As an atlantian I can confirm life sucks here. Every morning my wife wakes up and asks “is it going to get better?” I reply, “no, our node density per square kilometer is too low”
3:53 I don't think there was anything incorrect about Moscow and Saint Petersburg. Russian and ex-Soviet cities have much less roads than in many other countries since the city blocks are generally much bigger.
adding to that- the large size of these residential blocks - "microregions" - is because each block will have many buildings of various heights inside it connected by pedestrian paths and small streets and containing grocery stores and pharmacies, convenience stores, small parks/courtyards, schools, and playgrounds within easy walking distance of the residential buildings.
A dense grid certainly creates a lot of nodes and edges, but I don't think I'd consider it complex. Instead of average node degree I'd probably go for the standard deviation of node degree. A city that has a wide variety of funky intersections should score higher than a city where every intersection follows the same pattern.
Great video! Those node diagrams of the cities really remind me of the book "City Unseen" with their colors and the way only the road network is visible
08:52 Honest question: How would Bcn's grid be better with 2-way streets? (Seems to me like it would add even more to its induced traffic and thus to its congestion?)
This video was just looking at interesting data about roads, not making any genuine comments about how cities should be planned. That was just a joke about how surprisingly low Barcelona's node density was.
What were the *least* well-connected for the Netherlands? Regarding cities probably Arnhem, Deventer and Zwolle for sure, but there may be some lesser known towns that finished high?
Cool to see Atlanta here! Another way to interpret its low density is a very high degree of sprawl. Atlanta is seeing a lot of recent growth with few natural barriers, so your radius contains a few large parks and forested areas that are technically part of the city, but have seen very little development.
Did you consider that in well mapped areas single street might be divided into many individual segments. For example for speed limits, parking restrictions etc meaning that nodes of those segments have only two connected segments and are actually not changing the general connectivity since they are basically the same street. only real intersections should count
For example, my city (Cracow) has and OLD centre. But the newer parts have a not so dense but well designed road system, with the main roads built like a spider's net - "sun rays" crom the centre with ring-like roads circling it, the further from the center, the faster the rings.
I have a interesting Question, your accent reminded me a lot of the one from another youtuber. I did a bit of research, but couldn't really find anything that matches well. your channel location is set to cyprus, but turkish and greek accents, didnt really sound close to yours. So could you pls do me a favor and tell me what kind of accent you have?
Is the videos title image customized to viewers? I am from Berlin and the title image looks like Berlin to me. But it wasnt even mentioned or shown in the video. So I am wondering in general if youtube lets you customize this for geographical target groups.
I'm curious if you can compare how much a city's roads look like a matrix vs a space filling tree. A space filling tree would give an ideal connection to every point of a graph, but minimizes the amount of road needed.
Just curious if any of these metrics are considered better or worse. The reason I ask is that my own city, Toronto, doesn't stand out in any of these metrics, but we are one of the worst cities in the world for traffic congestion (only London and Dublin are worse). Of course these metrics are but a few of many factors that lead to congestion, just curious if any of them had any significance.
I really love how you approach the topic. I also have to deal with OSMdata bc of city planning studying. Although I do know how to get data out of osm, I do sometimes struggle with the sheer amount of it the more administrative levels I have to go up. In that moment I realize that I am studying city planning and not planning on larger scales (forgot terminology). Anyways, I really love the way you address the clear problems of the mess of administrative levels with an absolute definition of the boundaries (plus having the knowledge to actually automise the data output! That wouldve come really handy for my friend who just finished his bachelor's thesis) thanks for reading my ramblings
Wondering how La Plata scored given it's famously planned and gridded road network. Most interesting to me is the rigorous adherence to numbering the streets. There are small towns on the outskirts to the east of the city proper that have road numbers in the six and seven hundreds. Plenty of people in plenty of cities can claim to live on Tenth Street or Fiftieth or maybe even Hundredth. How many can say they live on a Seven-hundred and Seventh Street?
8:00 - Why is Hong Kong not included in the data, considering that only Mainland Chinese cities appear less dense on OpenStreetMap due to side road omission?
Low node density is what you want in residential areas. In other words, no cut-throughs or rat runs for commuters, and safe streets for children to play in, and general peace and quiet.
At least Manhattan is. Once you go into any of the outer boroughs, street grids become messy. I'd say I'm looking specifically at Queens and SI. But as someone who's lived in both the Bronx and Brooklyn, I can say it's pretty bad there too. Though Brooklyn is definitely better than the other 3.
Sounds like an interesting concept. But smaller cities can be more complex. So I think that looking at a city’s population isn’t the best way to do this. What I suggest is to look at countries instead. This will also include smaller cities. And this way we can see what country is the most complex. Or we could see what country has the best road system.
How about weighing the connections by the road types? (arterial, connector, local) this would vary the road system between urban sprawl with dead ends and well designed cul-de-sac networks.
