Methodology

The score is the product of independent probabilities — a naive-Bayes style “configuration match” estimate. It is honest about being simple: each axis is a real statistic; multiplying them treats your traits as unrelated, which they are not. Treat the result as entertainment with footnotes, not a peer review.

1. Birth second

P(a random human shares your exact birth second) = P(born in your year) × 1/(seconds in that year). Births in your year = interpolated world population × crude birth rate (anchors from UN World Population Prospects approximations). Humans ever born: ~117 billion (PRB-style estimate).

2. Birthplace

City metro population (or country population when no listed city matches) divided by current world population (~8.23 billion). Cities are a curated list of major metros; unlisted towns fall back to the national figure.

3. Body lottery

P(height ± 0.5 cm | sex) × P(eye color | region prior) × P(handedness). Height: normal distributions (male 171±7 cm, female 159±7 cm world means). Eye color: global vs European survey priors. Handedness: ~10% left-handed consensus.

4. First name

Approximate bearers per million, from US SSA national data (public domain) plus hand-set estimates for South Asia, the Muslim world and China. Unknown names default to 40 per million — deliberately conservative.

5. Galactic odometer (fact row)

Whole days lived × (Earth orbital 29.78 km/s + Solar System galactic orbit ~230 km/s + solar apex ~20 km/s). Speeds are summed as path magnitudes rather than vector-added — the true displacement is somewhat smaller and changes direction.

6. Stardust (fact row)

Body element mass fractions (O, C, H, N, Ca…) converted to atom counts via atomic mass and Avogadro's number. Everything except hydrogen (~10% of atoms) was forged in stars; hydrogen predates them.

7. Birth sky (fact row)

Your wall-clock birth time is converted to UTC using longitude/15 as an offset — a simplification that ignores historical timezone politics (good to roughly an hour). Stars from the Yale Bright Star Catalog (V ≤ 5.0) are transformed to altitude/azimuth for your location and moment; the brightest star within 30° of zenith is highlighted on your card.

8. Gift modes (couple / baby / pet)

Couple: both partners are scored independently, then multiplied by a “paths crossing” term — roughly 150 meaningful ties (Dunbar's number) against ~8.23 billion humans alive. Because we only ask for a partner's name and date, their body lottery runs on default priors (the lite set). Formula: R = R₁ × R₂ × max(150 / worldPopulation, 10⁻⁶).

Baby: identical engine minus the body-lottery dimension (newborns have no meaningful adult height/handedness priors). Eye color defaults apply when unknown.

Pet: same lite set as Baby, with the gotcha-date standing in for the birth second. Pet-population baselines are rough global estimates and clearly entertainment-grade; pet names use the conservative default frequency.

Known weaknesses

• Independence assumptions overstate rarity (your height and eye color don't care about your name).
• Name frequencies are region-blended estimates, not census-exact.
• Couple/baby/pet modes inherit every weakness above plus their own disclosed approximations.
• “Configuration of matter” is poetry wearing a lab coat.

Data sources

• UN World Population Prospects (approximate anchors)
• Population Reference Bureau — humans-ever-born estimate
• US SSA baby names (public domain)
• Yale Bright Star Catalog, 5th Rev. Ed. (Hoffleit & Warren 1991)
• NASA fact sheets — Earth orbital speed; galactic kinematics literature

Everything above runs client-side in one page of TypeScript. Nothing is uploaded, nothing is stored.