So I have been playing around with some citystrides data. I just posted it on my running-related blog and figured the people in this forum might care too. Not to be seen as advertising, I am just copy-pasting the contents. If you really want to read it on the blog, you can find that link on my profile.
Note: it seems copy pasting it from my text editor into this form has messed up the unicode characters. Apologies to the non-latin script cities.
I am creeping closer to finishing running all streets in Brussels. It got me curious. How many big cities have been completed or are near completion? I track my progress through Citystrides, as detailed before on this blog, so I decided to use that as a basis to answering my question. Will that skew the data? Sure. Do I care? No.
Warning: a lot of these will end up being tables of data, because that is what I like looking at. Don’t expect any groundbreaking conclusions to be drawn from the data.
There is no easy API to get this information out of Citystrides. Instead I
decided to scrape all the cities and go from there. The scraping of cities was done from 28 to 30 September 2022 from the main city index. Armed with a list of 170,000 cities, I then selected those with 5000 or more streets. For those I scraped the first page of striders running in that city, so that is a maximum of 12 top striders per city. I always made sure to play nice and not make more than six requests a minute.
There is no clear answer. Different countries define the boundaries in
different ways. Citystrides uses OpenStreetMap information, so we need to rely on those being present as well. Which boundaries Citystrides selects can also be a topic of discussion. Should you think about that more in a serious study? Yes. Is this a serious study? No.
Citystrides works with the “ways” present in OpenStreetMap. To try to make sense of them, Citystrides joins ways with the same name within one city into a street. So streets without a name do not count. Streets where only cars are allowed do not count (think highways). Forest paths and the like also tend to not count, though this is heavily dependent on how they happen to be tagged in OpenStreetMap. See my serious study remark from the previous paragraph.
For those curious about the entire city dataset, here are some descriptive statistics. These are calculated after some data cleanup:
- Cities with 0 streets were removed (some data is not fetched until someone runs in that city on Citystrides).
- Some deduplication was attempted. If two cities had the same name and the same street count, one of them was thrown out. This works reasonably well for cities with enough streets and likely fails quite often for cities with fewer than, say, 5 streets.
You might think the caveat of the second bullet surely cannot matter much, but it turns out there are a lot of cities with almost no streets. For the about 170k cities, these are the numbers.
So these are the cities currently in Citystrides with 5000 or more streets. Brussels just makes the cut with 5007 streets. Happy little accidents.
|0||36594||S√£o Paulo||S√£o Paulo, Brasil||44082|
|1||188749||Greater London||England, United Kingdom||39149|
|2||227615||Sydney||New South Wales, Australia||32375|
|4||225681||ŸÖÿ≠ÿßŸÅÿ∏ÿ© ÿßŸÑÿ¨Ÿäÿ≤ÿ©||ŸÖÿ≠ÿßŸÅÿ∏ÿ© ÿßŸÑÿ¨Ÿäÿ≤ÿ©, ŸÖÿµÿ±||22319|
|5||177521||City of Cape Town||Western Cape, South Africa||22308|
|6||225679||ŸÖÿ≠ÿßŸÅÿ∏ÿ© ÿßŸÑŸÇÿßŸáÿ±ÿ©||ŸÖÿ≠ÿßŸÅÿ∏ÿ© ÿßŸÑŸÇÿßŸáÿ±ÿ©, ŸÖÿµÿ±||20787|
|7||225680||ÿßŸÑŸÇÿßŸáÿ±ÿ©||ŸÖÿ≠ÿßŸÅÿ∏ÿ© ÿßŸÑŸÇÿßŸáÿ±ÿ©, ŸÖÿµÿ±||20734|
|8||177494||eThekwini Metropolitan Municipality||KwaZulu-Natal, South Africa||19706|
|9||33385||Rio de Janeiro||Rio de Janeiro, Brasil||17953|
|10||227695||ŸÖÿ≠ÿßŸÅÿ∏ÿ© ÿßŸÑÿ±Ÿäÿßÿ∂||ŸÖŸÜÿ∑ŸÇÿ© ÿßŸÑÿ±Ÿäÿßÿ∂, ÿßŸÑÿ≥ÿπŸàÿØŸäÿ©||17939|
|12||177491||City of Johannesburg Metropolitan Municipality||Gauteng, South Africa||16154|
|14||7696||Jacksonville||Florida, United States||14816|
|15||177488||City of Ekurhuleni Metropolitan Municipality||Gauteng, South Africa||13949|
|17||10538||San Antonio||Texas, United States||13509|
|18||221129||Surabaya||Jawa Timur, Indonesia||13235|
|19||172050||Auckland||Auckland, New Zealand / Aotearoa||13232|
|20||171368||Houston||Texas, United States||13182|
|21||171512||Census Divsion No. 6||Alberta, Canada||13153|
|23||177489||City of Tshwane Metropolitan Municipality||Gauteng, South Africa||11848|
|24||130998||Leeds||England, United Kingdom||11657|
|25||187894||Indianapolis||Indiana, United States||11622|
|26||35294||Belo Horizonte||Minas Gerais, Brasil||11276|
|27||171334||Charlotte||North Carolina, United States||11274|
|28||221380||Semarang||Jawa Tengah, Indonesia||11217|
