We don’t have in the data — the government list drops names given to fewer than five babies a year, and plenty of names are simply new. How common does it feel?
Where that lands
Try a room
For scale — coin-flip rooms for other names
How this works
Pick a name and we look up its share of recent US births — of all babies born 2020–2024 were named . Every extra kid in the room is one more independent chance of a namesake, so the odds climb like this:
P(at least one) = 1 − (1 − p)n · coin flip at n = ln(½) / ln(1 − p)
This is not the famous birthday problem. There, any two people matching on anything counts, and pairs pile up fast — 23 people is enough. Here the name is fixed in advance: you’re counting people, not pairs, so the room has to be much bigger.
The data. Names come from the Social Security Administration’s baby name files (public domain), all births 2020–2024 combined: 16.75 million listed births across 43,334 spellings. We ship the 4,000 most common. A name’s share is its count divided by the sum of all listed counts — SSA omits names given to fewer than 5 babies in a year, so true shares run a hair lower than shown.
Spellings don’t merge. Sofia and Sophia — or Aiden, Ayden and Aden — are separate rows here, because that’s how the data comes. To a kid in a classroom they collide, so for heavily-fragmented names the real coin-flip room is somewhat smaller than we report.
And the actual birthday problem? Once you have the whole distribution you can ask it: how many kids before any two of them share any name? Squaring every name’s share collapses 43,334 spellings into an effective pool of about 1,000 equally-common names — so a mixed room of just 38 kids has even odds of a shared name. All boys: about 25. All girls: about 30. The birthday problem’s famous answer is 23 — names cluster almost as tightly as birthdays do.