Method before mythology

Data sources & method

The source year matters. A generated name is a prompt; a historical count is a specific aggregate observation.

Given names

The first-name collections adapt U.S. Social Security Administration national birth-name data and a 2020 Census top-name selection. The familiar default contains 1,012 selected given names, emphasizing recent SSA records. The extended bank contains 19,378 distinct spellings after deduplication with the previous collection. New candidates were selected from the national 1880–2025 SSA archive when a spelling appeared in the top 3,000 of a recorded-sex/year field, had at least 250 published births, and appeared in at least five distinct years; obvious placeholders were excluded. Former reviewed entries were retained. Eligible names are drawn uniformly, not weighted by source frequency.

The annual national SSA files came from a 28 September 2026 archive capture of the original SSA names ZIP, because the direct official download returned 403 in our build environment. The archived ZIP has SHA-256 cd78e975ed7bb358e018dd62fbe14ced89295e9581c49172ca4eedcb011b3724. Its 2025 top names match the official SSA list; this does not prove a byte-for-byte match to the current official ZIP. The federal catalog marks national SSA data CC0.

Historical filters for generated given names

Seven generators can limit their given-name candidates by an exact birth year from 1880–2025 or a decade, and by the top 100, 500, or 2,000 published names in the selected period. The 2020s decade contains only the published years through 2025. These filters use the pinned national SSA files and count-sorted positions; they do not change the provenance of separately selected surnames or imply that a generated full name occurred in the source. Small cells are suppressed, so a narrow combination may not have enough names to draw the requested number.

2020 Census first-name frequency

The first-name frequency lookup uses all 53,615 published first-name rows from the 2020 U.S. Census name release. It reports national published count and rank, with male and female aggregate counts where supplied. These figures cover people enumerated in 2020 across ages, not births in 2020. The source applies publication thresholds and disclosure protection; a missing spelling does not imply zero people.

Surnames

The generator’s familiar selection has 1,000 surnames from a mirrored extraction of the 2020 Census frequently occurring surname table. Its extended bank contains 30,006 distinct spellings after combining the published 2020 Census top 30,000 with the previous selection, which includes entries from a 2010 Census copy in a U.S. government repository. A spelling retained only from that 2010 selection is not assigned a 2020 rank.

The surname-frequency lookup separately uses all 156,621 published rows from the official 2020 Census surname workbook, retaining only name, published count, and published rank. The source includes surnames meeting its pre-noise frequency threshold in both 2010 and 2020 and applies disclosure protection noise to 2020 counts; ranks can tie. Absence does not imply zero occurrences or even fewer than 100 in 2020. No individual-level, race, or ethnicity columns are used. See the Census data release and technical report.

Annual popularity records

The lookup searches all 146 national SSA birth-name files from 1880 through 2025. It includes every nationally published entry with at least five recorded births for a spelling and recorded sex field in a year; there is no top-2,000 cutoff. “Position” is calculated by sorting this site’s published entries by count within the same field and year, not an official SSA ranking. No published entry never means zero births: a small count may be suppressed, or the spelling may differ. The SSA background and privacy rules describe the limitations of historical applications. Raw counts are not adjusted for population or application coverage.

State-level births

The state lookup uses 1,045,300 published SSA state-specific birth-name rows for 2015–2025. A blank name field lists 10, 25, 50, or 100 leading spellings for one state or DC, birth year, and recorded-sex field, with a previous-year comparison where published. An entered spelling returns its published count and a position calculated by sorting the released entries by count. State cells under five are suppressed; a missing row is not a verified zero. State counts describe births tied to a state in that year, not residents living there today. Source: SSA national and state file guidance.

Historical mixed-use explorer

For a selected year, browsing requires a published count in both SSA recorded-sex fields, at least five in each and at least 100 combined. The broader filter accepts a female share of 20–80% of those two counts; the closer balance accepts 35–65%. An entered name instead returns all 146 annual female and male field records; missing values are shown as unpublished and share is calculated only when both fields have counts. Neither view makes a claim about anyone’s gender identity.

What the tools do not infer

First-name association labels come from aggregate U.S. sex fields and a broad mixed-use rule; they do not describe a person’s identity. First and last names are published separately, so a generated full-name combination has no measured frequency. We do not infer origin, ancestry, pronunciation, or cultural fit from a spelling. Middle-name, sibling-name, and pen-name generators apply explicit spelling and length rules rather than an inferred compatibility score. The source data and selection rules may omit names you know.

Questions or corrections: info@nameatlaslab.com.

U.S. annual ranked browsing

The boy-name and girl-name browsers show up to the first 1,000 spellings from each pinned national SSA birth-year file, 1880–2025. We calculate the position by sorting published counts and then spelling. A previous-year position is shown only when that spelling had a published row. The ranking list is a selection, not the complete file.

England and Wales birth registrations

The separate England and Wales browser uses the Office for National Statistics 1996–2025 workbook, released 9 July 2026. It shows exact-spelling published ranks up to 1,000 in each of the girls’ and boys’ tables, with counts and previous-year values where available. ONS suppresses counts of two or fewer, and a missing name may also be outside this top-rank selection. This is England and Wales, not all of the United Kingdom. We do not combine ONS births with U.S. SSA records. Contains public sector information licensed under the Open Government Licence v3.0. Questions about the source can be directed to ONS; questions about this implementation can be sent to us.

Nickname collection

The nickname finder has 48 manually selected familiar English-language full-name entries. It is finite editorial content rather than a statistical or etymological source. Forms are suggestions that vary across communities and people. A missing name or short form is not evidence that it is never used. Send corrections and additions to info@nameatlaslab.com.