City Spacing

[Artifex1]

How much overlap do you feel is OK between cities?

 

[Promethian]

All the overlap. The big factor is industrial districts. The factory and power plant buildings give their production to every city within 6 tiles of the district and it stacks. This adds up a lot. If you don't have at least one cluster of cities 4 or more strong in which all their industrial districts hit the entire cluster you are simply doing it wrong.

 

[Roller123]

i never overlap, due mostly to how messy it becomes to micro when its filled up. otherwise overlapping serves the exact same purpose it was serving in all the previous games, i.e sacrificing long-term benefits to get a short-term effect of more cities/terran. 6tile bonuses, meh, one Factory projects like 8 industry, that's 2 mine tiles, if your sacrificing 10 tiles for overlap that's a clear loss. You are also guaranteed a 2, maybe 3 cities coverage without overlap. 6tile bonus is not worth it. But like said overlapping is done to get the empire strong quick and early, that's where the game is namely decided. Specifically in Civ6 unused tiles between cities can actually be useful to fill up with farms of forest for more food and appeal to neighbouring tiles. So it's not even a question of overlap but it might even be useful to leave some space even.

 

[kampori]

I try and space at least 6 tiles away, for factory/zoo overlaps. I've had plenty of cities only 4 tiles away that were fantastic cities too. All pop went into specialists and all growth came from trade routes. Very sustainable.
6 is good though. Still get IZ overlap, can overlap huge swaths of farmland for incredible food growth, with +8-+10 farm tiles.

That's not to say cities miles away can't work either. Once they get their own IZ then they work just fine.

So, anything works really

 

[stiiknafuulia]

I've found that 4 tiles tends to be more practical than 3, before factories or aqueducts are available. Fresh water is a big factor early on, and often means that you have to deviate from the optimal 3-tile distance (it *is* optimal once factories are available, no question about it, as each factory will be able to hit 12(!) cities ). More than 4 should only be done in special circumstances (grabbing a harbor spot, making a canal city, etc).

 

[ZHONN]

It is always situational, but generally the more the better. There is no intrinsic advantage to big cities. Smaller cities grow quicker and require less amenities and housing. And as it stands, certain districts are better than others, so with more of the small cities you get more of the better districts.

 

[RealAntithesis]

I think area of effect industrial zones (and other districts) was a bad idea. It's driving a certain empire layout that doesn't really make sense. City location should be driven by the terrain, availability of resources with the goal of getting large megapolis', rather than squishing everything together in industrial slums just to get aoe benefits that don't really make sense in the real world. Cities are suppose to be more or less self contained.

 

[Roller123]

as each factory will be able to hit 12(!) cities ).

this is wrong by definition. I'm not even going to ask how you got 12, but the perimeter cities will only cover half, and since perimeter cities are the most numerous, the empire will be a bunch of junk cities used to power the core, which is nice i guess, but while you were busy building settlers, someone else has been building military instead. I don't think people realize that an ICS approach has just the same weaknesses as non-overlapping cities -- it is weak in the beginning. And i'm not even going to ask as well where did all that space for all these cities came from, unless it's a huge map vs AI, which, considering the AI isn't expanding beyond ~6 cities anyway, is quite redundant to play.

 

[stiiknafuulia]

this is wrong by definition. I'm not even going to ask how you got 12, but the perimeter cities will only cover half, and since perimeter cities are the most numerous, the empire will be a bunch of junk cities used to power the core, which is nice i guess, but while you were busy building settlers, someone else has been building military instead. I don't think people realize that an ICS approach has just the same weaknesses as non-overlapping cities -- it is weak in the beginning. And i'm not even going to ask as well where did all that space for all these cities came from, unless it's a huge map vs AI, which, considering the AI isn't expanding beyond ~6 cities anyway, is quite redundant to play.

I saw a graphic about it (on Reddit iirc) that showed one factory touching 12 city centers. I only took a passing glance at it though, so it may be possible that it was flawed. Anyway, the denser the pattern, the more cities you'll be hitting, so since production is so important in the game as it stands, it makes sense to try and maximize it. Hence, the optimal distance from the pov of production alone is 3 tiles. Obviously you cannot hit the max amount in a real game, due to mountains, coasts, fresh water considerations, etc. But it is the *ideal* to strive for, and the closer you get, the better.

