A probability study of human connection
One world. Billions of possibilities.

How many people
could be your person?

Begin with everyone. Add what matters. See what changes, and what it might take to actually meet.

01 / 07
First, make the world smaller

Where could love realistically live?

Choose any country you would genuinely date in, travel to, or relocate for. Start broad; you can always change it.

Type a country, then use the arrow keys and Enter to select it.

or
select one or more countries
High confidence Country populations update from the World Bank when available.
Desire, without assumptions

Tell us about you.

Your identity helps suggest a starting partner selection and determines which partner-preference research applies. You can change every suggestion on the next page.

I describe my gender as
My orientation is closest to
Coverage varies Gender-identity and orientation data are not collected consistently worldwide.
Who you are open to

Who could your person be?

Select every gender identity you would genuinely consider. We account for the chance that their orientation includes you.

I’m open to dating…
Estimated reciprocal dating opennessThe share of people selected above whose reported orientation and dating preferences could include someone with your identity.
94.1%
Measured proxy
About 94 in 100 gender-aligned men report an orientation that could include a gender-aligned woman.

Based on the UK Office for National Statistics 2024 sexual-identity distribution.

What this number meansIt removes people whose stated orientation or dating preferences probably would not include someone with your identity. It does not estimate whether they will be attracted to you personally.
Life stage

What age feels right?

Age is estimated from UN population distributions. This is a population share, not a judgment about compatibility.

27–42
1885+

The age estimate uses a trusted country profile where one is available; other selections retain the worldwide age curve. UN World Population Prospects 2024 ↗

38%
38% is a broad, adjustable starting point, not an age- or country-specific result. Partnership status varies with age, sex, local definitions of marriage and cohabitation, law, economics, and social norms.
UN World Marriage Data · age and sex, 232 countries ↗OECD Family Database · marriage and cohabitation ↗
Belief & background

What should feel familiar?

Both are optional. Trusted country profiles may refine individual shares; unsupported factors retain the worldwide proxy. “Open to any” remains the most statistically honest default.

Religion or worldview
Open to any religion or worldview
Select any number. Selected shares are added together; trusted country profiles may refine them.
Cultural or regional background
Open to any background
Using the worldwide fallback. Country-specific background data are currently applied only when the United States is selected. Select any number. The U.S. crosswalk converts overlapping Census race, Hispanic-origin, and ancestry measures into the model’s mutually exclusive broad categories. It is an approximation, not an identity census.
Mixed confidence Religion uses a measured baseline with trusted country refinements where available. Background uses a U.S. Census-based crosswalk for the United States and a coarse worldwide proxy elsewhere.
Practical compatibility

What does a shared life need?

Keep what matters. Leave the rest open. Income is modeled against local economic data when available.

No minimum
Open$200k+
Children
Lifestyle
The non-smoker estimate uses a trusted country profile where one is available. Other selections use 2022 U.S. current-cigarette-smoking rates, with the U.S. overall rate as a limited nonbinary fallback. CDC smoking data ↗
Now, introduce chance

How often do you meet someone new?

A “new person” means someone you speak with long enough to learn whether you might want to see them again. These inputs remain editable and recalculate the results instantly.

6
65%
4%
This is the most personal assumption in the model. Try 2–8% to see a realistic range.
YOUR RESULT / READY TO SHARE
THE FINE PRINT / 01

Honest numbers,
soft edges.

LoveRace is a what-if model, not a forecast. The encounter calculation treats new people as independent random draws, although real social circles, dating apps, neighborhoods, workplaces, and repeated encounters are highly clustered. Actively seeking a partner also changes the process through where you go, whom you approach, which platforms you use, and how often you create new opportunities.

What is being modeled

The model begins with the world population or the populations of selected countries, then applies broad shares for partner gender, reciprocal dating openness, age, relationship availability, religion, background, income, and lifestyle preferences. Selected countries are calculated independently with trusted local refinements where available, then their eligible pools are added together. Unsupported fields retain the worldwide or external proxy.

The calculation

pool = Σ(country population × applicable factors)

chance per encounter = pool share × geographic relevance × connection assumption

weekly chance = 1 − (1 − chance per encounter)new people per week

P(by week n) = 1 − (1 − weekly chance)n

What the model simplifies

  • Filters are treated as independent. In reality, age, country, income, religion, background, orientation, relationship status, gender identity, and family preferences influence one another. Multiplying separate population shares can make the combined pool too large or too small.
  • Several inputs are broad defaults. The 38% relationship-availability value is not calculated from age or country. Background categories can overlap even though their shares are added, and the current categories total 98%. Gender population shares remain assumptions where a trusted country refinement is unavailable.
  • Orientation and dating openness are separate ideas. The user's orientation suggests an initial partner-gender selection, which the user can override. The selected partner genders drive the reciprocal dating-openness calculation. Estimates involving trans and nonbinary people rely on geographically limited studies and are marked as limited evidence.
  • Smoking coverage varies by country. Trusted country profiles refine supported partner groups. Other groups retain the 2022 CDC U.S. proxy, including the overall 11.6% rate as a limited nonbinary fallback.
  • U.S. background categories require a crosswalk. Census race, Hispanic-origin, and ancestry measures overlap and do not align exactly with LoveRace’s broad categories. The model converts them into a mutually exclusive distribution so selected shares can be combined without double counting.
  • The range is not a confidence interval. It is a sensitivity cue that broadens as assumptions are added. Meeting a filter also does not establish chemistry, timing, emotional availability, safety, reciprocal effort, or compatibility.

Data sources

The sources below use different years, definitions, sampling methods, and geographic coverage. Together they form an informed approximation, not a unified census.