How the odds are calculated

The calculator answers one question: of all the adults you could be looking for, what share meets every standard you set? This page explains exactly how it gets there, where each number comes from, and where it’s likely to be off.

The short version

1. For the U.S., age, race, marital status and income are counted together, from the same people, in the Census Bureau’s own survey records. “Unmarried and earning $100,000 at 25” is a real count, not two percentages multiplied.
2. Everything else (height, build, children, orientation, religion and gambling) is looked up for each single year of age and sex, then applied on top.
3. The answer is shown two ways: as a percentage of people in your chosen age range, and as an estimated number of real people.

A worked example

Say you’re looking for a man aged 25 to 35 who is at least 6 feet tall, earns at least $100,000 a year, and isn’t married.

Notice the last row. Multiplying 20.1% by 63.9% would say 12.9%. But married men earn more, so the real overlap is only 8.8%. Applying height on top leaves about 1.60% of men in that age range. Across the U.S., that works out to around 408,000 people.

That’s roughly 1 man in 63 in that age range. Every standard you add shrinks the pool again, which is why results can fall so quickly.

Where every number comes from

All figures are for U.S. adults unless you switch to another country. The shares below are for all adults; the calculator itself uses each figure for the exact ages you pick, because almost all of them change with age. Nothing is invented to fill a gap.

A note on “build”

The quiz lets you pick a build: normal, athletic or obese. But only obesity (a BMI of 30 or more) is actually measured nationally. “Athletic” isn’t a BMI category at all: a weightlifter and someone who never exercises can have the same BMI. So the calculator starts from the real obesity figure and then estimates that roughly a third of non-obese men and a quarter of non-obese women are athletic, based on the share of adults who meet federal guidelines for both cardio and strength exercise. That split is an assumption rather than a measurement, and it’s the least certain number on this page.

Other countries

The Global Dream Partner Report runs the same calculation for 198 countries. For 177 of them it uses that country’s own figures, drawn from national statistics offices, UN population data, the World Health Organization, UN marriage data and World Bank and IMF income data. For the other 21, where reliable national figures aren’t published, it combines the country’s real population with its region’s typical profile, and the report labels those as regional estimates rather than national statistics.

Most countries don’t collect race or ethnicity data at all (France and much of Europe are barred from doing so by law), so that filter only appears where a country actually publishes it.

Where the estimates can go wrong

This is a real calculation from real data, but it is an estimate. Worth keeping in mind:

• Some traits are still combined as if unrelated. Age, race, marriage and income are counted together from real people. Height, build, orientation, religion and gambling come from different surveys, so they’re applied on top for each age as if they had nothing to do with income or marriage. Children is tied to marital status but not to income. Taller men do earn a little more on average, so for a height-plus-income search the true figure is probably slightly higher.
• Surveys are samples. The Census survey behind the income and marriage counts covers about 108,000 adults, so very narrow searches (a single year of age, a rare combination, a $500,000+ income) rest on few people and are least precise.
• Height is a smooth curve for each age group. It matches the measured heights within about one percentage point at every height, but the very tallest heights are the least precise.
• A couple of figures don’t split by everything. Religion is by age but not by sex, and gambling is the same at every age. The men’s children figures are from 2014, the latest the Census Bureau has published.
• States and metros are estimates relative to the nation. They start from the national count for your search and are adjusted by how that place’s own published figures differ from the national average.
• It counts people, not availability. “Not married” isn’t the same as single, and nothing here knows who’s looking, who’s nearby or who’d be interested.

What happens to your answers

The calculation runs entirely in your browser. Your answers aren’t sent anywhere to work out the result. We do record which filters get used, anonymously, to see what people are actually looking for. The privacy policy has the details.