50% chance that two of them share a birthday. Not a certainty — but more likely than not, with a year of 365 days to spread across.
How many people do you need in a room before it's more likely than not that two share a birthday?
Most people guess somewhere around 180 — half of 365. The real answer is 23.
Wait, That Can't Be Right
It feels wrong because of a very natural mistake: thinking about your birthday matching someone else's. If you ask "what's the chance someone else in this room shares my birthday?", you do need a lot of people — dozens, even hundreds — because every comparison is anchored to one fixed date.
But the Birthday Problem isn't asking that. It's asking whether any two people at all share a birthday — and with 23 people, there are far more pairs to check than you'd think.
Here's Why: Count the Pairs, Not the People
With 23 people, how many different pairs of people are there? Every person can be paired with every other person, which works out to:
(23 × 22) ÷ 2 = 253 pairs
253 separate chances for a birthday match, from just 23 people. That's the whole trick: the number of pairs grows much faster than the number of people.
The actual calculation works by finding the probability that nobody shares a birthday, then subtracting from 1. The first person can have any birthday. The second must avoid the first's (364/365 chance). The third must avoid both (363/365). And so on, all the way down to the 23rd person avoiding the 22 birthdays already taken (343/365).
Multiply all 23 of those fractions together, and the chance that nobody shares a birthday comes out to about 49.3%. Subtract that from 100%, and the chance that at least two people do share a birthday is about 50.7% — just past a coin flip.
Try It Yourself
The probability climbs faster than most people expect:
10 people → 12% chance of a shared birthday
20 people → 41%
23 people → 51% (the tipping point)
30 people → 71%
50 people → 97%
57 people → 99%
By 57 people — fewer than a typical school class and a bit — a shared birthday is almost guaranteed. You never need anywhere close to 365 people, because you were never checking one birthday against the calendar. You were checking every person against every other person, all at once.
The OddMaths Takeaway
The Birthday Problem feels impossible because our intuition quietly swaps the real question ("do any two of these people match?") for a much harder one ("does anyone match me specifically?"). Once you count pairs instead of people, 23 stops looking tiny and starts looking exactly right: 253 separate rolls of the dice, all trying to land on one of only 365 slots.
It's a genuinely useful piece of maths too — the same logic underpins real-world collision problems, from hash-table clashes in computer science to how quickly password or ID systems need extra digits to stay unique.
Quick Check
- With 23 people, how many distinct pairs of people are there?
- Roughly what's the probability that two people share a birthday in a room of 23?
- Why does the Birthday Problem need far fewer than 365 people to reach a 50% chance?
- True or false: you need over 100 people before a shared birthday becomes more likely than not.
Answers
- 253 pairs — (23 × 22) ÷ 2.
- About 50.7%, just over a coin flip.
- Because the question is about any two people matching, not one specific person — and the number of pairs to check (253, for 23 people) grows much faster than the number of people.
- False. The 50% mark is reached at just 23 people, not 100+.