São Paulo is chaotic, but in a way it is efficient because the public transport system works, even though I go to work with 30 people on top of me KKkkkk. Why is it chaotic? Simple, many people travelling at the same time in the morning or late afternoon/early evening, I know people who make a journey from the beach (Santos - SP, 96km) to São Paulo and take less time than I do to get to the service living about 30-40 minutes without traffic.
It's only the northern Chinese cities that are not that dense but rather gridlike. I'm disappointed at least 20% of this video wasn't about Chongqing and similar places. If you yourself said some data is unreachable, which means you eliminated 17% of the world population, don't give your videos such all-encompassing titles. Still, good idea for a video
also true in india , while he included the tier 1 cities india has way more population in tier 2 like patna. the road network is also way better and planned in comparison to say delhi or bangalore
Chat gpt isn’t research! Please just look it up at the least, having verified information is very valuable and relying on LLMs for information is a bad reflection on you and your intellectual capacity
As an atlantian I can confirm life sucks here. Every morning my wife wakes up and asks “is it going to get better?” I reply, “no, our node density per square kilometer is too low”
Suburbs at fault
Your wife has a great dedication at being annoying then...
@@ensiehsafary7633 you take that back. i love my wife. she is soft and she says nice things to me sometimes.
3:53 I don't think there was anything incorrect about Moscow and Saint Petersburg. Russian and ex-Soviet cities have much less roads than in many other countries since the city blocks are generally much bigger.
adding to that- the large size of these residential blocks - "microregions" - is because each block will have many buildings of various heights inside it connected by pedestrian paths and small streets and containing grocery stores and pharmacies, convenience stores, small parks/courtyards, schools, and playgrounds within easy walking distance of the residential buildings.
Barcelona isn't full of one-way streets if you're a pedestrian
Never thought I see my hometown making it top 1 in any list, great video man
Which one was your hometown?
@ensiehsafary7633 San Luis
@@ensiehsafary7633I'm guessing it's San Luis Rio Colorado
A dense grid certainly creates a lot of nodes and edges, but I don't think I'd consider it complex. Instead of average node degree I'd probably go for the standard deviation of node degree. A city that has a wide variety of funky intersections should score higher than a city where every intersection follows the same pattern.
Interesting concept, though certainly a tough one to measure without compromises
i feel like node degree isn't so much a measure of complexity but rather the likelihood of being hit by a car
So people really be playing permanent battleships in San Luis Río Colorado
Always bringing original stuff. Well done!
Squarest country: ...
@@隋隋隋 Almost
Great video! Those node diagrams of the cities really remind me of the book "City Unseen" with their colors and the way only the road network is visible
08:52 Honest question:
How would Bcn's grid be better with 2-way streets?
(Seems to me like it would add even more to its induced traffic and thus to its congestion?)
This video was just looking at interesting data about roads, not making any genuine comments about how cities should be planned. That was just a joke about how surprisingly low Barcelona's node density was.
More road capacity induces traffic as far as I know and one-way streets reduce capacity it so... ?
Ome way road do not reduce capacity, one way roads have two lanes to the same direction instead of two lanes on different dorection. @@iliqiliev
5:22 Holy Sh*t, you flooded our airport!!!
The Jorge Chavez passed from airport to submarine hub lol
I vote Metro Manila cause that city is clearly built by people who don’t know what a city is and what is the purpose of a road.
What were the *least* well-connected for the Netherlands? Regarding cities probably Arnhem, Deventer and Zwolle for sure, but there may be some lesser known towns that finished high?
this channel goes so hard dude
Cool to see Atlanta here! Another way to interpret its low density is a very high degree of sprawl. Atlanta is seeing a lot of recent growth with few natural barriers, so your radius contains a few large parks and forested areas that are technically part of the city, but have seen very little development.
Did you consider that in well mapped areas single street might be divided into many individual segments. For example for speed limits, parking restrictions etc meaning that nodes of those segments have only two connected segments and are actually not changing the general connectivity since they are basically the same street. only real intersections should count
Very cool video!
I'd love to see my city's score on the list
For example, my city (Cracow) has and OLD centre. But the newer parts have a not so dense but well designed road system, with the main roads built like a spider's net - "sun rays" crom the centre with ring-like roads circling it, the further from the center, the faster the rings.
Can we see the list of all the cities.
I have a interesting Question, your accent reminded me a lot of the one from another youtuber. I did a bit of research, but couldn't really find anything that matches well. your channel location is set to cyprus, but turkish and greek accents, didnt really sound close to yours. So could you pls do me a favor and tell me what kind of accent you have?
Is the videos title image customized to viewers? I am from Berlin and the title image looks like Berlin to me. But it wasnt even mentioned or shown in the video. So I am wondering in general if youtube lets you customize this for geographical target groups.
same lol, I waited for Berlin the whole time
Edit: it's probably in the thumbnail because it is objectively the best city
your job is so exciting
I love geospatial data, I wish I knew python too
I'm curious if you can compare how much a city's roads look like a matrix vs a space filling tree. A space filling tree would give an ideal connection to every point of a graph, but minimizes the amount of road needed.
