Citystrides in Big Cities: Some Numbers

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.

The Data

Getting It

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.

What Is A City?

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.

What Is A Street?

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.

All Cities

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.

street count
mean 97.409907
std 394.551961
min 1
25% percentile 10
50% percentile 27
75% percentile 76
90% percentile 194
95% percentile 349
99% percentile 1182.43
99.9% percentile 4314.172
max 44082

The Final Selection

So these are the cities currently in Citystrides with 5000 or more streets. Brussels just makes the cut with 5007 streets. Happy little accidents.

cs_id name region_country street_count
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
3 225687 القليوبية القليوبية, مصر 30858
4 225681 محافظة الجيزة محافظة الجيزة, مصر 22319
5 177521 City of Cape Town Western Cape, South Africa 22308
6 225679 محافظة القاهرة محافظة القاهرة, مصر 20787
7 225680 القاهرة محافظة القاهرة, مصر 20734
8 177494 eThekwini Metropolitan Municipality KwaZulu-Natal, South Africa 19706
9 33385 Rio de Janeiro Rio de Janeiro, Brasil 17953
10 227695 محافظة الرياض منطقة الرياض, السعودية 17939
11 167292 Loire Auvergne-Rhône-Alpes, France 16959
12 177491 City of Johannesburg Metropolitan Municipality Gauteng, South Africa 16154
13 94322 Roma Lazio, Italia 15970
14 7696 Jacksonville Florida, United States 14816
15 177488 City of Ekurhuleni Metropolitan Municipality Gauteng, South Africa 13949
16 188747 Brisbane Queensland, Australia 13813
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
22 223551 กรุงเทพมหานคร กรุงเทพมหานคร, ประเทศไทย 12330
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
32 131268 Toronto Ontario, Canada 10443
33 33516 Goi√¢nia Goi√°s, Brasil 10193
34 2550 San Diego California, United States 10112
35 132395 Berlin Berlin, Deutschland 10075
36 36171 Salvador Bahia, Brasil 9801
37 218665 Petaling Selangor, Malaysia 9717
38 32499 Curitiba Paran√°, Brasil 9698
39 5637 Las Vegas Nevada, United States 9665
40 173609 Fort Worth Texas, United States 9628
41 160001 Juárez Chihuahua, México 9575
42 171388 Calgary Alberta, Canada 9492
43 32406 Manaus Amazonas, Brasil 9365
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, España 8817
49 221066 Jakarta Timur Daerah Khusus Ibukota Jakarta, Indonesia 8710
50 160222 Monterrey Nuevo León, México 8632
51 225689 محافظة الإسكندرية البحيرة, مصر 8492
52 131267 Ottawa Ontario, Canada 8479
53 36434 Campinas S√£o Paulo, Brasil 8342
54 22080 Louisville Kentucky, United States 8295
55 159441 San Luis Potosí San Luis Potosí, México 8218
56 140129 Hamburg Hamburg, Deutschland 8125
57 159392 Puebla Puebla, México 8032
58 160068 Morelia Michoacán de Ocampo, México 7991
59 131066 Birmingham England, United Kingdom 7990
60 35801 Fortaleza Cear√°, Brasil 7797
61 18074 Nashville-Davidson Tennessee, United States 7766
62 132539 Praha Praha, Česko 7595
63 31374 Recife Pernambuco, Brasil 7545
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
68 159898 Chihuahua Chihuahua, México 7321
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
73 42638 Budapest Budapest, Magyarorsz√°g 7027
74 159833 Zapopan Jalisco, México 6948
75 187823 Wiltshire England, United Kingdom 6846
76 130668 Osmangazi Bursa, Türkiye 6815
77 132430 Ecatepec de Morelos Estado de México, México 6762
78 221065 Jakarta Selatan Daerah Khusus Ibukota Jakarta, Indonesia 6708
79 159332 Aguascalientes Aguascalientes, México 6592
80 5732 El Paso New Mexico, United States 6541
81 132396 Wien Wien, Österreich 6497
82 218707 Kinta Perak, Malaysia 6484
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
89 207703 عمان عمان, الأردن 6088
90 132529 Iztapalapa Ciudad de México, México 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
95 171349 München Bayern, Deutschland 5905
96 130669 Yıldırım Bursa, Türkiye 5758
97 131085 Liverpool England, United Kingdom 5695
98 48513 Paris Île-de-France, France 5673
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 Querétaro Querétaro, México 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
111 124926 Rotterdam Zuid-Holland, Nederland 5318
112 188751 Torreón Coahuila de Zaragoza, México 5318
113 213880 Saltillo Coahuila de Zaragoza, México 5312
114 159171 Acapulco de Juárez Guerrero, México 5255
115 159265 Hermosillo Sonora, México 5225
116 187804 Cheshire West England, United Kingdom 5184
117 189835 Warszawa województwo mazowieckie, Polska 5184
118 113291 Göteborgs Stad Västra Götalands län, Sverige 5162
119 95259 Durango Ciudad Durango, México 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 João Pessoa Paraíba, 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.

