City distances

[phipfwe76]

Hello all. I usually don't start threads because I find 99% of what I need using the search feature, but I'm trying to get some input on how far apart you like to space your cities. I'm fairly new at this, and it took me about a dozen plays just to get a grip on settler, but now 10 plays later I'm winning comfortably on prince, so I'm heading in the right direction.

I always place my cities a minimum of 6 tiles apart, simply because they will grow 3 tiles in every direction, and I don't want my cities having to fight over resources. But I notice a lot of videos and a lot of comments here seem to go about 4-ish tiles apart. Am I missing/overlooking something important?

 

[JesseS]

Yes, you’re missing 2 big things that make clustering cities about 4 tiles apart generally optimal.

1) Cities can practically never grow to the size to fully use all of the tiles in their vicinity. To actually work the entirety of 3 rings of tiles, a city needs 6 (first ring)+ 12 (second ring)+ 18 (third ring)= 36 population. Even after accounting for districts and unworkable tiles, you’re looking at ~20 population to work all the tiles, which is way too large to grow a city in civ 6. Most cities only need to be 10 population and to do so, they really only need a few strong starting tiles and some chops.

2) There are significant advantages to clustering cities. It lets them stack districts next to each other, providing far higher adjacency than a city can get on its own. It lets you fit in far more cities on any given piece of land. Districts are very powerful, and being able to get 1 more campus or 1 more trade route is worth far more than a marginal tile. If you’re willing to engage in micromanagement, swapping around a few high growth and high production tiles between cities depending on their housing status and what they are building is useful.

However, there are times when you want to space cities out. In domination games, you usually don’t settle too many cities overall, so you might as well space them out a bit more. In cultural games, settling some late game cities for national parks and/or seaside resorts is worthwhile, but these cities should be spaced out to efficiently claim land with as few cities as possible. This is because the power of late game settled cities is having tiles with tourism improvements, not districts as is the case in most other situations.

 

[phipfwe76]

Yes, you’re missing 2 big things that make clustering cities about 4 tiles apart generally optimal.

1) Cities can practically never grow to the size to fully use all of the tiles in their vicinity. To actually work the entirety of 3 rings of tiles, a city needs 6 (first ring)+ 12 (second ring)+ 18 (third ring)= 36 population. Even after accounting for districts and unworkable tiles, you’re looking at ~20 population to work all the tiles, which is way too large to grow a city in civ 6. Most cities only need to be 10 population and to do so, they really only need a few strong starting tiles and some chops.

Awesome input! I must be doing something wrong, because I usually finish the game with at least 4 cities that have a population of 20+. Right now I'm playing as Dido, and my capital is already at 21, and I'm on turn 450 of 750. Perhaps I'm making sure they're too well fed? I don't really like seeing the growth slow down, so I'm always making sure plenty of food is available.

2) There are significant advantages to clustering cities. It lets them stack districts next to each other, providing far higher adjacency than a city can get on its own. It lets you fit in far more cities on any given piece of land. Districts are very powerful, and being able to get 1 more campus or 1 more trade route is worth far more than a marginal tile.

I played as Germany a few weeks ago, and made sure I did this with Hansas, commercial hubs, and resources. It was probably my quickest win yet. I guess I never really though about doing it with other civs, but it makes perfect sense. Thanks!

 

[Noble Zarkon]

I must be doing something wrong, because I usually finish the game with at least 4 cities that have a population of 20+. Right now I'm playing as Dido, and my capital is already at 21, and I'm on turn 450 of 750.

Not so much you're doing it wrong but that most experienced players would have won by then, many of the better ones before turn 300 on Epic .....

 

[phipfwe76]

Not so much you're doing it wrong but that most experienced players would have won by then, many of the better ones before turn 300 on Epic .....

Sigh. And here I was thinking I was getting better.

 

[JesseS]

Right now I'm playing as Dido, and my capital is already at 21, and I'm on turn 450 of 750. Perhaps I'm making sure they're too well fed? I don't really like seeing the growth slow down, so I'm always making sure plenty of food is available.

The main thing is that food and population growth should be a means to an end, rather than an end itself. There are situations where growing a mega city is good (tons of mines to work in a future spaceport city, growing a city with Pingala or Mausoleum for the direct extra culture and science). But the bulk of your cities only need to get to pop 10 for 4 full districts and the rationalism boost. Population after that just requires more amenities and housing, and they would be better off instead capping growth and running projects on repeat.

I suggest looking at the tiles of a 10 pop city and asking “what’s the next tile they’re going to work after growing, and do I really need to grow to work it?” I would also suggest analyzing tiles by subtracting 2 food (the amount to support the citizen working it). Overall, working farms to grow more population to work more farms is not good play. And investing in neighborhoods, sewers, and entertainment complexes to support an endlessly growing population of farmers compounds the production inefficiency.

 

[Bostock]

One other reason to go with 4-5 tiles apart: since cities don't start acquiring third-ring tiles until the second ring is full, getting third-ring tiles (that aren't second-ring tiles of other cities) means either throwing tons of money at purchasing tiles or waiting for-ehhhh-ver.

