Methodology
Credible estimates require transparent methods. We intersect criteria on microdata and weight with PWGTP — we do not multiply independent percentages.
1. What the calculator measures
It estimates how large a population slice is given your criteria — population rarity. It does not measure dating activity, interest, response rates, or compatibility.
2. Data sources
Primary engine: U.S. Census Bureau ACS 5-Year Public Use Microdata Sample (PUMS), processed locally to Parquet and queried with weighted intersections.
3. Census PUMS
PUMS provides anonymized person records with demographic and economic variables. We filter records that match all selected criteria simultaneously.
4. Population weighting
Each PUMS record represents a number of people in the underlying population via the person weight PWGTP. We estimate counts as SUM(PWGTP), never by counting rows alone.
5. Income methodology
Person income uses PINCP adjusted by ADJINC: PINCP × (ADJINC / 1,000,000). Thresholds are applied in survey-year dollars unless a documented CPI series is configured.
6. Geography
MVP supports United States and state (and PUMA when present). City/radius claims are not treated as exact PUMS geography.
7. Statistical uncertainty
When replicate weights PWGTP1–PWGTP80 are available, the API can compute ACS successive difference replication standard errors and confidence intervals.
8. Limitations
ACS does not identify who is actively dating. Height/weight are not in ACS PUMS. Estimates depend on sample size and geography. Do not treat results as exact counts of available partners.
9. Future datasets
Planned integrations include NHANES (measured anthropometrics), BLS, IPUMS, and peer-reviewed dating-research parameters — always labeled separately from Census population rarity.