Most Completed Cities

Let’s look at how many of these are completed or near completion. Also considering some variations on this idea.

Percent of Top Striders

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
1 Stephen Peck 1 100 Toronto 10443
2 Monkey Man 1 100 Manchester 6320
3 Michael Shanks 1 99.9 Glasgow 6161
4 Tony Woods üå± 1 99.77 Liverpool 5695
5 Denis Bafounta 1 99.11 Berlin 10075
6 Ward Muylaert 1 98.86 Brussels 5007
7 Dan Schoettinger 1 96.28 Colorado Springs 5944
8 William Skorupinski 1 91.05 Ottawa 8479
9 Krzysztof Kasprzyk 2 88.6 Brussels 5007
10 Richard Denis 2 85.48 Toronto 10443
11 Just_one_more_street 1 77.31 Wien 6497
12 Anonymous 1 76.08 Austin 9236
13 John Teal 1 75.99 Columbus 10750
14 Graham Holland 2 75.01 Liverpool 5695
15 Al Vermette 2 71.43 Ottawa 8479
16 Jim Fulton 1 71.07 Sheffield 6342
17 Matthew Stroh 3 65.27 Toronto 10443
18 rick richard 1 64.12 Calgary 9492
19 Fabrice DUPOND 1 63.71 Paris 5673
20 Anonymous 2 61.55 Calgary 9492

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.

Number of Streets by Top Striders

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
2 Stephen Peck 1 100 10443 Toronto
3 Denis Bafounta 1 99.11 9985 Berlin
4 Richard Denis 2 85.48 8927 Toronto
5 Marshall Butler 1 56.58 8383 Jacksonville
6 John Teal 1 75.99 8169 Columbus
7 William Skorupinski 1 91.05 7720 Ottawa
8 Anonymous 1 76.08 7027 Austin
9 Matthew Stroh 3 65.27 6816 Toronto
10 Sean P 2 17.41 6816 Greater London
11 Monkey Man 1 100 6320 Manchester
12 Brian Grinnell 2 58.08 6244 Columbus
13 Michael Shanks 1 99.9 6155 Glasgow
14 rick richard 1 64.12 6086 Calgary
15 Al Vermette 2 71.43 6057 Ottawa
16 Anonymous 2 61.55 5842 Calgary
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
20 Anonymous 3 59.17 5616 Calgary

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.

First Page Mean

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.

City Average Pct
1 Toronto 41.3167
2 Brussels 40.1708
3 Ottawa 37.7842
4 Wien 32.6258
5 Glasgow 32.345
6 Paris 32.3417
7 Sheffield 32.0925
8 Liverpool 31.6767
9 Austin 30.575
10 Calgary 26.2125

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.

Third Place Strider

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.

Name Pct City
Matthew Stroh 65.27 Toronto
Anonymous 59.17 Calgary
Barbara Dundas 58.98 Ottawa
Tobias Barnett 55.06 Brussels
Jackie Howard 50.57 Austin
Sobras 47.23 Wien
Joe Carruthers 45.7 Sheffield
JB BOMMIER 44 Paris
Gabe Boer 41.46 Rotterdam
Geof Hileman 40.78 Raleigh

I Just Want To Be First Page

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
Andrew Greenberg Toronto 2068
Heather Jones Sydney 1340
Jesse Blondin Ottawa 1319
Anonymous Hamburg 1131
Anonymous Sheffield 1130
Anonymous Glasgow 1074
Anonymous Leeds 1026
Anonymous Paris 1013
Dena Childs Austin 976

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.

Big Cities vs Top Striders

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.

Caveats

  • 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.

Show Your Work

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.

Conclusion

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.

17 Likes

Thanks for sharing this, great write-up!

I thought your point about “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.” was interesting. Looking at the first list I was surprised at the number of US cities on there. Places like Charlotte and Indianapolis are very street-heavy for cities with <1m population. So more and longer streets? Tough being a strider in the states :laughing:

I can speak for Indianapolis at least since I live nearby. The city actually takes up an entire county. Most of the long streets are part of the grid beginning downtown, but as you get further and further away it turns more into random neighborhoods connected to the “main” streets of the grid. This runs up the street count.

Compare a city like Chicago, which is almost only a grid. It has 700 more miles than Indianapolis, but only 2200 streets!

3 Likes

Same with another grid city, Philadelphia. I’ve run there a few times but rarely get streets because they continue on either side of the Schuylkill River…very difficult to ‘complete’ most parts of the grid. The top Philly strider is only at ~15%!

1 Like

Very interesting read. Lots of good points raised.

I believe New York City was left off your list. It comprises of 5 counties/boroughs in it of itself (Brooklyn, Queens, Staten Island, Bronx and Manhattan).

1 Like

Correct, because it doesn’t seem to be in Citystrides. It does exist on OSM. The five parts that make it up are in CS. Combined they would be close to 10k streets.

1 Like

Great bit of data journalism there! Nice to see my name in a few of the lists, and you’re right, 99.9% is the closest I can get at the moment having been at 100% at the start of the year. Will be a constant battle to keep up with housebuilding in the city!

1 Like