As for military, in the beginning your cities will be small enough that production isn't really affected by their close distance, as every city will still have enough tiles to work without significant overlap. Later on (yet before factories) this effect is noticeable, but should be compensated by the sheer number of cities, provided you start building the troops early enough. It is possible to screw this up and be steamrolled, but with the bad AI, it is mostly a threat in multiplayer or on Deity vs Gilgamesh (etc).

There is usually plenty of space to expand to, unless playing on an island map, a crowded map (with extra AIs) or having an unlucky spawn with close neighbors. In the latter case, might as well start early with racking up the denouncements (I view them as a badge of honor these days ), and make them relinquish their inappropriate aspirations to greatness.

I will concede that one 'super city' with a good amount of tiles is useful to have, for military emergencies, building wonders and especially spaceship projects, which are prohibitively expensive for your main 'hive' of small-to-medium sized cities. A desert city that built the Petra is ideal for this purpose, if you can get it ofc.

 

[Xyriach]

Seeing the full city grid at placement (similar to the great person area of effect) would be handy for this. If you're going to overlap, it would be useful to see your settlers total potential zone.

It's workable and not exactly a problem, but it would be handy.

 

[Aenthin]

Sometimes it's situational. I would space 3-4 tiles if I get Japan, for example. If I'm planning on building on more improvements, like Ziggurats for Sumeria, I might consider 4-6 tiles away.

 

[LDiCesare]

this is wrong by definition. I'm not even going to ask how you got 12, but the perimeter cities will only cover half, and since perimeter cities are the most numerous, the empire will be a bunch of junk cities used to power the core, which is nice i guess, but while you were busy building settlers, someone else has been building military instead. I don't think people realize that an ICS approach has just the same weaknesses as non-overlapping cities -- it is weak in the beginning. And i'm not even going to ask as well where did all that space for all these cities came from, unless it's a huge map vs AI, which, considering the AI isn't expanding beyond ~6 cities anyway, is quite redundant to play.

How is this wrong?
Half the settlers you don't build yourself but steal from the AI to begin with. I suspect it's even easier on Deity since the ai has more settlers. I've even seen city-state settlers.
It doesn't matter that cities are junk. My last game, the best tech researcher had zero campus but a lot of cities. ICS works fine because your last (junk) cities produce settlers while the rest produce whatever you like (military, factories...).
The overlap is all the more interesting if you build some of the stuff that increases range (some great persons do that too).

 

[Roller123]

How is this wrong?

i was referring to every factory covering 12 cities being an impossible statement, because perimeter cities cover half the cities, and if perimeter cities would cover 12, then the center would cover 24. See what i'm saying? In any case this number is only possible(and i'm assuming that statement as true) with select, not every, city, which have completed a wonder or consumed a great person, extending the range of the factories. If you just normally take 4 generic non-overlapping cities in a rhombus, and put their factories to the center, they easily cover themselves. So you can have 4 cities with 23 workable tiles and 4 factories overlap without any prereqs. Have a cake and eat it too.

 

[joncnunn]

Highly situational to me; mostly based on river and other fresh water sources locations; but to me, both "tight city placement" (one tile beyond minimum allowed city placement) and "loose city placement" (two tiles beyond minimum allowed city placement / one tile closer than no overlap) feel better in most situations than either "ICS placement" (as close as the game allows) and "OCP" ("optimum city placement" / no overlap but no underlap either)

 

[Zyxpsilon]

Two words; Specialized Sub-Urbia.

For me, any core of the first 3-4 cities should concentrate on pop & productivity while anything else around that stable central area *CAN* use the proximity trick to overlap major bonuses while touching a few if need be... but that is situational & involves various other conditions such as the Victory type & a whole bunch of supplemental concerns.
Ideally -- you want optimal rather than swift (but still limited) power at the expense of TOO rapid expansion in some cases.

Again.. simply because of the newest systems of Districts & Wonders built ON surrounding city tiles.. terrain (luxury and strategic resources included) comes at a very high premium in this game, more so than ever.