About NameYear
Most of us can hear a name and place it in time. Linda sounds like someone’s grandmother; Madison sounds like a twentysomething. NameYear turns that hunch into a daily game.
Why it exists
First names rise and fall in waves. Linda was at its most popular in 1948, Jennifer in 1974, Madison in 2001. Put six names from the same year side by side and the era usually comes through, even when no single name gives it away. We wanted a quick daily puzzle built on that feeling: something you can play in a couple of minutes, compare with friends, and learn a little from.
Every name statistic on the site comes from public government data. There is no AI-written trivia, and we don’t guess at name meanings or famous namesakes.
How the puzzles are made
A script builds the puzzles ahead of time. Nobody hand-picks rosters. For each day it:
- Picks five answer years between 1935 and 2025, at least 8 years apart, with at least one year from 1965 or earlier and one from 2000 or later. Years used in the past week are skipped, and years before 1950 are drawn half as often because the early data is thinner.
- Draws three girls’ names and three boys’ names for each round from that year’s rankings: the top 10 for round one, deeper cuts later. Every name was given to at least 50 babies that year, has at least three letters, and isn’t on a short blocklist of placeholders like “Unknown” and names tied to slurs or notorious figures. A name is skipped if the other sex used it more than a third as often that year. No name appears twice in one day.
- Runs a fairness check. For every year from 1925 to 2025, it works out how likely it is that six kids born that year would have exactly these names, using each name’s share of U.S. births that year. The roster is only kept if the data’s best guess is within 2 years of the real answer and the range holding 80% of the likelihood is narrow (10 to 18 years wide, depending on the round). The two hint names must keep the best guess within 2 years as well. If a year can’t produce a fair roster after 200 tries, the script picks a different year.
- Saves the puzzle with the answers scrambled, so they don’t show up if you peek at the page source. Once a puzzle has gone live, it stays as it is, even when new data comes out.
That’s also why your results show what the data would have guessed. Sometimes you’ll beat it.
What the data can’t tell you
- It only lists names with at least 5 births in a given year for a given sex. Rarer names, and rare spellings, are left out to protect privacy.
- Counts before 1937 are incomplete. Social Security started in 1936, and many people born earlier never applied for a card, or applied much later in life.
- It’s U.S. only. It covers babies born in the United States who got a Social Security card. Kids born elsewhere aren’t in it, so a real classroom may sound different.
- It comes from Social Security card applications. Names are recorded as written on the application, so every spelling counts separately (Jon and John are different names here), and sex is recorded only as female or male.
The Data page has the full source citation, the years covered (1880–2025 in the national files) and checksums for the exact files we used.
Independent project
NameYear is an independent project. It is not affiliated with or endorsed by the Social Security Administration or any other government agency. Found a bug or have a question about a name? hello@nameyear.com.