Which world cities were likely to exist, and which were made by chance?

I've since updated this post — in hindsight I was wrong about a few things. A newer version from July 2025 exists.

Like I said in the post before this one, I like to waste time staring at maps. I play around with Google Maps all the time, zooming in and out, clicking on random restaurants to read the reviews, and noticing the terrain that shaped the geopolitical borders.

I'm often asking myself things as I do this. What's it like there? Who lives here? Why is this city even here? One of those questions has led me down more than a few rabbit holes.

Why is this city even here?, I ask myself while reading the Wikipedia page for the Toledo Mud Hens. Turns out Toledo, Ohio was once one of the most important cities in America — the somewhat-lucky placement of the Erie Canal made it a useful trade center, and later became one of the country's largest manufacturing hubs. There's a world where it's one of the largest cities in the country today.

My Question

Of course, Toledo isn't one of the largest cities in the country. There's a huge mix of factors which led to New York, LA, Chicago, Dallas, and the rest of them being what they are today. It starts with geography, plus a big dash of sociology and history, and finishes with a huuuuuge amount of luck.

You can really scale up that why-is-this-city-even-here question into a larger question about the role of chance not just in our day-to-day lives, but in society as a whole. I ask myself this all the time:

If we went back in time to the start of human civilization and let history play itself out a second time, which of today's global cities would exist again, and which are only around now from random chance?

I've thought about this a lot, and while I'm no anthropologist or geographer, there are some obvious answers. I'll get to a list in a moment, including a few places that aren't global cities today that maybe could have been.

The Criteria

For me, the simplest form of this question is judging important, global cities. If a city is in the top 300 or so cities, in terms of importance, there has to be good reason for that, right?

There are a few measures of city classification, but I'm going to use the Global and World Cities Research Network classification from 2020, which I've mapped.

There are 394 cities on this list, with New York and London tied at the top as the world's two most connected cities, all the way down to a hundred or so cities simply deemed "self-sufficient", including such esteemed locales as Tulsa, Oklahoma and Lomé, Togo.

Though there's undeniably room for debate regarding the order of this list (is London more "important" in 2022 than Beijing? Probably not), it's a good proxy for a binary question of important-or-not-important. So, let's use it.

I'm seeking out cities on this list which, if we spun back time and threw another roll of the anthropological dice, would be founded, built, and populated a second time, eventually appearing on a similar list of cities when humanity again hits ~7.7 billion people (assuming that would happen — but that's its own question).

Of course, I don't mean exactly the same. At all. None of these cities would have the same names, the same streets, the same population sizes in this alternate timeline. But many of their geographical locations would feature an alternate city at the same physical place, and that's what I mean.

A couple cities on the GaWC list would be likely (say, greater than 90% chance) to be "recreated" and make the list again. Some cities on the list would be very unlikely to exist again in their current state (making the new list less than 5% of the time). And some places that aren't on today's list could make it big if they got a 2nd chance.

So we're looking at two lists: the real list from GaWC, and a hypothetical list from an alternate universe.

Why would there even be a difference between these two lists?

Good question, Riley. Why am I assuming that the locations of global cities would change? Weren't today's cities founded and built where they are for good reason? Why would that reason not happen if we redid history?

Well, I'd actually posit that few of the cities on this list would actually exist a second time! There's far too much random chance in human history.

Forget going back to the beginning of humanity. Let's roll back the clock just to 569 CE and give Muhammad's mother a well-timed fever — no Muhammad, no Islam. How different are the last 1500 years? Scores of cities and towns would never be founded. Empires would never be built, and their settlements never created. Instead, completely different empires would have grown and fallen, leaving behind completely different buildings and roads.

We don't even have to go back that far to change the list. Aleppo, Syria was a city of 4.6 million in 2010. Is there a chance that the last few decades could have gone completely differently for that city, and it's important today? Absolutely.

Cities are made somewhat organically, but also at times extremely intentionally. It's this divide that really informs my original question. Physical locations that naturally have geographic, resource, and economic importance — completely independent to the lives of any individual human — would be likely to spawn cities. But they're rare. Places that lack one or more of those advantages likely grew mostly on human whim.

So which ones are they?

Cities Likely to Exist Again

Like I said above, I think few global cities would exist twice. But there are a couple that would, and then a solid 10-20 more that at least have a good chance to exist (maybe in 1-in-3 universes, or more). This deserves more thought — and probably a data-heavy dissertation, honestly — but here's a handful at the top of my mind.

