Starlight Tools

Birthday Paradox Calculator — Probability of Shared Birthdays

Pick a group size and instantly see the chance that at least two people share a birthday. Private by design — all calculations run locally in your browser.

Fast answer: with the classic 365-day model, 23 people gives about a 50.73% chance of at least one shared birthday.

23 people50.73% chance
47 people95.48% chance
57 people99.01% chance

Calculator mode

Probability for a group

Tip: 23 people is about 50.73% with the 365-day assumption.
Use custom outcomes to model hash buckets, random IDs, rooms, months, or categories.

Results

Required group size

Smallest Group for Common Thresholds

Target ProbabilitySmallest Group Size
≥ 50%—
≥ 75%—
≥ 90%—
≥ 99%—

Outputs are rounded for readability. Exact computations use high-precision log products to avoid underflow.

Probability Curve

Shared-birthday probability rises quickly as the number of pair comparisons grows.

Monte Carlo Simulation

Run random groups and compare the simulated result with the theoretical value.

Theoretical: —
Simulated: —

Common Values

These values use the current birthday model or custom outcome count.

Group size P(shared birthday) P(no match) Expected matching pairs

How the Birthday Paradox Calculator Works

The birthday paradox asks for the chance that at least two people in a group share a birthday. The standard method is to calculate the easier complement first: the probability that everyone has a different birthday.

P(no match) = product from k = 0 to n - 1 of (D - k) / D
P(match) = 1 - P(no match)
For D = 365: P(match) = 1 - 365! / ((365 - n)! * 365^n)
Expected matching pairs = n(n - 1) / (2D)

For n = 23 and D = 365, the no-match product is (365/365) x (364/365) x (363/365) x ... x (343/365). That product is about 0.4927, so the chance of at least one shared birthday is 1 - 0.4927 = 0.5073, or 50.73%.

Methodology and accuracy: the default model assumes uniform, independent birthdays over 365 days. The 366-day option treats leap day as equally likely, and custom outcomes use the same generic collision model. Calculations use log-space products for numerical stability. Last reviewed and updated: June 7, 2026.

The expected matching pairs result is separate from the probability of at least one match. Probability answers "does any collision happen?" while expected pairs estimates the average number of matching pairs over many repeated groups.

How to use it step by step

  1. Enter the group size (number of people).
  2. Choose 365 days, 366 days, or custom generic buckets / outcomes.
  3. Review the probability, no-match complement, odds, and expected matching pairs.
  4. Switch to Required group size to find the smallest group for 50%, 90%, 95%, 99%, or any target.
  5. Use the common-values table, probability curve, or simulation for comparison and teaching.

Real-world uses include teaching probability in classrooms, explaining why collisions happen in hashing, or understanding why duplicate events appear in large datasets. The same logic shows up in cybersecurity when discussing hash collisions and in analytics when checking for overlapping dates. It is also a great conversation starter for team events or workshops because the result feels counterintuitive.

For example, with just 23 people, the chance of a shared birthday is about 50%. By the time you reach 50 people, the probability is already above 97%. These numbers highlight how quickly combinations grow as group size increases. The calculator uses precise math internally so the results stay accurate even for large groups.

Birthday Paradox: FAQs

Why is it called a paradox?

Because the result feels surprising: our intuition often confuses the question “someone matches someone” with “someone matches me.” The calculator answers the former.

How many people are needed for 95%?

Under the 365-day model, 47 people are needed for at least a 95% chance. At 46 people the probability is about 94.83%, and at 47 people it is about 95.48%.

Is this the same as someone sharing my birthday?

No. This calculator checks whether any pair in the group shares a birthday. Matching one specific person's birthday is a different and much smaller probability.

Does at least one match mean everyone shares a birthday?

No. It means at least one pair matches. A group can have one matching pair while everyone else has unique birthdays.

Why is 183 not the answer?

183 feels natural if you are thinking about half of 365 or matching one fixed birthday. The paradox counts every possible pair in the group, and the number of pair comparisons grows as n(n - 1) / 2.

Does it assume birthdays are equally likely?

Yes. Real-world data varies by month and region, but the uniform assumption is standard for the paradox and keeps the math transparent.

Do real birthday distributions change the result?

They can change it slightly because births are not perfectly uniform across all calendar days. The classic 365-day model is still the usual reference because it is simple, transparent, and close enough for the headline paradox.

How does this relate to hash collisions?

Hash buckets, random IDs, rooms, months, and categories are all versions of the same collision problem. When many items are assigned to a limited number of outcomes, duplicate outcomes appear sooner than simple intuition suggests.

What if the group size exceeds the number of days?

If n > D, then by the pigeonhole principle the probability is 100%.

Is my data private?

Yes. All calculations run locally in your browser; no data is uploaded or stored on a server.

5 Fun Facts about the Birthday Paradox

52 people ≈ 97%

Double the “23 people” headline and you’re nearly guaranteed: 52 people is about a 97% chance of a shared birthday.

Quick math

Real data can shift

Birthdays are not perfectly uniform in real populations, so observed odds can differ slightly from the clean 365-day teaching model.

Real-world skew

Poker paradox cousin

The same math explains why pair collisions in hash functions happen sooner than intuition says—just like matching birthdays in a crowd.

Crypto parallel

Leap day barely moves it

Including 29 February keeps the 50% threshold at 23 people; it only nudges the 99% threshold from 57 to 58 people.

Leap tweak

Three-way matches

The chance of at least three people sharing a birthday passes 50% around 88 people (365-day assumption).

Higher collisions

Explore more tools