23

In a room of just 23 people,
the chance of a shared birthday
exceeds 50%.

Most people guess "around 183."
The real answer is 1/8th of that. See how far your intuition is off.

StatPlay Columns The Birthday Paradox

"How many people does it take for two to share a birthday?"
With 365 days in a year, you'd think you need a lot.
But probability gives a completely different answer.

01

// Guess First

For a 50% chance of a shared birthday, how many people do you need?

50 people

Why does this happen? Try the simulation below.

02

// Hands-On Simulator

Add people one by one. Each birthday lights up on the 365-day calendar.
When two dots share a day, that's a match.

0 people

Auto-starts on scroll

03

// Why? — The Pair Explosion

The reason your intuition fails is simple.
You think about "me vs. someone," but it's actually "anyone vs. anyone" — every pair counts.
Increase n and watch the lines (pairs) explode.

5
Pairs = n(n−1)/2 10
23 people → 23 × 22 ÷ 2 = 253 pairs
With 253 chances for a match, hitting at least one is more likely than not
04

// Probability Curve

X-axis: people. Y-axis: P(at least one match). Watch the curve cross 50% at n = 23.

23
People 23
Match probability %
Intuition
P(match) = 1 − (364/365) × (363/365) × … × ((365−n+1)/365)
Subtract the "all different" probability from 1 — the complement rule in action
05

// 1,000 Trials

Theory is one thing. Run 1,000 actual simulations and see where
the first match tends to land.

Trials 0
Avg first match

// KEY TAKEAWAY

FAQ

// FAQ

In a group of 23 people, the probability that at least two share a birthday exceeds 50%. It's called a "paradox" not because of a logical contradiction, but because human intuition badly underestimates the probability.

With 23 people there are 23×22÷2 = 253 possible pairs. People think about "me vs. someone," but the real question is "anyone vs. anyone." The explosion of pairs drives the probability past 50%.

Over 99.9%. At 57 people it's 99%, at 40 it's 89%. The probability rises steeply with each additional person. Try the curve slider above to see for yourself.

The Birthday Paradox itself is rarely tested, but the underlying concepts — the complement rule P(A) = 1 − P(Aᶜ) and multiplication of independent probabilities — are fundamental and frequently tested.

Master the complement rule, and probability stops being scary.

Dive deeper into the probability concepts from this page.
StatPlay's interactive tools let you learn statistics by touching.