1. Istanbul

Above every other city in the world, Istanbul is the one I come back to again and again as the most likely to exist on a large scale. When I think about rolling back all of human history, things could turn out massively different. Maybe Africa outpaces Europe in economic growth long term. Maybe the Americas are densely settled far earlier — or far later. But one thing that will never change? The Bosporus Strait being one of the most important crossroads for three continents.

Most of Europe needs to cross this strait to get to most of Asia and Africa, and vice versa. No matter what happens in any 2nd iteration of human history, and the kings, religions, cities, and countries that dominate, this geography will remain constant. It's not even that the most common pathway of travel will pass through today's Istanbul — it's that every traveler will have to stop on one side of this strait to get to the other. Travelers from most places on three continents will have to stop in one single place over and over… trade (and settlement) is inevitable.

Maybe the "major" part of a Bosporus city ends up on the Asian side the 2nd time, or closer to the Black Sea. And it might not grow to the same size if Constantine isn't around to name it the capital of the Roman Empire. But there will be a significant city along this strait almost every time history is restarted.

2. Cairo

Cairo has some of the advantages of Istanbul. It's an inevitable crossroads of travel, with nearly all of Africa funneling towards the Sinai to get to Europe and Asia. The Suez Isthmus will always be critical, as it was even before a canal was built there. It's a crucial location for trade.

More importantly, this includes all of the people and goods sent up and down the world's longest river, the Nile. Cairo is at the base of the Nile's delta, a highly fertile area between two deserts, so settlers will come as agriculture does. Plus, its proximity to the actual origin of humanity means Cairo is destined to be a point of travel for even the very first humans to venture away from their home continent. It's going to be as likely and as old as anywhere.

3. New York

When you think about things that make New York New York, they revolve around the incredible diversity and huge mass of humanity that has come to exist in a relatively small area. Ships of immigrants from every country on earth have docked at Ellis Island for a couple of centuries. And it's not just people — New York is one of the busiest container ports in the world, which is impressive for a place that (in 2022, at least) does relatively little exporting. This is all a flywheel. Ships bring people, people need stuff, ships bring stuff, stuff attracts people, and so on and so on.

But how did this flywheel start? New York wasn't the first city on the east coast of the Americas, or more historically important than Boston or Philadelphia or other places. A lot of those immigrants simply wanted to get to America in general, not necessarily New York. But this place had one big thing going for it: a vast natural harbor.

The water between Manhattan, New Jersey, Long Island, and Staten Island is often called the best natural harbor in the world. It's huge, calm, deep, reliable, and protected from the elements. It leads to wide rivers that themselves can accommodate ships. It's safe from invasion, with ample local resources, and the ability to trade up the Hudson. New York has a great location.

This geographic advantage would exist no matter what happens with humans, so assuming there are people and goods on one side of the Atlantic that want to get to the other, it's likely that New York exists in a significant capacity. Is it as likely as Istanbul? Not by a long shot, in my opinion. Would it be the number one global city in most alternate timelines? Nope, almost never. But would it exist? I'd say yes, usually.

Honorable mention to some other major cities with natural harbors and shipping advantages that have an argument for their existence: Hong Kong, New Orleans, Sydney, and Boston to name a few.

4. Panama City

Another important shipping locale, this time in Central America. While sending Asians around Africa to get to Europe is enormously time-consuming and dangerous, it's even more so to get from the west coast of North America to the east coast via Patagonia. Canals were built to ease both of those burdens, and you can expect that to happen again as technology reaches that level of sophistication.

This is a point, though, where the rock-solid likelihoods of Cairo and Istanbul, and decent likelihood of strong harbor cities, begin to fade. Does a version of Panama City exist in 90% of alternate universes? Honestly, probably not. But this is more-or-less always going to be the narrowest isthmus between the Atlantic and Pacific. That's more than most cities have going for them.

5. Caracas

Ok, we're at the point where this begins to feel like a reach, but hear me out to keep this interesting. Sure, Caracas lacks the attributes of the choices above. It's not on a common trade route, and it's not a good location for shipping. It doesn't seem "destined" like Istanbul or Cairo (and probably isn't the 5th most likely, I just want to talk about it). But it has a few things going for it.

Caracas is in an extremely comfortable valley, near to the ocean but protected, with lots of the basic resources for a 19th-20th century settlement. When it was settled by the Spanish, it was considered the ideal place for them to set up a New World home base. It's a solid location for a city. There's almost certainly a settlement of some kind here in any alternate universe — in fact there's probably one that's bigger than most local competitors.

