The formula. For every country, every sex and every five-year age band we take the real population count, then multiply it by the share of that cell that passes each filter: the fraction of the age band inside your range (under-18s are always removed), the share whose relationship status, parenthood and lifetime-partner count you accepted, the normal-curve probability of a height inside your range, the normal-curve probability of a BMI under your cut-off, the share at or above your build level, and the log-normal probability of earning at least your income. Then we add all 218 countries × 2 sexes × 17 age bands back up.

The big assumption: the filters are treated as independent inside a country-sex-age cell. In the real world they are correlated (taller people earn a little more, richer countries are older, and so on), so treat the result as an order of magnitude, not a census. Each tab's ×0.49 chip is that filter's own cut, applied in tab order.

The squares. Area is proportional to people: the cream frame is all 8.06B on Earth, the espresso square is who fits your filters, the butter square is your type. When a square gets too small to see we magnify that corner in a lens and say by how much — the squares themselves never change scale.

Sources

Built 2026-09-15 · prototype.