Everything people ask after seeing their number — including the uncomfortable ones.
The big number is your dating pool: how many local adults match your standards. The second number is your odds: how many of them would swipe right on you back. The first is about your pickiness; the second is about your competitiveness. Both matter.
No — it's a directional estimate, not a prophecy. It can't see your charm, your photos, or that thing you do with your eyebrows. What it can do is show how brutally the math compounds: each filter multiplies, and attraction has to go both ways.
Population and demographics come from the Census Bureau's American Community Survey (ACS) 2024 across 75 U.S. metros. Body-size bands use CDC data; religion uses Pew 2023–24; drinking and smoking come from CDC BRFSS SMART 2023 per metro and by sex, and drug use leans on SAMHSA/NSDUH; relationship intent uses Pew (dating-intent filter) and Singles in America 2026 (long-term-intent seed). Swipe and dating-funnel behavior draws on published studies (Tyson et al., Bruch & Newman, Hitsch et al., Celdir et al.).
Filter by filter:
Full source list also lives at the bottom of the calculator page.
Because filters multiply. Ten filters that each keep 70% of people leave you with about 2.8% — that's not a bug, it's arithmetic. The people with the smallest numbers aren't unlucky; they're specific. Loosen one filter and watch what happens.
If you lined up local adults at random, about one in N would clear all your filters. Bigger N = pickier you.
The model computed effectively zero people under your current assumptions. It doesn't mean zero humans exist — it means your combination of filters and funnel assumptions rounds to nobody. Loosen something.
Your profile (age, height, income, education, and more) is scored against the same kinds of filters others set, producing a profile-fit multiplier. That's combined with three adjustable assumptions: what share of people would swipe right on you, how active they are on dating apps, and what share of matches turn into first dates. You control all three — the defaults are deliberately middle-of-the-road.
The big number is your pool filtered by your standards, then run through a 90-day funnel:
The headline number is a midpoint, not a promise: the modeled 90-day range runs from roughly 0.45× to 1.8× that midpoint, and the midpoint itself comes from your expected positive-response rate across the estimated profile views in 90 days.
The quiet ranges under the big numbers come from each filter's data uncertainty, compounded: Census-measured filters count about ±4%, national survey filters ±8%, measured conditional tables ±10%, modeled or approximated joints ±15%. That uncertainty flows through every funnel stage into your odds. The assumption sliders are scenarios you control, so they sit outside the range. These percentages are rough judgment calls about data quality, not statistical confidence intervals.
The income effect is anchored to Hitsch, Hortaçsu & Ariely's study of ~22,000 online daters: men above $50k received 34%–151% more first contacts as income rose, while women's outcomes were mostly flat. The model conservatively rescales that to your metro earnings percentile. Swipe selectivity comes from Tyson et al. (2016): 59% of women liked 10% or fewer profiles, versus 9% of men.
Where we have real joint data, no: demographics (age × marital × education × race × income × kids) come from a measured PUMS joint table per metro and sex, and religion → drinking and wants-kids → childlessness use measured conditional tables. Marijuana × other drugs combines SAMHSA's past-year marijuana rate (23.4%) with its past-month other-illicit-drug rate (3.38%): because the two span different reference periods, the four combinations use an independence approximation over the mixed periods and are labeled approximations. For politics and smoking we multiply independently and say so on the page — it’s the honest approximation, and the methodology section flags exactly where it’s used.
75 U.S. metros, from New York to Boise. Pick yours at the top of the calculator — every number recomputes live.
There's no account and nothing identifies you. When you share, we store your filter settings and result numbers under a random link ID so the preview card works and friends can open your exact setup. That's it — no name, no email unless you voluntarily sign up for updates.
Yes — open a friend's share link and use Friend Compare to run their filters against your city, or yours against theirs. Nothing says friendship like quantifying each other's standards.
An optional, affectionately brutal reading of your standards. It roasts the filters, not you. Mostly.
Hit Share anywhere on the page. You'll get a short link plus a card image with both your numbers — built for iMessage, X, and everywhere else people argue about dating.
The City dating market section ranks all 75 metros by your expected mutual matches, adjusted twice. First, the market tilt: how many rivals are chasing the same people where you live. Following the Date-onomics argument (Jon Birger, building on Guttentag & Secord), the ratio that sets a dating market’s terms is the college-educated one — so bachelor’s+ profiles use college-educated men per women 25+ (ACS 2024 B15002), everyone else uses the overall 18–44 ratio (B01001).
Second, the short-vs-long lean: the scarcer sex sets the market’s terms. Where men are scarce the culture skews casual; where women are scarce it skews toward commitment. Each city’s long-term share of your matches shifts with that lean, and the card also shows your short-term mutual number for comparison. Both adjustments are capped so they nudge rather than decide — theory-informed proxies, not predictions. Spark is the share of people you’d like who’d swipe right back, before app-activity mechanics shrink it.