Just curious if any of these metrics are considered better or worse. The reason I ask is that my own city, Toronto, doesn't stand out in any of these metrics, but we are one of the worst cities in the world for traffic congestion (only London and Dublin are worse). Of course these metrics are but a few of many factors that lead to congestion, just curious if any of them had any significance.
cool video man
I really love how you approach the topic. I also have to deal with OSMdata bc of city planning studying. Although I do know how to get data out of osm, I do sometimes struggle with the sheer amount of it the more administrative levels I have to go up. In that moment I realize that I am studying city planning and not planning on larger scales (forgot terminology).
Anyways, I really love the way you address the clear problems of the mess of administrative levels with an absolute definition of the boundaries (plus having the knowledge to actually automise the data output! That wouldve come really handy for my friend who just finished his bachelor's thesis) thanks for reading my ramblings
That was awesome! I cant imagine how much work that 10 minute video mustve taken!
Wondering how La Plata scored given it's famously planned and gridded road network. Most interesting to me is the rigorous adherence to numbering the streets. There are small towns on the outskirts to the east of the city proper that have road numbers in the six and seven hundreds. Plenty of people in plenty of cities can claim to live on Tenth Street or Fiftieth or maybe even Hundredth. How many can say they live on a Seven-hundred and Seventh Street?
8:00 - Why is Hong Kong not included in the data, considering that only Mainland Chinese cities appear less dense on OpenStreetMap due to side road omission?
Brilliant
I was very surprised to see Wellington at 2:47
Nice, could you talk about biking infrastructure next?
@JohnnySmartie should be possible. at least for Europe and the US, bike infrastructure is reasonably well mapped in OpenStreetMap.
3:12 I live in Tehran and our population isn't 16 million where did you get that data?
2016 Census notes the metropolitan area as have 14 million. 2024/25 estimates show 16 million in the metro area
Really cool video! Surprised you have so few subs (at least for now 😊). Carry on! This content looks very original and roads are generally cool 😁
Low node density is what you want in residential areas. In other words, no cut-throughs or rat runs for commuters, and safe streets for children to play in, and general peace and quiet.
Peruvian from Lima here. Our traffic jams are so bad that they are used to train AI
What about Athens, Greece? Are there any results for the cities on a table? Could the approach apply for smaller ones?
I clicked on the video because I recognized the city in the thumbnail as Berlin. Did that stand out particulary?
I live in NYC, so I think we're pretty gridded and well connected
At least Manhattan is. Once you go into any of the outer boroughs, street grids become messy. I'd say I'm looking specifically at Queens and SI. But as someone who's lived in both the Bronx and Brooklyn, I can say it's pretty bad there too. Though Brooklyn is definitely better than the other 3.
Sounds like an interesting concept. But smaller cities can be more complex. So I think that looking at a city’s population isn’t the best way to do this. What I suggest is to look at countries instead. This will also include smaller cities. And this way we can see what country is the most complex. Or we could see what country has the best road system.
that's a different question though
How about weighing the connections by the road types? (arterial, connector, local) this would vary the road system between urban sprawl with dead ends and well designed cul-de-sac networks.
Does higher node degree means higher congestion?
Great Video! Could we have the full list of cities, I want to see where mine ranks?
How was Monterrey, Mexico not included. It has a larger population than Guadalajara
San Luis Río Colorado????? That one???
Look at Varanasi's street map, no where in the world has streets as complex as varanasi
São Paulo is chaotic, but in a way it is efficient because the public transport system works, even though I go to work with 30 people on top of me KKkkkk.
Why is it chaotic? Simple, many people travelling at the same time in the morning or late afternoon/early evening, I know people who make a journey from the beach (Santos - SP, 96km) to São Paulo and take less time than I do to get to the service living about 30-40 minutes without traffic.
Can you do railways next?
One way streets ftw
*insert commenter’s hometown here*
>thumbnail is berlin
>doesn't mention berlin
:(
Very cool
Good video!
Wait omg another conlanger omg omg
@@Patriasburg Yeah
i didnt know monowi is just a badly drawn tetris
It's only the northern Chinese cities that are not that dense but rather gridlike. I'm disappointed at least 20% of this video wasn't about Chongqing and similar places. If you yourself said some data is unreachable, which means you eliminated 17% of the world population, don't give your videos such all-encompassing titles. Still, good idea for a video
also true in india , while he included the tier 1 cities india has way more population in tier 2 like patna.
the road network is also way better and planned in comparison to say delhi or bangalore
Dubai is crazy
21 pilots?
Chat gpt isn’t research! Please just look it up at the least, having verified information is very valuable and relying on LLMs for information is a bad reflection on you and your intellectual capacity
having a nose ring also reflects badly on you and your intellectual capacity, but I doubt you'll take it out
pulling out actual peer-reviewed research content from an international group specializing in remote sensing and geographic data isn’t research?
He was making a joke. Did you no see the literal peer reviewed article he cut to right after?
this was great - buy a new microphone
Wow im surprised not to see mumbai or dhaka on node density those cities are pure choas
Data guy