Where Caracas goes from "large regional town" to "important global city" comes later, with technology and the nearby discovery of some of the largest oil reserves in the world. As that oil is located and the ability to extract it is developed, northern Venezuela is going to become durably rich. Even in today's universe, where Venezuela's economy has collapsed due to bad government, Caracas remains globally important through the vast wealth accumulated through oil production. Oil-rich regions are going to be oil-rich in every alternate universe.

Put those two together, and it's why Caracas has a solid shot of existing twice. This region as a whole is going to almost always produce a global city as it extracts the natural resource wealth around it. The question is, where does that city lie? Caracas doesn't have to be a top-notch locale for a global city, it just has to be one of the better places in the region. If this valley is so nice to live in, it could be the biggest benefactor of that oil wealth time and time again, creating one of the world's fastest-growing cities many times over.

Cities Unlikely to Exist Again

I would argue that most of the world's 394 "most connected" cities exist through some degree of chance. Maybe a fleck of gold was found at just the right time, or a czar's daughter happened to marry some local prince over a dozen other suitors.

Even as cities are often built in "good" locations, many cities were built in places that just don't make sense — willed into existence by human whim. The circumstances that led to their creation aren't going to reappear if human history ran back from the beginning.

1. Salt Lake City

The greater Salt Lake region is one that I think of as an inverse Istanbul — it pops up over and over as I think through this exercise, but for less positive reasons. The banks of the Great Salt Lake would almost never produce a global city, and the fact that it did in this reality is truly an accident of history.

SLC exists right between the edge of a desert, a fish-less and undrinkable lake, and a burly, snowy, cold (yet relatively narrow) mountain range. It has little going for it geographically — few resources, a barrier to east-west travel, and a climate that isn't especially hospitable.

So why is there a city there today? All thanks to one extraordinarily unlikely religion, which lives in SLC simply because it wasn't welcome anywhere else. It's worth reading up more on the history of the Mormon Church. I'm no expert, but the basics are enough here. A New York farmboy declared himself a prophet and his new religion an "updated" version of Christianity. While he amassed a following, other Christians hated him for it, and ran him and his followers out of town repeatedly. At times they were more-or-less hunted, treated as less than human, unwelcome everywhere (their leader was eventually killed by one of his critics). They ended up living where no one else wanted to — between a desert, a mountain range, and a salty lake.

Without any one of a million tiny decisions, the Mormons don't exist at all, let alone in this specific location. And if they were any less of a success, that settlement would never become a true city, let alone a "global city" (which, yes, the GaWC includes SLC). While there's likely to be a small, outdoorsy small city or two in what's now the SLC Metropolitan Area, I'm thinking it'd be more the scale of Bend, Oregon or Idaho Falls — max.

As an aside, Las Vegas is also in a terrible location for a city (a desert, really? The water crisis there is no surprise). But Vegas began as a trading point between SLC and LA. No Mormons, no Salt Lake City, no Sin City.

2. Canberra

The capital of Australia, Canberra, is one of a series of planned capitals along with places like Brasilia, Naypyitaw, and our very own Washington, D.C.

It was specifically built to be its nation's capital — there was no previous city there that was named capital. It was a compromise between Sydney and Melbourne, each of which wanted to be capital but were on more-or-less equal footing. So a new one was built between them.

As we're thinking about all of human history restarting, we would never again get to the point of a political squabble between a city at Sydney and a city at Melbourne that leads to a planned capital between the two. And even if we did, Canberra wasn't even the first choice! The rule was that the Sydney government could choose any site towards Melbourne, provided it was at least 100km away, which they did. And their choice was overridden anyway. It's even unlikely that Canberra exists in this universe.

If either Sydney or Melbourne is slightly larger at the time of this debate, there's no debate. If Australia became multiple countries instead of one, perhaps there's no debate. Planned capitals are so unlikely to be repeated — they're a product of a bunch of random choices in an infinitesimally unlikely and hyper-specific power struggle, ending in a quid pro quo wherever there happens to be land available. There's a far less than 1% chance Canberra exists a second time if we start history over.

3. Riyadh

In another situation of a capital city that arose from highly unlikely circumstances, Riyadh somehow exists as an important global city. Despite being in the middle of the desert, not too too far from more fertile and livable locales, Riyadh stands today.

