Methods and validation
What our life simulation is built from, how its numbers are set, what the data says when we measure it, what to watch out for, and how to give credit.
What the world rests on
Every birth, death, school year, wage and family in this simulation is drawn from public statistics: UN World Population Prospects 2024 for births, population and life tables; WHO Global Health Estimates for causes of death; World Bank WDI for income, cities, schooling and employment; UNESCO UIS and UNDP HDR for education; ILO for work; UNICEF, UNAIDS and UNODC for child marriage, HIV and homicide; and the World Happiness Report for life satisfaction. The sources block on the game page names them all.
Each simulated life is one person’s whole story, told one year at a time across 172 countries and territories. The birthplace is drawn from each country’s real share of world births, together with a sex, a town or a village, and the family’s place in the country’s income spread. Whatever can happen in a year happens with the chances the data records for that place, that age and that situation.
How choices work
On the site, the reader steers a life: a focus for each year, and a choice whenever a decision comes up. A choice moves the odds of what follows; it never decides the outcome. Where lives must run unattended — the classroom lab and the validation below — each decision is drawn uniformly among the choices that are open at that moment, and each year keeps the game’s default focus. The policy is stated in the model line that every classroom run and every table here carries.
How the numbers are set
Each country’s life table is built so that life expectancy at birth matches the UN figure for that country, separately for women and men. Mortality from causes follows the WHO Global Health Estimates’ regional and age patterns. A life’s own circumstances — its income group, where it lives, what it goes through — move its risk around that national level.
The chance of finishing each stage of school comes from UNESCO and UNDP figures, read through the family’s income group and its town or village. Incomes are drawn from the country’s income distribution, shaped by its inequality. Marriage, births and moves follow the country’s own rates.
Measured against the data
Measured on model engine-2026-ff037793de331faf, 2026-11-06: 240 seeded lives per country below, choices drawn as described above. These numbers are reproducible — the same model and the same seeds give the same table.
| simulated | UN WPP 2024 | ± | |
|---|---|---|---|
| Nigeria | 56.3 | 54.5 | +1.8 |
| Ethiopia | 67.3 | 67.3 | 0.0 |
| India | 71.5 | 72.0 | -0.5 |
| Egypt | 71.3 | 71.6 | -0.3 |
| Brazil | 75.8 | 75.8 | 0.0 |
| Mexico | 73.4 | 75.1 | -1.7 |
| China | 75.5 | 77.9 | -2.4 |
| United States | 76.6 | 79.3 | -2.7 |
| Germany | 78.3 | 81.4 | -3.1 |
| South Korea | 80.4 | 84.3 | -3.9 |
| Japan | 80.0 | 84.7 | -4.7 |
| Lesotho | 57.8 | 57.4 | +0.4 |
| simulated | UN WPP 2024 | ± | |
|---|---|---|---|
| Nigeria | 3.30 | 4.48 | -1.18 |
| Ethiopia | 3.08 | 3.99 | -0.91 |
| India | 1.57 | 1.98 | -0.41 |
| Egypt | 2.42 | 2.75 | -0.33 |
| Brazil | 1.61 | 1.62 | -0.01 |
| Mexico | 1.64 | 1.91 | -0.27 |
| China | 0.77 | 1.00 | -0.23 |
| United States | 1.53 | 1.62 | -0.10 |
| Germany | 0.83 | 1.44 | -0.61 |
| South Korea | 0.86 | 0.72 | +0.14 |
| Japan | 0.83 | 1.21 | -0.38 |
| Lesotho | 2.36 | 2.69 | -0.33 |
| Primary | Lower secondary | Upper secondary | University | |
|---|---|---|---|---|
| Nigeria | 87.9% | 82.1% | 58.3% | 17.5% |
| Ethiopia | 57.5% | 45.8% | 28.8% | 10.4% |
| India | 94.2% | 90.0% | 67.9% | 32.5% |
| Egypt | 97.9% | 97.1% | 74.6% | 39.2% |
| Brazil | 98.3% | 97.5% | 88.8% | 52.1% |
| Mexico | 98.8% | 97.9% | 91.7% | 55.0% |
| China | 97.9% | 96.3% | 87.5% | 46.3% |
| United States | 97.9% | 97.1% | 89.2% | 55.4% |
| Germany | 98.8% | 97.9% | 90.8% | 51.7% |
| South Korea | 98.3% | 97.9% | 90.4% | 62.5% |
| Japan | 98.3% | 97.1% | 88.8% | 51.3% |
| Lesotho | 50.4% | 42.9% | 31.7% | 11.3% |
The data column is the UN’s figure for the country; the simulated column is the mean age at death of the lives drawn. Countries at the top of the world’s distribution land short of the data — the simulation’s layer of events and circumstances pulls realised lifespans down where most lives would otherwise run long. We publish it as it is.
Stages reached by the lives drawn, in shares. The model reads school completion from UNESCO and UNDP inputs through each family’s circumstances, so these shares are the outcome of the fitted model, not a restatement of the inputs.
What to keep in mind
- A result about these simulated lives is a result about our model. Check it against real-world numbers before drawing conclusions about the world.
- Most rates are national averages, divided by sex, age, income group and town or village only where the data goes that far. Life inside one country is exactly as detailed as the data behind the model.
- How one event shapes the next is modelled simply, taken from published studies where they exist, and estimated where they do not.
- The model changes as this site grows. For work that must be run again, quote the exact model version from the table above — the classroom lab reports the same version with every run.
Citing this work
Cite the site, the page and the exact model version, for example: “1in8B (in8b.com), life simulation, model engine-2026-ff037793de331faf, methods and validation page, accessed 2026-11-06.” The classroom lab and its exported CSV carry the same model version, so a reader can check which engine produced a run.
The Developer API documents the shared notes feed and share links for researchers who want to read what players chose to publish; nothing private is exposed there.
In the classroom
The teachers’ guide on the edu page sets out lesson plans, and the classroom lab runs batches of lives with the same engine shown here — with lesson rules that ride in the student link, and anonymous ten-letter class codes that let a class compare its lives with the data without giving up a name.