In a room of 30 kids, what are the odds two of them share a birthday?
Did this make sense?
Your thumbs help us make it clearer.
Thanks! That helps us make it clearer. 💜
See it another way
Check your intuition: The birthday problem - David Knuffke
TED-Ed animates the same any-pair birthday question, derives the no-match complement, and shows why pair opportunities grow so quickly.
Requires internet. You’re leaving Clickory in a new tab; YouTube may show ads or recommendations.
Keep going
What else makes you wonder?
What changes if birthdays bunch up in some seasons?
Picture several cards landing in calendar spaces that are already crowded.
How big must a room be before two different pairs often match?
One match is only the first collision; a second needs another counted event.
Would matching first initials happen sooner than matching birthdays?
Compare how many possible letters and possible dates the pairs can land on.
After you watchIn a room of 30 kids, what are the odds two of them share a birthday?
The short answer
This story keeps two questions separate: does someone match one chosen birthday, or does any pair in the room match? Predict whether neither, both, or only one path crosses one-half in a 30-person room, then earn the exact result by filling both paths.
Try this next
- What if you track only one chosen person's date? Keep one person fixed and count exactly one comparison for every other person.
- What if the room grows beyond the story's slider? Extend the complement calculation one person at a time and watch the no-match product change.
- What if a year had fewer possible dates? Replace 365 in both formulas with a smaller number and predict the direction before calculating.
Now you — bend it
- What if What if a pretend planet had only 100 possible birthday dates?Keep the group fixed and ask what fewer available dates would do to the chance of a collision.
- What if What if birthdays cluster in some seasons instead of spreading evenly?Imagine moving several date cards into already crowded calendar spaces.
- What if What if you matched first initials instead of birthdays?Count the possible labels, then compare one chosen label with any pair of labels.
Can you prove it?One-person checks and all-pair checks grow in different patterns. — Draw a dot for each person. First connect everyone only to one chosen dot. Then start over and connect every distinct pair exactly once. Compare the link counts for several small groups.
Design your own test:Choose a smaller or larger set of possible labels, predict which probability path changes more, then adapt the no-match calculation.
Explain it to a 6-year-old: Checking who matches one person is a different game from checking every pair in the whole group.
The whole story
How it works
For one chosen birthday, each added person creates one new comparison with that date. For the room-wide question, each new person forms a pair with everyone already present. The story calculates the two probabilities separately under a simple 365-equally-likely-days model and lets the reader watch how the paths diverge.
What people get wrong
The phrase 'two people share a birthday' is easy to hear as 'someone shares my birthday.' Those are different events. The story uses distinct colors, labels, and formulas so the personal question never gets blended into the any-pair question.
The catch
A room-wide probability only describes whether some pair exists; it does not tell you which pair, and it never promises that anyone matches one chosen person. The clean calculation also simplifies real calendars by treating 365 dates as equally likely and independent.
Questions kids ask
Is matching one chosen birthday the same as any two people matching?
No. The chosen-birthday question checks everyone against one fixed date. The any-pair question compares every distinct pair in the group. The story keeps both questions visible side by side.
Why count pairs instead of people?
A shared birthday belongs to two people, so each distinct pair is one possible place a match could appear. Adding a person creates links to everyone already in the room.
What assumptions does the birthday model make?
It uses 365 possible dates, ignores leap day, gives each date an equal chance, and treats people's birthdays as independent. Real birthday patterns are a little bumpier.
Is the birthday problem a trick?
No. It is a probability comparison that feels surprising because one personal comparison and every possible pair are easy to mix together. The exact room-size slider runs the calculation directly.
Talk about it
- Which question are we asking: does anyone match you, or does any pair match?
- How does one new person change the number of spokes and the number of all-pair links?
- Before running the final room, do you expect neither, both, or only one meter to cross halfway?
For grown-ups
For n people, the one-chosen-date event uses 1−(364/365)^(n−1). The any-pair event is easier to compute through its complement: 1−∏ from i=0 to n−1 of (365−i)/365. The story renders both formulas deterministically from the committed room-size value.