Methodology and Data Sources

The short version

For each single year of age in your range, the calculator looks up the share of US adults of that sex and age who meet each preference, using official government tables, and multiplies those shares together.

It then adds up the matching people across every age to get the total, and divides by the size of the whole group. The result is shown as a percentage, as "1 in X" and as an approximate number of people.

Data sources

Every figure comes from a public table or public-use data file published by a US government agency. We do not use our own surveys, user submissions or invented numbers. Each dataset is stored in the site's code together with the name of the table it came from, and is rebuilt by a script when the agency releases new data.

Height and obesity are calculated from the NHANES public-use files, which contain measurements taken by trained staff rather than self-reported figures, using the survey's examination weights. As a check, our obesity figures for adults 20 and over match every age and sex figure published in NCHS Data Brief No. 508.

How filters are combined

1. Split the group by age. The calculation covers US adults of the sex you are looking for, one year of age at a time, across the range you chose (anywhere from 18 to 80).
2. Look up each filter at each age. For every age, we take the share who meet each preference from the source band that contains that age.
3. Multiply within each age. At each age, the shares are multiplied together. For example, if half of 30-year-olds meet one preference and a fifth meet another, the estimate for both is one tenth of 30-year-olds.
4. Add up across ages. Each age's share is applied to the number of people of that age, and the results are summed. That total is the headcount, and dividing it by the size of the whole group gives the percentage and the "1 in X" figure.

Working age by age keeps traits that change with age paired up correctly. Younger adults are more often single and older adults more often earn high incomes, so multiplying range-wide averages would pair the singleness of 22-year-olds with the incomes of 45-year-olds. It also guarantees that widening your age range never lowers the number of matching people, although the percentage can fall when the added ages match less often.

The "pass on its own" figure for each filter on the results page is the share of the whole group who meet that one preference, and "match if dropped" repeats the age-by-age calculation without it.

Changed on 23 October 2026: earlier versions multiplied range-wide averages. Results for ranges that span several source age bands changed slightly; for example, single men 6'0" or taller earning $100,000 or more went from 2.49% to 2.33% for ages 25 to 37.

Matching your age range to the source tables

Each source groups ages into its own bands, such as 25 to 29 or 45 to 64. Every single year of age takes its figure from the band that contains it, so a 27-year-old uses the 25 to 29 figure for marital status and the 25 to 34 figure for education.

The number of people at each age comes from the Census Bureau's sex by age table, spread evenly across the years of each band, which are at most five years wide. Height and obesity use the three age bands NCHS reports: 18 to 39, 40 to 59, and 60 and over.

How each filter is defined

• Height: the share whose measured height is at least your minimum and, if you set one, no more than your maximum.
• Income: total personal income before tax in the previous calendar year, from all sources, in $2,500 steps up to $100,000. People with no income are included in the group, so they count as below any minimum. The Census table does not split incomes above $100,000 by age, so we do not offer higher thresholds rather than estimate them.
• Singles only: people who have never married, are widowed or are divorced. People who are separated are still legally married and are counted as married, as in the Census table.
• Exclude obese: removes people with a measured body mass index of 30 or more, the standard CDC definition of obesity.
• Education: the share whose highest completed level is the one you choose or higher.

The independence assumption

Multiplying shares assumes that the filters are independent of each other within the chosen sex and age group: for example, that knowing someone's height tells you nothing about their income. In reality, some traits are linked. Education and income are strongly related, and height and income are weakly related. Official tables rarely publish these traits cross-tabulated together, so we cannot measure the overlap directly.

Because of this, results are best read as an approximate order of magnitude rather than an exact figure. When filters are positively related, as education and income are, the true share is likely to be somewhat higher than the estimate.

Rarity tiers and the score

The 1 to 10 score reads the final share on a logarithmic scale. A score of 1 means everyone in the group matches, 4 means about 1 in 10, 7 means about 1 in 100, and 10 means about 1 in 1,000 or rarer.

Results show the score as a rarity tier: Common (1 to 2), Uncommon (3 to 4), Rare (5 to 6), Epic (7 to 8) and Legendary (9 to 10). Tiers describe how rare a combination of preferences is. They are not a rating of the person setting them.

The person grid shows the same share as icons: one icon per percent out of 100, or, below 1%, one highlighted person in a crowd of X. Very large crowds are capped at 1,000 icons, and the caption says how many such grids it would take to find one match.

Very rare combinations

Height comes from a survey that measured a few thousand adults, so very tall or very short thresholds rest on only a handful of people. When fewer than 30 surveyed people meet a height filter, we do not show exact-looking figures.

• If the estimate is below 1 in 1,000, the result reads "fewer than 1 in 1,000, too rare to measure reliably in the CDC survey" and gets the rarest tier, Legendary. For example, almost no women aged 18 to 39 in the survey measured 6'0" or taller.
• Otherwise, the figures are shown with "about" and marked as a rough estimate. For example, men 6'4" or taller are uncommon but not that rare.

Other limitations

• US adults only. All figures describe the adult population of the United States.
• Survey data. The sources are large, well designed surveys, but every survey has sampling error, and income is self-reported.
• Different survey years. Agencies publish on their own schedules, so filters come from different years. Each source above names its survey year.
• Youngest income band. The income table's youngest band is 15 to 24, so for ages 18 to 24 it includes teenagers under 18, most of whom have little or no income. Income shares for ranges that start below 25 are therefore likely to be a little low.
• No judgment intended. A small match pool describes how common a combination of traits is. It is not a statement about anyone's worth or chances.

Corrections and updates

We review the sources when agencies publish new releases and update the data when they do. If you think a figure is wrong, please contact us with the page and the source you are comparing it with.