Where your house stands
The typical ghost
half died before, half after
The first people here
The column of the dead
Raised all at once, they rise as a column of light. The oldest have drifted highest and faintest; the most recent stay close to the ground.
Under whose rule they died
Everyone who ruled this spot, and how many of its dead died in their time.
Notable dead nearby
The register
A few of them, drawn at random with the odds of their time. The details are typical of their age and place, not records of real people.
Elsewhere in the census
Tap a place to fly there and raise its dead.
Your census return
How this census was taken
What is counted. Everyone estimated to have died within 100 metres of the house (about 31,400 m²), from the first humans to Halloween 1926: the century-old dead. People are assumed to have died roughly where they lived. These are modelled averages, not records of individuals.
Where you are. Your country is read from the map inside your browser. Whether you are in a historic city, and since when, comes from the data below. Elsewhere, the kind of place is guessed from the buildings OpenStreetMap shows within 100 m and left for you to correct.
Who lived here. The countryside comes from the HYDE 3.2.1 history of population, a world grid of about 9 km cells from 10,000 BCE onward (Klein Goldewijk et al., 2017). Before that, the Ice Age is spread over the land people had reached by then: Europe from about 43,000 BCE, Australia from about 63,000 BCE, the Americas from about 16,000 BCE. Cities come from the geocoded populations of some 1,600 historic cities (Reba, Reitsma & Seto, 2016, after Chandler and Modelski), with corrections where modern estimates differ, such as Tenochtitlan at about 212,500 (Smith, 2005). Each city is drawn as a disc of people around its centre, densest in the middle. Your home counts as inside the city from the year the disc reaches it.
How many died. Deaths are people × death rate × time. Death rates come from life tables, not from guessed birth rates: about 4% of people a year in antiquity and 3.4% around 1800, with a higher rate in crowded premodern cities. By 1900–1926 the rate was about 1.6% a year in richer countries and 3.6% in poorer ones (Riley, 2005; Gurven & Kaplan, 2007; Woods, 2003). Summed over the world, that comes to about 56 billion deaths before Halloween 1926 (plausibly 45–75 billion). The often-quoted 117 billion "ever born" also counts everyone born since, and assumes prehistoric birth rates higher than any population has been seen to sustain.
Events. About 300 battles, sieges, epidemics and disasters before 1926 that killed many people in one place are added where their footprint reaches your home, using published low, central and high death tolls. Epidemics add only their excess over ordinary deaths. Wars and epidemics before 1926 are all inside the count. Massacres are not named on the map.
Who ruled here. From the historical-basemaps world maps by André Ourednik (GPL-3.0), drawn at world scale, so borders and names are approximate.
The range. Each answer is run 1,500 times, varying death rates, population, city size and density, and event tolls. The range shown is the middle 90% of those runs. It is widest for antiquity.
What it can't know. A churchyard, a plague pit or a battlefield trench can hold far more than any average; a quarter abandoned for centuries holds less. These are modelled averages, not records of individuals.
Privacy. Your pin never leaves your browser. Map tiles come from OpenFreeMap, the census data from this site, and searches go to Photon (komoot). The share image leaves out your streets unless you ask for them.
Population: HYDE 3.2 (CC BY 3.0) · cities: Reba et al. 2016 (CC BY 4.0) · rulers: historical-basemaps (GPL-3.0) · death rates: Riley 2005, Gurven & Kaplan 2007, Woods 2003, Scheidel 2001; PRB (Kaneda & Haub 2022) for comparison · events: published tolls, sources listed in the data · map data © OpenStreetMap contributors, tiles by OpenFreeMap and OpenMapTiles · a companion to The Veil of History.