What this calculator does
The birthday paradox is the gap between the question people think they are asking and the question the math answers. It feels like it should take dozens of people to find a shared birthday, but with 23 people there is already a 50.7% chance. The reason is pairs: 23 people make 253 possible pairings, and every pairing is a small chance to match. This calculator computes the exact probability for the number of people and days per year you enter, then reports the thresholds for 50% and 99%. It also explains what the model assumes, including the standard simplifying assumption of 365 equally likely birthdays and the fact that leap years and real birth-rate patterns move the answer only slightly. It is a probability demonstration, not a forecast about a specific group.
How to use it
- Enter the number of people in the room, including yourself.
- Keep days per year at 365 for the classic birthday problem.
- Read the probability that at least two people share a birthday.
- Check the two thresholds to see how many people are needed for 50% and 99% odds.
- Use the no-shared-birthday figure if you want to see the same result from the other side.
- Switch days per year to 366 to see how a leap-year cycle changes the thresholds.
How it works
Example calculations
Example 1: The classic 23
With 23 people and 365 days, the chance of at least one shared birthday is 50.7%. There are 253 pairs to check, and while each individual pair matches only rarely, the sheer number of pairs makes a match more likely than not.
Example 2: A smaller room
With 10 people there are only 45 pairs, and the chance of a shared birthday is 11.7% - low enough that a match would feel like a coincidence. This is why the 23-person result surprises people: the probability grows quickly once the pair count starts to climb.
Example 3: Pushing to 99%
At 40 people the chance is 89.1%, at 57 people it is 99.0%, and at 70 people it is 99.9%. Each added person brings n new pairs, so the last few percentage points arrive much faster than the first few.
What the result means
The result is counterintuitive because people compare themselves with each other person one at a time, while the math compares every pair at once. With 23 people there are 253 pairs, and even a small per-pair match probability compounds across all of them. The model assumes each birthday is independent and equally likely across 365 days; real births are not perfectly uniform, and a seasonal pattern actually makes a match slightly more likely, not less. Leap years add a 366th possible day, which reduces the probability by a small amount and barely moves the thresholds: with 366 days, 23 people still cross 50% (50.6%) and 58 people cross 99%. Twins, deliberate planning, and hospital clusters break the assumptions in ways the model cannot capture. Use the calculator to build intuition, not to make decisions about a real group.
Common mistakes
- Comparing the entered number of people with 365 days and expecting the probability to be that fraction; pairs grow quadratically, not linearly.
- Assuming the 50% threshold is around 183 people, which is the answer to a different question: the chance that someone shares your specific birthday.
- Forgetting that the model needs independent, roughly uniform birthdays, which twins and planned births violate.
- Ignoring leap years entirely and then treating a 366-day answer as exact; it is a small correction, not a different result.
- Reading the probability as deterministic: 50.7% does not mean a match will happen, it means a match is more likely than not.
Assumptions and limits
- Birthdays are independent of each other across people in the room.
- Each day of the year is equally likely, with no seasonal or weekday pattern.
- The year has 365 days by default; using 366 models a leap-year cycle.
- Every person in the group is counted once, and the group size is known exactly.
- Twins, planned births, and hospital clustering are ignored because they violate independence.
Frequently asked questions
How many people do you need for a 50% chance of a shared birthday?
23 people with 365 equally likely birthdays, which gives 50.7%. That is the classic birthday paradox result and the reason the problem is famous: the answer is much smaller than most people guess.
Why is the birthday paradox true?
Because the number of pairs grows with the square of the group size. Two people make one pair; 23 people make 253 pairs. Each pair has a small chance of matching, but 253 small chances add up to more than a 50% probability of at least one match.
How many people are needed for a 99% chance?
57 people with 365 days give 99.0%. With 366 days it takes 58 people for 99%. The jump from 50% at 23 people to 99% at 57 people shows how the probability saturates as the pairs pile up.
Do leap years change the answer?
Only a little. A 366-day year adds one possible birthday, which slightly lowers the probability and moves the 99% threshold from 57 to 58 people. The 50% threshold stays at 23 people, at 50.6% instead of 50.7%. Real birth-date patterns matter more than the leap day.
What is the difference between sharing any birthday and sharing my birthday?
Sharing any birthday among any pair in the room is the birthday-paradox question, and it needs about 23 people for 50%. Sharing your own specific birthday is a different, much rarer event: you need roughly 253 other people before the chance that one of them matches your exact date reaches about 50%.
Can I use this with a different number of days?
Yes. The days-per-year input lets you model a simplified year with 28 to 366 days, which is useful for teaching or for testing how the thresholds shift. The formula is the same; only the number of possible birthdays changes.
Sources
2026 update
How to use the professional output
This page is not just a formula box. The result is designed to be checked, compared, and taken away.
- Professional worksheet. The result includes the inputs used, the method and scope, and a one-variable sensitivity check for finance calculators.
- A/B scenarios. Save two input sets in the browser and compare the primary, secondary, and breakdown figures side by side.
- Print / PDF and CSV. Print a branded worksheet, or export the year-by-year table when the calculator produces one.
The formulas and assumptions above still govern the number. The worksheet makes the estimate auditable; it does not make it a lender quote, tax return, or personalized advice.
FROM THE EXPERIMENT TO THE REAL ANSWER
If you need the real number
The Lab is for playing. These calculators are built for decisions that matter.