How we count people you have not met

01 The short version

We start with everyone in the country you choose, keep the people of the gender you are looking for, cut to your age range, then apply each requirement you add. At the end we remove people who are probably already married or in a relationship. What is left is the number you see.

The same maths runs for around 150 countries, so we can also add the world up and rank countries against each other.

02 What goes in

Population by age and sex, marriage and partnership shares by age group, height distributions, income and net worth distributions, body-mass classes, employment status, education levels, lifestyle shares (smoking, drinking, pets, car and home ownership and so on), religion and sexual orientation.

Typical sources: UN World Population Prospects, World Bank, WHO and NCD-RisC, the UBS Global Wealth Report, OECD, Eurostat, ILO, Pew Research, Gallup and national censuses. We use the latest figures we could find and round them like good adults.

03 How the numbers are combined

Height is modelled as a normal distribution per country and sex. Income and net worth follow a log-normal shape fitted to medians and inequality figures, and both rise with age. Body shape comes from BMI classes. Hair and eye colour come from regional population tables, and hair greys with age.

Naively multiplying probabilities is wrong when traits are linked (older people earn more, own more and are more often married), so we apply age-aware effects and a correlation damping factor. That keeps long lists from collapsing to zero faster than reality would.

We also show an uncertainty band. Treat the headline figure as the middle of a range, not a promise.

04 Orientation and attraction

We only count people who could plausibly be attracted to someone of your gender. That uses survey-based shares of straight, gay, lesbian and bisexual adults per country. In places where people may under-report, real figures are likely higher.

05 Where it gets wobbly

Data quality varies. For countries with sparse statistics we borrow from similar neighbours and flag lower confidence. Surveys reflect the past, people change, and real life is full of correlations no table captures.

The model works with two sexes because most official statistics do. It leaves out important things such as chemistry, kindness, humour and timing. It cannot tell you who will like you back.

06 What we deliberately leave out

No ethnicity or race filter, ever. We also never single out individuals: all figures are aggregate.