While I don't know much about its history, it's easy enough to see how wild those chances might be. In 1744, a local emir formed some key alliances and set out to unify the Saudi peninsula. There were some setbacks, wars, and later key rulers, but with time, this grew into what we now know as Saudi Arabia. The original nation started near Riyadh and grew from there, so it was the capital at the time and it remains the capital today.

This is obviously not a sustainable reason for the location of a capital city. Saudi Arabia had been host to various empires large and small for thousands of years. Rulers came and went regularly. Even this first state was destroyed in the early 1800s and the founder's grandson had to pick up the pieces. The reason this particular iteration of that recurring conquest is important today is largely because of timing: it happened relatively recently, and was still in existence when oil was eventually discovered there in the 1930s.

If these highly specific circumstances didn't lead to Riyadh being the capital of a large nation-state, it would never have global importance. It would be another dusty desert town amongst countless others on the Arabian peninsula. It just happened to be the birthplace of the right person at the right time.

4. London

Like in the previous section with Caracas, this probably looks like the point where I've jumped the shark. But again, hear me out. London's importance as a truly global city is not as likely as it may seem. Unlike New York (the only other "Alpha ++" city, according to GaWC), London is not at a highly strategic geographic location serving an entire continent. It's just a city on a river, like hundreds of others around the world.

London's existence today wasn't assured — it was destroyed numerous times along the years, and kept getting rebuilt. The fact it's as old and resilient as it is matters for how likely it might be to exist, sure. But it's not truly any more a strategic location than other places up and down the Thames. In fact, its access to the English Channel made it vulnerable to viking attacks for centuries, and there's a world where the largest city in the area ended up built a few miles away from the river. Some Brussels-style canals, anyone?

Ultimately, like other unlikely-to-exist-again cities, London's importance has historically drawn from political levers, not geographic advantages. It's not assured by any means that this particular island would again create the largest dynasty in the history of the world, leading to its capital's hypergrowth. And even if it did, and we're starting ALL of history over back to the days where neanderthals roamed, who's to say the leader of that island nation is even located along the Thames? Is there a world where Great Britain is again unified, but led by a city along the southern coast of the island, or on a river north of the Thames, or even — dear god — in Ireland?

Possible Global Cities — That Aren't Today

If, as I've claimed, most of the 394 GaWC classified cities don't exist if history were restarted, then what exists in their place? What cities aren't in the 394 most important today, but sneak in when Salt Lake, Canberra, and London cease to be?

Well, a lot of these regions would still produce a city, it just wouldn't be in the same place. No Riyadh, but maybe Buraidah is that much bigger. Say Chicago doesn't get lucky but Gary does instead.

Other cities would get created in the extremely human processes that lead to all of these unlikely cities in the first place. Maybe, just as Brasilia was ordained into importance in our universe, a hubristic nation-state builds a nonsensical capital on the Atacama plateau in another one.

But there are a handful of places today that might have been likely to produce a city — their strong cards simply didn't win this hand.

1. Charleston

When talking about New York above, I went on about its natural harbor and how that led to trade. I called this a durable, geographic advantage, that would still exist as long as there was trade across the Atlantic. Well, there are other places like that along the eastern seaboard of the US. Charleston is one of them.

The natural harbor in Charleston is real, though not nearly as large and useful as New York's. But it's good enough, and that led to Charleston being founded early in American history, and put it in the top ten American cities in size heading into the Civil War (let's not get into what exactly they were trading at the time). The war and its aftermath put a stop to the city's growth, and South Carolina as a whole stagnated for the rest of the century, as did much of the southern US. If history works out differently, Charleston is significantly larger and more important today.

(As an aside, that's not to say that slavery would need to exist today for that to happen. If we started back many millennia ago, who knows what's different? The largest indigenous American city: Charleston, South Carolina!)

Regardless of how human history plays out, there will usually be a settlement of some kind in Charleston's place, and that is enough for it to be relatively likely to be a global city. Most coordinates where North America meets the Atlantic will not reliably produce a town, let alone a city, but Charleston (plus honorable mentions Savannah and Jacksonville) will always have a fighting chance.

2. Santa Barbara

The reasoning for Santa Barbara is pretty similar to that for Charleston: it's a likely settlement due to some geographical advantages, and that's enough to make it a city more often than other places. But I don't think people realize how close Santa Barbara was to becoming what Los Angeles is today — the predominant metropolis of the southwestern US.