|29||2525||Los Angeles||California, United States||11098|
|30||218739||Johor Bahru||Johor, Malaysia||10916|
|31||12129||Columbus||Ohio, United States||10750|
|34||2550||San Diego||California, United States||10112|
|39||5637||Las Vegas||Nevada, United States||9665|
|40||173609||Fort Worth||Texas, United States||9628|
|44||5074||Austin||Texas, United States||9236|
|45||2098||San Jose||California, United States||9205|
|46||14065||Dallas||Texas, United States||9061|
|47||15734||Memphis||Tennessee, United States||8886|
|48||104756||Madrid||Comunidad de Madrid, Espa√±a||8817|
|49||221066||Jakarta Timur||Daerah Khusus Ibukota Jakarta, Indonesia||8710|
|50||160222||Monterrey||Nuevo Le√≥n, M√©xico||8632|
|51||225689||ŸÖÿ≠ÿßŸÅÿ∏ÿ© ÿßŸÑÿ•ÿ≥ŸÉŸÜÿØÿ±Ÿäÿ©||ÿßŸÑÿ®ÿ≠Ÿäÿ±ÿ©, ŸÖÿµÿ±||8492|
|53||36434||Campinas||S√£o Paulo, Brasil||8342|
|54||22080||Louisville||Kentucky, United States||8295|
|55||159441||San Luis Potos√≠||San Luis Potos√≠, M√©xico||8218|
|58||160068||Morelia||Michoac√°n de Ocampo, M√©xico||7991|
|59||131066||Birmingham||England, United Kingdom||7990|
|61||18074||Nashville-Davidson||Tennessee, United States||7766|
|64||171347||Virginia Beach||Virginia, United States||7522|
|65||130996||Bradford||England, United Kingdom||7519|
|66||9842||Raleigh||North Carolina, United States||7427|
|67||9252||Albuquerque||New Mexico, United States||7325|
|69||31703||Campo Grande||Mato Grosso do Sul, Brasil||7129|
|70||227966||–ì—Ä–∞–¥ –ë–µ–æ–≥—Ä–∞–¥||–°—Ä–±–∏—ò–∞, –°—Ä–±–∏—ò–∞||7126|
|71||227612||Honolulu County||Hawaii, United States||7080|
|72||97060||Nelson Mandela Bay Metropolitan Municipality||Eastern Cape, South Africa||7031|
|75||187823||Wiltshire||England, United Kingdom||6846|
|77||132430||Ecatepec de Morelos||Estado de M√©xico, M√©xico||6762|
|78||221065||Jakarta Selatan||Daerah Khusus Ibukota Jakarta, Indonesia||6708|
|80||5732||El Paso||New Mexico, United States||6541|
|83||130967||Sheffield||England, United Kingdom||6342|
|84||131026||Manchester||England, United Kingdom||6320|
|85||187843||Durham||England, United Kingdom||6263|
|86||10126||Oklahoma City||Oklahoma, United States||6224|
|87||171363||Glasgow||Scotland, United Kingdom||6161|
|88||34030||Porto Alegre||Rio Grande do Sul, Brasil||6117|
|90||132529||Iztapalapa||Ciudad de M√©xico, M√©xico||6012|
|91||36891||Sorocaba||S√£o Paulo, Brasil||5965|
|92||25864||City of Casey||Victoria, Australia||5960|
|93||46528||Bogot√°||Distrito Capital, Colombia||5957|
|94||7563||Colorado Springs||Colorado, United States||5944|
|97||131085||Liverpool||England, United Kingdom||5695|
|99||130985||Kirklees||England, United Kingdom||5667|
|100||221069||Jakarta Barat||Daerah Khusus Ibukota Jakarta, Indonesia||5666|
|101||187842||Dorset||England, United Kingdom||5613|
|102||187803||Cheshire East||England, United Kingdom||5605|
|103||36151||Guarulhos||S√£o Paulo, Brasil||5581|
|104||2421||Bakersfield||California, United States||5575|
|105||124652||Santiago de Quer√©taro||Quer√©taro, M√©xico||5538|
|106||25809||City of Greater Geelong||Victoria, Australia||5520|
|107||225540||Kuala Lumpur||Kuala Lumpur, Malaysia||5490|
|108||6836||Tucson||Arizona, United States||5459|
|109||218670||Hulu Langat||Selangor, Malaysia||5422|
|110||32830||Natal||Rio Grande do Norte, Brasil||5362|
|112||188751||Torre√≥n||Coahuila de Zaragoza, M√©xico||5318|
|113||213880||Saltillo||Coahuila de Zaragoza, M√©xico||5312|
|114||159171||Acapulco de Ju√°rez||Guerrero, M√©xico||5255|
|116||187804||Cheshire West||England, United Kingdom||5184|
|117||189835||Warszawa||wojew√≥dztwo mazowieckie, Polska||5184|
|118||113291||G√∂teborgs Stad||V√§stra G√∂talands l√§n, Sverige||5162|
|119||95259||Durango Ciudad||Durango, M√©xico||5134|
|120||5630||Henderson||Nevada, United States||5129|
|121||229537||Municipiul Bucure»ôti||Municipiul Bucure»ôti, Rom√¢nia||5072|
|122||213874||Municipiul Bucure»ôti||Bucure»ôti - Ilfov, Rom√¢nia||5070|
|123||25808||City of Wyndham||Victoria, Australia||5048|
|124||32136||Jo√£o Pessoa||Para√≠ba, Brasil||5023|
|125||207687||Brussel-Hoofdstad - Bruxelles-Capitale||R√©gion de Bruxelles-Capitale - Brussels Hoofds…||5007|
The astute reader might spot some oddities. Calling the Loire (spot 11) a city is a stretch. Indeed, I think this is a Citystrides issue that I will report separately. Similarly, the top ones are various gigantic swats of land in and around Cairo, Egypt, though I do not see an alternative for those areas. Spots 6 and 7 seem to be the same city, though with slightly different boundaries. One of them will have to be removed. Same for spots 121 and 122. Those, and many other duplicates I bumped into, have been reported in this CS forum post.
