For every building in the cores of six towns we measured how many minutes' walk on the footpath network it is to the nearest grocery, public-transport stop, kindergarten, primary school, park or playground, restaurant or café, post office or bank, and pharmacy. The minutes were then folded into one score.
That is 88,947 buildings and 216,651 dwellings. The map below shows every building but names none: the score speaks of a place, not of the people who live in it.
What the score measures
One nearest destination of each type, measured in minutes on foot. The route is on the footpath network, at a speed that counts the climb: a hill takes longer than the flat. Every minute takes a little away: full credit at the door, none at 20 minutes, and a smooth curve between. It is the "twenty-minute city", without a plateau at the start: a score that stayed full for five minutes painted even Ljubljana a single colour.
The types are weighted: groceries 20, public transport 15, pharmacy 15, and kindergarten, school, park, food and drink and everyday services 10 each. The weights are a choice, so you can change them yourself under the map.
Six cores, one rule
A comparison is only as fair as its frame. So no town was outlined by hand: the core is the contiguous 500 m cells of at least 1,000 residents per km² around the central square. The settlement boundary does not work: the settlement of Nova Gorica stops short of Solkan, which is the same town on foot, while Ljubljana's runs into the fields.
That makes Ljubljana's core 49.3 km² with 245,396 residents, Maribor's 21.3 km² (90,263), Kranj's 11.0 km², Celje's 7.5 km², Nova Gorica's 4.8 km² and Koper's 4.5 km².
The map, town by town
Every building in the chosen town's core. The colour is the score or the minutes to the chosen type; buildings without dwellings are only outlined. The weights below change the score at once.
Loading …
The score is a weighted sum; the weights are scaled to a whole, however many you set. The weights are a choice, not a measurement. There is no doctor: OSM holds 44 in Ljubljana, a floor rather than a census, so the health weight goes to pharmacies, which come from the register.
What is near and what is not
Almost every dwelling in all six cores has a stop within ten minutes, and a grocery too. Scarcer are the school and the pharmacy. Only 69% of dwellings in Koper's core have a school within ten minutes, against 87% in Nova Gorica's. A pharmacy: 72% in Kranj.
A park or playground is within ten minutes for 72% of dwellings in Kranj's core and nine in ten or more elsewhere. Food and drink is scarcest in Celje (76%). A caution here: restaurants and parks come from OSM, which volunteers draw, so part of a difference can be who mapped what.
Share of dwellings (not buildings) in the urban core with a destination of each type within 5, 10 or 15 minutes' walk. Public transport is a stop, not a line or a frequency; outside Ljubljana the stops are OSM's alone.
Source: OSM, GURS, SURS, registers · bloki.si
Towns are alike, streets are not
The six towns' medians differ by only 7 points. Inside each town, the range between the worst tenth of dwellings and the best tenth runs from 26 points (Nova Gorica) to 47 points (Kranj).
Someone choosing where to live is choosing a street more than a town. Kranj stands out: its median is 85, but 13% of its dwellings score under 50, against between 3% and 5% in the other cores.
Score 0–100 per dwelling, at the study's weights. The urban core is the contiguous area of at least 1,000 residents per km² around the central square.
Source: OSM, GURS, SURS, registers · bloki.si
From the centre outwards
Within the first five hundred metres of the central square the score is high everywhere: from 84 in Kranj to 95 in Ljubljana. How fast it falls differs a great deal from town to town. Ljubljana stays high far out, while in Kranj the score falls to 38 at the edge of the core.
Median and interquartile range of the score per dwelling by straight-line distance from the central square, in 500 m bands; bands with fewer than 200 dwellings are left out.
Source: OSM, GURS, SURS, registers · bloki.si
Per dwelling or per building
Every figure in this article is per dwelling, not per building, because a dwelling is where a family lives. The difference is not small: the median per dwelling is 88 in Ljubljana and 81 per building; in Kranj 85 against 69. Blocks of fifty flats stand where the shops are; houses at the edge are one building for one family.
The map draws buildings, because buildings are what we see; the statistics count dwellings.
What the score does not say
One nearest point, not a count. A street with one café and a street with forty score alike.
What is not mapped is far away. Groceries, pharmacies, schools and Ljubljana's public transport come from registers; restaurants, parks and stops outside Ljubljana come from OSM, where a missing point means nobody drew it, not necessarily that it is not there.
A stop is not a bus. There are no lines and no timetables, so the score cannot tell a stop with a bus every five minutes from one with two a day.
There is no doctor. OSM holds only a handful of surgeries in Ljubljana, a floor rather than a census, so the health weight goes to pharmacies.
Nova Gorica borders Gorizia, whose shops and schools are in no Slovenian register, so a building near the border reads farther from everything than it is.
§ How it is calculated
The network. The pedestrian network from OSM, where ways join only where they share a node (a path under a bridge does not join the road above it), with slope from the elevation model. It is not the car network: Koper's old town is walked. Speed is Tobler's: about five kilometres an hour on the flat, less uphill and down a steep slope. A building and a destination attach to the nearest point of the nearest path.
The nearest destination. For each type one Dijkstra search from all destinations at once on the reversed network, so every point of the network gets its time to the nearest destination. The network and the destinations reach three kilometres beyond the core, so a destination just past the edge counts.
The score. A type's credit is full at the door and zero at 20 minutes, on a cosine curve between; the weighted sum gives a score from 0 to 100. Moving the sliders rescales the weights to a whole.
The core. Population cells of 500 m with at least 1,000 residents per km², connected to the central square (diagonals included). Buildings are from the building register, dwellings from the building parts used as dwellings.
The map computes for itself. When you change the weights, the browser recomputes each building's score from the exported minutes. The minutes are exported to a fifth of a minute, so the browser's score differs from the exported one by under a point. The figures in the text are always the exported ones, at the study's weights.
Can the Bloki server answer questions like these?
For any address the server says what is nearby and how far, but in straight-line metres, not minutes on foot.
- Yes
What is around my address?
Shops, schools, kindergartens, pharmacies, banks, public transport and more, with the nearest distance and the count within a radius, from OSM and from the registers separately, never added together.
neighbourhood - Yes
Which school is my child assigned to?
The catchment the building falls in, not the nearest school. The two often differ.
neighbourhood - Partly
How many minutes' walk is the shop?
The straight-line distance and a measured factor for how much longer the route is: on the flat, and more on a slope. Not walking minutes.
neighbourhood - No
What is my address's walking score?
The server does not serve walking scores or isochrones; they are only in this article, for six town cores.
data_limitations