Methodology and data sources
The calculator produces an illustrative estimate of how common a selected combination of demographic traits is in the U.S. adult population. It is not a live Census query, a database of available partners, or a peer-reviewed prediction. This page separates the official benchmarks from the assumptions added by our model.
Source register
Government agencies listed below do not endorse this site. Their publications are reference points; the model's fitted distributions and combinations are our own.
| Model input | Official reference | How version 1.0 uses it |
|---|---|---|
| Age and population | 2024 American Community Survey, S0101: Age and Sex | Benchmark for the size and age structure of the U.S. population. |
| Personal income | U.S. Census Bureau, Current Population Survey personal-income tables | Benchmark for income differences by age and sex. Version 1.0 fits simplified log-normal curves rather than querying these tables live. |
| Marital status | U.S. Census Bureau, Historical Marital Status Tables | Benchmark for sex differences in current marital status. The current age-band rates are simplified approximations. |
| Height | CDC/NCHS FastStats: Body Measurements | Benchmark for measured adult height from NHANES. Height is not a Census Bureau variable. |
| Obesity | CDC/NCHS Data Brief 508: adult obesity, August 2021–August 2023 | Benchmark for measured BMI-defined obesity by age and sex. Calculator bands are mapped approximately to the published age groups. |
| Future joint model | 2024 American Community Survey Public Use Microdata Sample documentation | Intended source for a future weighted person-level model of variables available together, including age, income, sex, race/ethnicity, and marital status. |
Calculation steps
- Divide adults into seven age bands: 18–24, 25–34, 35–44, 45–54, 55–64, 65–74, and 75–90. The selected age range determines how much of each band is included.
- Weight the included bands by their modeled share of the male or female adult population. A partially included band is prorated uniformly by year of age.
- Model height as a normal distribution for each sex and personal income as a log-normal distribution for each sex and age band. Calculate the share inside the selected minimum and maximum.
- If selected, apply the age-band marital-status and obesity pass rates. Apply the selected race/ethnicity share as a separate national multiplier.
- Sum the weighted age-band results, multiply by the estimated population in the selected age range, and round the headline population count to a whole person.
Worked example
With version 1.0 defaults—U.S. women ages 25–45, height 5′4″–6′0″, modeled personal income $50,000–$200,000, any ethnicity, and no marital-status or obesity exclusion—the implementation returns about 17.0%. The value is obtained by weighting the income pass rate across the included age bands, multiplying it by the modeled height pass rate, and summing the bands. Changing the reference population to men with the same numeric filters produces a different result because the fitted height and income distributions differ.
This example documents the current code's behavior; it is not a claim that exactly 17.0% of real-world women meet those criteria.
Assumptions and likely error
What the model captures
Income, marital status, and obesity assumptions vary by age band and sex. This is more realistic than applying one national rate to every age.
What remains independent
Within each age band, height, income, marital status, and obesity are multiplied as if independent. Race/ethnicity is also applied independently. Real traits are correlated.
Definitions do not equal availability
“Unmarried” includes never married, divorced, separated, or widowed people and does not establish that someone is single, dating, interested, or geographically reachable.
The count is especially approximate
The headline count scales a national percentage. It does not account for local population, sexual orientation, mutual preference, or the size of an actual dating market.
Version history
Version 1.0
Age-stratified aggregate approximation. Added an explicit source register, worked example, assumptions, and a clear distinction between official benchmarks and fitted model values. A future version should recompute compatible variables from weighted ACS PUMS person records and publish a reproducible transformation file.
How to use the result responsibly
Use the percentage to understand how several demographic filters compound, not as a verdict on your standards or dating prospects. The model does not measure personality, values, communication, chemistry, location, or mutual interest. For how we handle sourcing and corrections, read the editorial policy.