Let’s look at how many of these are completed or near completion. Also considering some variations on this idea.
For this list, I sorted everyone by percent in their respective cities. Here is the top 20. The ones listed as “Anonymous” is due to their privacy settings in Citystrides. I added the total number of streets for each city as well for perspective.
|Strider||City pos||Pct||City||City streets|
|4||Tony Woods üå±||1||99.77||Liverpool||5695|
|7||Dan Schoettinger||1||96.28||Colorado Springs||5944|
So at the time of writing, only two big cities, as per my definition, are
completed by someone. My guess is both the 99.9% and 99.77% counts are cities that had been at 100%, but an update made those striders lose some streets. Perhaps the 99.11% one in Berlin as well. Then in spot six you have me, hoping to finish Brussels in the next few weeks. Toronto also has quite the strider interest, with three people showing up in this little top 20. Other doubles cities are Liverpool, Brussels, Ottawa, and Calgary.
Similar, but now sorted by number of streets that were completed by the strider in that city. Note that I cut some extras from the first name cause it was messing up my table layout.
|Strider||City Pos||Pct||Completed Streets||City|
|1||James Salmon||1||26.9||10531||Greater London|
|10||Sean P||2||17.41||6816||Greater London|
|17||Bravo Foxtrot||3||14.82||5802||Greater London|
|18||Dan Schoettinger||1||96.28||5723||Colorado Springs|
|19||Tony Woods üå±||1||99.77||5682||Liverpool|
That changes things up a little bit. The cheer size of some of the cities makes it so you need to complete a lot of streets and still get seemingly nowhere. Our new first ranked person has only about a quarter of the streets in Greater London, but Greater London is the second largest city on Citystrides with just shy of 40,000 streets, so, ye, they did a massive amount of striding just to get there.
On a personal note, this drops me down to 27th place. Not unexpected, Brussels is the smallest of these big cities after all.
I take the percentages completed of the first page of striders and average them out. I also ignore all cities that do not have at least 12 striders. Surprisingly that is around half of our original list of big cities! Taking the mean can feel like an arbitrary measurement, but my reasoning is as follows. If you open a city and see a crowded first page in terms of completions, then it just feels like there is some good competition and liveliness going on. It is nice to know that you are not the only idiot going around running all the streets.
We keep seeing the same names pop up, but it is expected. If you have some people that have most of the city completed, then it will help a lot in bringing the mean up.
Bit of a weirder one perhaps to look at. I wanted something like taking the median from the first page, but I also did not really care about the percentage of the sixth and seventh placed person. So instead I made the executive decision to pick number three.
While we are at it, which city should you probably avoid if you just want to get first page on a big city? In other words, which city’s 12th place has the most number of streets completed?
|Strider||City||Streets for 12th spot|
|Marcus Schodorf||Greater London||3320|
That said, finishing a short street in the dense old city centre of a European city is probably easier to come by than a long grid-like American one.
It stands to reason that the global top striders would also live (or at least, run) in big cities. There are just more streets there to easily get to. However, the data does not seem to necessarily agree with this. Of the global top 25 (as looked at on 9 October 2022), 2 striders are hidden by privacy settings. Of the remaining 23, just 9 striders appear in the first page of any of the big cities. So somewhat evenly split.
- I have an arbitrary cut-off for what is a “big” city.
- I only consider the first page of striders for a city.
- There is no worldwide definition of where to draw city borders.
- City design and how to name streets is not consistent.
So the SQLite database is 11 MB. It is not massive, but I also do not want to host it myself. If you want the data, drop me a message. You can get the Jupyter Notebook with it too.
That is all for now. I was in a mood to look at these numbers and do not know when I will be in a mood again, so just throwing this one out there before I start overthinking things.
I warned you there would not be any groundbreaking conclusions.