When California was originally settled by the Spanish, three sites were chosen to build the first settlements and forts: Los Angeles, San Jose, and Santa Barbara. Obviously one of those didn't have fortunes like the other two. There are valid reasons for that — Santa Barbara doesn't have as sizable a valley for a city as LA or San Jose, for example. But it also simply didn't get lucky. The fort in LA was built first, so LA began growing before the Spanish built in Santa Barbara. Reversing that order could have easily led to a shift in the preferred destination at the time, giving the first city of the two a fair bit of inertia above the other.

One big advantage to Santa Barbara is its proximity to the largest natural harbor between the Bay Area and San Diego, operating today as Port Hueneme. This harbor historically didn't serve LA due to the mountains in between the two (LA instead built a pair of high-traffic artificial harbors), but it could have served a hypothetical Santa Barbara metropolis. That New York flywheel would spin and spin. It's really not that farfetched.

3. Izabal, Guatemala

Here we go again — the part where I ask you to hang with me. "But Riley," you say, "there isn't even a city along Lake Izabal now. There's barely a town! What makes you think there'd be a major global metropolis if human history started over?"

To that I say this: if my guesses above panned out, and London and Las Vegas and Canberra and all the others didn't exist, the idea of them being cities would seem as ridiculous as this proposal in Guatemala. "There's no way an important city would ever exist in the Nevada desert", alternate universe you is saying to alternate universe me.

So like I said, hang with me.

This starts with a recurring theme — shipping. This corner of the Gulf of Mexico is a strong area for harbor building. There are a few minor ports here, and the fact that they aren't bigger probably says more about the exporting output of Belize and Guatemala than about the geography here. If all of human history starts over, this area as a whole might be significantly more populated than today, which would make a protected freshwater lake with direct access to desirable shipping lanes highly, highly desirable for settlement in the area. There's even a perfect little valley for a city on the western banks! It's more than possible.

But where would all those people even come from? The gold, of course. I'm joking a bit, but it's worth noting that the mountains on either side of that valley are home to some of the world's largest known gold reserves. Gold rushes have seeded a number of GaWC cities — San Francisco, Melbourne, Denver, Johannesburg, Belo Horizonte… why not Izabal?

4. Aleppo and more

As I noted earlier, Aleppo was a city of 4.6 million in 2010 before the current civil war took hold. In 2006 it was called the Global Capital of Islam, a noteworthy title. Today, it's unrecognizable, devastated. This means more than its status as a global city, obviously, but it's in a place where it can't be considered "important" by that definition today. A few different outcomes over the last few decades and it could look a lot different, as complicated as the situation is. Bad luck played a role.

This gets back to details I've mentioned about many of the other cities here: arbitrary political outcomes make or break a lot of cities. I don't have to argue Aleppo's potential as a global city. That's clear. Just as political luck made Riyadh, the inverse happened here.

Aleppo is just one of many cities around the world that have clear potential as an important, global city, but for political reasons (that I won't get too much into) aren't able to realize that potential.

Havana is larger than many GaWC cities, and well-known around the globe. Its ceiling in a different world is what Miami is today — a rich, highly desirable, world-famous vacation destination.

Pyongyang was the second largest city in Korea before the country split. Seoul has grown into one of the largest cities in the world, and it's easy to think that Pyongyang would have joined it to some degree.

Tripoli and Tehran, like Aleppo, are historically important cities. Instability has hindered their connection to the rest of the world for the last few decades, but a list of "important global cities" made in the 1970s might have included both.

End

This question is interesting to me for a couple of reasons.

One, it gets at the role of random chance in the shaping of our world. This is something that we all know to some degree — the butterfly effect is the famous metaphor here. Going back in time 10,000 years and whacking any random person on the head creates an entirely different world, and it's interesting to think about the implications of that regarding global cities.

Two, it's a fun way to think about cities in general. Why do they exist? What makes them? Which factors are important in a big city? The answer to what's important? is often simply that someone decided it's important! There's a city built on a narrow stretch of land between the Black Sea and the Mediterranean that's enormously unique and useful, meanwhile some guy named Saud accidentally created a city in the middle of the desert that ended up in a similar range of importance. When thinking about building things in the real world today, it's important that you decide to build them.

I think it's worth digging into this question a bit more, and I'll try to do so with time. There's surely a data-first way to address it at scale, maybe with a Bayesian search theory approach to determine ideal city locales. I also might attempt to crowdsource some opinions from anthropologists, or even just from a hypothetical twitter following (we'll see if I ever get one of those).