In a room of 30 kids, what are the odds two of them share a birthday?
After you watchIn a room of 30 kids, what are the odds two of them share a birthday?
The short answer
Much higher than it feels. With just 23 people the chance that some two share a birthday is already better than 50/50 (about 50.7%), and with 30 people it is about 70%. The surprise comes from counting pairs, not people.
Try this next
- What if you only counted matches with YOUR birthday instead of any two kids? Watch the lower bar — the 'matches MY birthday' line. Predict how high it climbs by 30 kids before you check. (It barely reaches 8%.)
- What if the room kept filling — how many kids until a match is almost certain? Drag the room past 50 and then toward 70. Guess where the 'any two match' bar nearly touches the top before you slide there.
- What if a year had only 100 days instead of 365? Think about pairs versus days: fewer days means matches come even faster. Predict whether 23 kids would now be way past 50/50.
Now you — bend it
- What if What if you wanted a 99% chance of a match — how many kids would the room need?Keep filling the room and watch the 'any two match' bar; it passes 97% near 50 and creeps toward 99.9% by 70.
- What if What if real birthdays aren't spread evenly across the year?Uneven days only make matches MORE likely, because crowded dates give pairs extra chances to land together.
- What if What if you counted matching first initials instead of birthdays?There are far fewer 'days' (letters), so the pile of pairs beats them even sooner — predict how few kids it takes.
Can you prove it?A small room hides a huge number of pairs, and pairs are what make matches likely. — Count the pairs yourself: with n kids there are n times (n−1) divided by 2 pairs. For 23 that's 253, for 30 it's 435 — way more chances than 365 days can dodge.
Design your own test:Pick a room size, predict whether a shared birthday is more likely than a coin flip, then slide to that size and see if your guess holds.
Explain it to a 6-year-old: In a big group there are tons of pairs of kids, and with so many pairs it's easy for two to share a birthday — even if it's not yours.
The whole story
How it works
A match needs a pair — two people with the same birthday — and a small group hides a huge number of pairs. In a room of 23 people there are 253 different pairs that could match, and in a room of 30 there are 435. With that many chances, the odds that at least one pair lands on the same day climb past one-half by 23 people and reach about 97% by 50. The number of pairs grows with the square of the group size, so the probability rises far faster than your gut expects.
What people get wrong
People assume a shared birthday in a small group must be rare because there are 365 days, so they picture matching their own specific birthday. Matching YOUR exact day does stay rare — only about 8% even with 30 people. But the real question is whether ANY two people match, which counts every pair in the room, and that becomes likely very quickly.
The catch
Your gut is right that matching one exact day stays rare, which is a true and useful fact. The catch with the pair-counting answer is that it only promises somebody shares with somebody — it does not say anyone matches you, and it cannot tell you which two people it will be, only that a match almost certainly exists.
Questions kids ask
Why does only 23 people give better than 50/50 odds?
Because you are counting pairs, not people. 23 people make 253 different pairs, and each pair is a fresh chance for a match. With 253 chances against 365 days, the odds that at least one pair shares a birthday tip just past one-half.
Does this mean someone in the room shares MY birthday?
No. Matching your one exact day is still rare — only about 8% with 30 people. The high odds are for ANY two people matching each other on ANY day, which is a much easier thing to hit because there are so many pairs.
How many people until a shared birthday is almost certain?
About 50 people gives roughly a 97% chance, and 70 people gives about 99.9%. You do not need anywhere near 365 people, because pairs pile up far faster than the days run out.
Is this a trick or real math?
It is real math, called the birthday paradox. It is only a 'paradox' because the true answer feels too high — but counting all the pairs shows exactly why a shared birthday is so likely in a surprisingly small group.
Talk about it
- Guess first: in our family gathering of about 20 people, would you bet on two sharing a birthday?
- Why do you think matching anybody feels easy but matching YOUR exact day stays hard?
- If you had to pick the smallest room where a match is more likely than not, what number would you guess?
For grown-ups
With n people there are C(n,2) = n(n−1)/2 distinct pairs. Assuming 365 equally likely birthdays, the chance of NO shared birthday is (365/365)(364/365)…((365−n+1)/365); for n=23 that product is about 0.493, so P(at least one shared birthday) ≈ 0.507. By n=50 it is ~97% and by n=70 ~99.9%. The intuition fails because we imagine matching our own birthday — about 1−(364/365)^(n−1), still only ~8% at 30 — instead of all C(n,2) pairs, which grows quadratically.
Keep going
What else makes you wonder?
- If birthdays were spread evenly across the year, would a match get more or less likely than with the real bumpy pattern?
- What other things in your class might secretly match more often than you'd guess — same first letter, same favorite color, same shoe size?
- How big would a room have to get before TWO different pairs share birthdays, not just one?