 

[Siddharth Venkatesh]

Another reason to pack cities in more densely is the way global yields work in this game (science, culture, gold, faith). These yields scale better with more cities than "better cities", because a large percentage of the yields come from the adjacency, base value of buildings and envoy bonuses. All of these scale linearly with number of cities. Additionally, trade routes are very strong in civ 6 and one of the best way to centralize food and production in one city or generate vast amounts of gold. Ironically if you want a particular city to get very strong, the best way to do so is spam a lot of cities, build Commerce Hubs and then put all the trade routes in the city you want to strengthen.

On the other hand, having a "better" city gives bonuses only via:
1. +3 adjacency: this still largely doesn't care about how good the city is, but it might make you space a city out a bit more to guarantee enough adjacency. In fact, this tends to work better for cities settled closer to each other, as you can stack more districts next to each other for the +0.5 adjacency.
2. 10 pop: This is such a low threshold that it doesn't matter if you have a lot of tiles overlapping each other. It's also trivial to get a city to 10 with a few food chops.
3. Specialists: Unlike civ 5, where secularism in the rationalism social tree allowed very tall cities to be competitive in science via working specialist slots, specialists in civ 5 are weak.
4. More production: Having a city grow taller does give more production yields to the city, but cities tend not to grow to their third ring very quickly and production via extra population can be replaced by trade routes.

All this tends to make compact settlements better than spread out ones.

 

[Bibor]

Aiming for a few population 13 cities is reasonable.

For these bigger cities, that's:
13 for population (realistically 15-16 since you want to skip from farms to production at one point)
5 for districts
2-3 for wonders
1 for the Aqueduct
1-X mountains, desert, lakes and/or barren flatland.

For a total of what's usually around 26. So no, two rings are not enough if you plan even pop 13 cities.
Around population 16 is the cap that you can reasonably well use 3 full rings.

 

[Bostock]

On the other hand...

Sometimes 5-6 tiles is the only path to fresh water.
Sometimes major adjacencies gained outweigh minor adjacencies lost.
Sometimes closer spacing won't make room for another city.
Sometimes an extra tile away puts a supertile into the first or second ring.

<small>and some of us LARP a bit and thus don't chop, or "worse" yet, harvest like mad either...</small>

 

[Bibor]

On the other hand...

Sometimes 5-6 tiles is the only path to fresh water.
Sometimes major adjacencies gained outweigh minor adjacencies lost.
Sometimes closer spacing won't make room for another city.
Sometimes an extra tile away puts a supertile into the first or second ring.

<small>and some of us LARP a bit and thus don't chop, or "worse" yet, harvest like mad either...</small>

yes, exactly.

small text not working? hm...

 

[phipfwe76]

On the other hand...

Sometimes 5-6 tiles is the only path to fresh water.
Sometimes major adjacencies gained outweigh minor adjacencies lost.
Sometimes closer spacing won't make room for another city.
Sometimes an extra tile away puts a supertile into the first or second ring.

<small>and some of us LARP a bit and thus don't chop, or "worse" yet, harvest like mad either...</small>

This is the problem I've been having. I'm trying the 3-4 tiles away approach now, and the locations are usually garbage. It feels like the productionless, lack of resource tiles outweigh the benefit of the adjacency bonuses.

 

[Bibor]

This is the problem I've been having. I'm trying the 3-4 tiles away approach now, and the locations are usually garbage. It feels like the productionless, lack of resource tiles outweigh the benefit of the adjacency bonuses.

My new approach is to have 3 core cities (small map) and everything else gets just the essentials. I space these 3 out more and these cities get all the goodies, while the spacing of others is irrelevant, as long as they get at least a +3 in one key district and can grow/chop to pop 10.

 

[chazzycat]

the main reason people cram cities close together is that the incentives in this game are set up to reward having more cities than having less cities. Critically you can only have 1 of each district per city and districts are tied closely to the victory conditions. So the more cities you have the faster, the more districts you have, the faster you win.

 

[Bibor]

the main reason people cram cities close together is that the incentives in this game are set up to reward having more cities than having less cities. Critically you can only have 1 of each district per city and districts are tied closely to the victory conditions. So the more cities you have the faster, the more districts you have, the faster you win.

thats not true, though. A district’s yield is tiny compared to its buildings, which are not trivial to build.

 

[chazzycat]

ok, but the buildings face the same limitation of 1 per city so that really doesn't change anything about my point. More cities = faster victory

 

[TheFinalChiTown]

Lots of good comments here. I would suggest being less concerned about the tile distance and more about the purpose for each city.

For example, if you're going for a domination victory as Rome, here's a way to think about it. You're going to need iron for your legions. Either you settle your 2nd/3rd city where there may be iron or wait until bronze working (reveals iron) and settle near the iron. You will also need production/chops for your encampments, army, and gov plaza/warlord's throne. Lastly, don't overextend yourself so far that you get attacked before you can get your army out.

 

[phipfwe76]

Lastly, don't overextend yourself so far that you get attacked before you can get your army out.

This has actually happened to me twice in the 3 times since I bought NFP last week. I was rushing to get settlers out because I feel like I can only get maybe 4 cities before all the land is claimed by my neighbors. Once instance has Cyrus come at me on turn 80, and completely wipe me out, which was my own dumb fault knowing that's his favorite thing. Then Around turn 80 on my next playthrough I had Brazil come at me with what seemed like an impossibly large army that early into the game. As I was fending them off, Mongolia also declared surprise war and wiped out my remaining cities. The AI seems way more aggressive in NFP, and I'm struggling to see how they can simultaneously pump out so many units and settlers that quickly.