About the Birthday Paradox Room
Twenty-three people in a room and it is more likely than not that two of them share a birthday. Nobody believes it until they see it, which is why this tool makes it personal: add the real birthdays of your family, your team, or the classroom, and the odds recompute with every name — the exact formula, not a simulation — while the people who actually match light up in the room.
The engine is the one every probability textbook uses, checked against the published values and against sixty-thousand-trial random cross-checks before it shipped. The subscriber layer goes where the textbooks stop: near misses (birthdays a day or two apart, with their own exact 1970 formula — 50% at just 14 people), the full odds curve with a solver for any target, and the genuinely rare three-way match (50% at 88). Nothing you type leaves your browser; the room is kept locally so it survives a refresh.
Birthdays within a day or two of each other, on the circular year (31 December and 1 January are neighbours). The odds use the exact Abramson–Moser formula, so they are as honest as the headline.
How to Use the Birthday Paradox Room
Add people one at a time — a name if you like, then month and day — and watch the headline. The number is the exact probability that at least two people in a room of that size share a birthday; the second card tells you whether your room actually has a match and lists the pairs. The “one more person” card is the one to read aloud at a party: it says what the odds become if the next guest walks in. The room is kept in your browser, so you can build it over an evening.
Why Intuition Fails Here
People instinctively answer a different question: “how likely is it that someone shares my birthday?” That one really does need a crowd — 253 people for even odds. But the paradox asks whether any two match, and the number of pairs grows with the square of the group: 23 people make 253 pairs, 50 make 1,225, 70 make 2,415. Each pair is a 1-in-365 shot, and thousands of 1-in-365 shots do not miss. The exact calculation multiplies the chance that each new arrival avoids everyone already present — 364/365, then 363/365, and so on — and the product collapses faster than anyone expects.
The Numbers Worth Memorising
- 23 people — 50.7%, the famous crossing point.
- 30 people — 70.6%; a typical classroom is a better-than-two-thirds bet.
- 41 people — 90.3%.
- 57 people — 99.0%; at a wedding, a match is nearly certain.
- 70 people — 99.9%.
- 367 people — 100%, by the pigeonhole principle, since there are only 366 possible days.
Near Misses Are Even More Common
Loosen “the same day” to “within a day” and the crossing point drops from 23 people to 14; within two days, to 11; within a week, to just 7. This is the version people actually experience — “mine’s the day after yours!” — and it has its own exact formula, published by Abramson and Moser in 1970 for birthdays on a circular year. The subscriber panel applies it to your room and lists the near-miss pairs, which is usually where a small group finds its first surprise.
The Assumptions, Stated
The formula treats 365 days as equally likely and ignores 29 February. Real births are not uniform: in the United States they peak in late summer and dip sharply on Christmas Day and other holidays, when scheduled deliveries are avoided. Uneven days make coincidences slightly more likely, so the textbook figure is a conservative floor, and the difference is a fraction of a percentage point at 23 people. Leap-day birthdays are accepted in the room and counted in the real matches; they simply are not modelled in the probability, which is the standard convention.
Where the Paradox Actually Matters
The same arithmetic governs hash collisions in computing (a 64-bit hash has an even chance of a collision after roughly 5 billion items, not 18 quintillion), DNA-profile coincidences in forensic databases, and the surprising frequency of “impossible” coincidences in any large data set. The birthday room is the friendliest way to feel the effect in your gut; the lesson — that pairs multiply while intuition counts individuals — is the transferable part.
Related tools: Birthday Twins for who famous shares your exact date, The Day You Were Born for the weather and moon that day, and Coin Flip for another fair-randomness engine from the same family. Browse every Fun & Novelty tool for more.
Exact closed-form probabilities for a 365-day uniform year (near-miss odds via Abramson & Moser, 1970; triple odds via direct counting). Validated against published values and Monte Carlo before release. Birthdays you enter stay in this browser; the share link carries only the room’s size and headline numbers unless you choose to include names.
Frequently Asked Questions
Why does it only take 23 people to reach 50%?
Because the question is not "does anyone share MY birthday" but "does ANY pair share a birthday" — and pairs multiply fast. Twenty-three people make 253 pairs, each with a 1-in-365 chance, and 253 chances at 1-in-365 add up to better than even. The exact figure is 1 minus the probability that all 23 birthdays are different, which is 365/365 × 364/365 × 363/365 … × 343/365 = 49.3%, so a match sits at 50.7%. By 41 people it is 90%; by 57 it is 99%; by 70 it is 99.9%.
Is this tool using the real formula or a simulation?
The real formula, computed exactly in your browser: for n people and 365 equally likely days, P(no match) is the product of (365 − i)/365 for i from 0 to n − 1, and the tool reports 1 minus that. It was checked against the published values (23 → 50.7297%, 50 → 97.0374%, 70 → 99.9160%) and against 60,000-trial Monte Carlo runs before shipping. The near-miss and triple probabilities use exact closed forms too — no random simulation stands in for a number anywhere on this page.
My room has a match at only 12 people. Was that lucky?
At 12 people the exact probability of at least one shared birthday is 16.7%, so it happens in roughly one room in six — noticeable, not remarkable. What feels remarkable is the comparison with the intuition most people bring ("surely 1 in 30 or so"), which is the whole point of the paradox. The tool shows both the odds for your exact room size and what one more person would do to them, so "lucky" has a number attached.
What counts as a near miss, and are those odds different?
Birthdays within a day or two of each other on the circular year (31 December and 1 January are one day apart). The odds are far higher than for exact matches: a match within one day passes 50% at just 14 people, within two days at 11. The subscriber near-miss panel uses the exact 1970 Abramson–Moser formula for "no two birthdays within k days", so the near-miss odds are as honest as the exact ones — and it is why a party of a dozen almost always has two people saying "mine's the day after yours".
Does the maths assume birthdays are evenly spread? They aren't.
Yes, it assumes all 365 days are equally likely and ignores 29 February, which the tool states plainly. Real US births cluster in late summer and dip on holidays, and uneven days make a match slightly MORE likely, not less — so the uniform figure is a conservative floor. The effect is small (a fraction of a percentage point at 23 people); the 365-day formula is the standard textbook version for good reason, and the room's actual matches are counted from the real dates regardless.
