Help us pick her name
The familiars draw two name cards at a time; pick the one you like better, and the standings build below.
Fig. 1
The two names on the table, since 1900
Every American girl given each name, year by year since 1900, from the Social Security records.
Source: Social Security Administration national baby-names dataset (names.zip), births 1900 through 2025, counted from Social Security card applications. Each stream is scaled to births per 100,000 girls that year, so a name's height means the same thing in 1918 and in 2024.
Fig. 2
The standings
Every name starts at a rating of 1,500. Beating a favorite earns more than beating a long shot.
Fig. 3
Where each name ranks in the country and on this ballot
Left, where each name ranks among all American girls born 2015 through 2025. Right, where it stands on this ballot.
National ranks from the same Social Security dataset, births 2015 through 2025. Ballot ranks are the Elo standings in Fig. 2, names with at least one matchup.
How the ranking works
Each pick you make is one matchup, and both sides of it are saved, the name you chose and the name it beat. The standings use Elo, the chess-rating method. Every name starts at 1,500. When a name wins, it takes points from the loser, and upsetting a strong name pays more than beating a weak one, so a name can't climb the board by feasting on easy draws.
We seeded the ballot with the thirty most common American girls' names of 2015 through 2025, from the Social Security Administration's public dataset. The search box reaches the top 10,000, so a name only joins the ballot if real parents have used it.
Two cautions on the figures. Early standings swing hard, since a name that has only played two matchups hasn't earned its rating yet. And the Social Security series undercounts births before 1937, when getting a card became routine, so the left edge of Fig. 1 runs low.