← MathIsland
23생일 문제 Lab

The shared-birthday odds that pass 50% with just 23 people.

How many people must gather before two of them share a birthday? With 365 possible days, it feels like you'd need a crowd. Run the experiment and your mind will change.

The birthday problem is not about you — it is about pairs.

Experiment

Hands-on experiment

People5

Chance of a shared birthday (any two people)

2.7%

Still under one half. Push the slider up.

🎲 Classroom simulation

Draw random birthdays for this many people and see if any really collide.

📖 Read more — why it exists · insights · common mistakes · formulasExpand ▾

Why

Why does this exist?

Probability is where intuition fails most often. The birthday problem shows that gap at its most dramatic: the answer you feel and the answer you compute are wildly different.

Counting collisions doesn't stop at birthdays. Passwords, ID numbers, and cryptography all hinge on 'what are the odds two random values coincide?' — a core question of modern security.

When a probability is hard to count head-on, count the opposite. 'At least one pair' is messy; 'no pairs at all' is one clean chain of multiplications. That complement trick is this lab's second gift.

Insight

Insights from the video

Your eyes count people; probability counts pairs.

Twenty-three people look like few, but pairing them up gives 253 pairs. Collision chances ride on the pair count, not the head count. Step out of the protagonist's seat and the probability appears.

When 'at least one' is hard, count 'none at all'.

Matching cases come in too many flavors to enumerate. The opposite — everyone different — is a single line of multiplication. Subtract from 1 and you're done. The complement is probability's shortcut.

Misconception

Common misconceptions

365 birthdays, so you'd need about 183 people for a 50% chance.

That intuition computes 'someone matching ME'. The question asks about 'any two people'. There are n people but n(n−1)/2 pairs, so the opportunities are far more numerous.

A match in a small group must be coincidence or a trick.

Repeat the simulation and matches appear about half the time at that group size. It isn't a spooky coincidence — it's the natural consequence of how many pairs there are.

Formula

Writing it as math

The formula only summarizes the experiment. Instead of counting matches head-on, compute 'everyone different' as a chain of multiplications and subtract from 1. The slider you moved becomes the letter n.

The complement

As each person enters, multiply the chance they differ from everyone before. Subtract the product from 1 to get the chance of a match.

Counting pairs

The number of two-person pairs among n people. With 23 people: 253 pairs. Collision chances grow with this number.

The √N estimate

With N possible values, about 1.2√N random picks give a 50% collision chance. N = 365 gives about 23. Security engineers use this exact estimate for hash collisions.

In Real Life

Where you meet it in real life

Hash collisions and security

Computers summarize files and passwords into hashes. However many values exist, about √N attempts yield a 50% collision — a calculation cryptographers call the 'birthday attack'.

Duplicate ID numbers

Randomly assigned IDs, student numbers, and coupon codes collide sooner than expected. Designers of numbering systems compute this probability first.

Lottery numbers repeating

Winning combinations number in the millions, yet as draws accumulate, the chance of a repeat combination grows fast. Several national lotteries have made headlines with repeated winning numbers.

The illusion of coincidence

'Two classmates share a birthday', 'two strangers share a name' — amazing at first, common once you count pairs. You gain an eye for fact-checking the news's 'incredible coincidences'.

Practice

Practice — conquer the types

✏️ 4 practice problems — solve to conquerTap to solve ▾
Conquered 0 / 4
1

In a class of 50, how likely is a shared birthday?

2

What is the real reason 23 people are enough to pass one half?

3

How many two-person pairs do 10 people form?

pairs
4

For a 50% chance that someone shares YOUR birthday, about how many people are needed?

Watch

Related video

Why 23 people are enough — the birthday problemThe video link is coming soonBrowse the YouTube channel →

Connection

Concepts connect

Previous concept

P(A)

Probability

The sense that long repetition reveals a ratio underlies this whole simulation.

← Probability lab

Leads to next

The Monty Hall Problem

The next door where probability betrays intuition — this time with three doors and goats.

Go to the The Monty Hall Problem lab →

Related

Labs worth exploring together

Related lab

nCr

Counting Cases

The pair-counting grammar n(n−1)/2 comes from here.

Go to the Counting Cases lab →

Related lab

aⁿ

Exponents

Even a number near 1 shrinks fast when multiplied many times — the feel of a multiplication chain.

Go to the Exponents lab →