How dense to pack cities?

SeanTM

Chieftain
Joined
Nov 22, 2017
Messages
5
I've been playing Civ since the very beginning and have built in biases based on previous versions of the game. One of these is for placement of cities. In Civ VI I place my cities so that they overlap 1, 2 or 3 hexes with long term growth in mind. The AI places them MUCH closer together and seems not to worry about territory being contiguous. The result is that I have this beautifully laid out civ with 3 or 4 cities by turn 50 or so, then the next 50 turns the AI starts plopping their own cities wherever they can find a couple spare hexes.

Should I be packing them in tighter like the AI? Should I be concerned with territory being contiguous?
 
You generally want to pack cities tightly as a city doesn't really need to be larger than 5-7 pop to work well; buildings give flat yields so you just want as many of them as possible without getting crushed by maintenance. Another advantage to many ciites is the overlapping city bombard zones that can stop invaders more easily

In general, if the land looks sorta usable, and isn't just desert/tundra/ocean/mountain you should settle it, and in general if a city fits in the vicinity in your empire, you probably should.

If you build Colosseum, a wonder that gives amenities and culture to all cities within 6 tiles, any city that fits that range should be settled.
 
One good use for the Settler lens is to put your cities so that the AI can't stick anything inbetween. So I try to put them so that everything is red. I don't mind some overlap in the 3-rd ring, as generally your cities aren't going to work all their tiles.

I'd maybe let the AI do the work for me (settle the best spots and then conquer whatever they stick in free spaces), but they tend to choose the most awful spot when forward settling :D
 
Welcome to fanatics @SeanTM
If you want the best value from a coliseum or a factory then packing close is good for that purpose.
If you want to stop the AI settling between your cities then packing close is good.
My normal reason for packing close is I will have one city with great tiles 3 away and if I settle a city there I can swap the tiles with my high pop city without buying them for high gold prices... If I have to spend 200 gold on a tile and 3 gold is worth 1 production for purchasing buildings (which it roughly is) then have just spent 66 production on a tile that gives 1-2 extra production a turn. Of course the tile may be iron or niter or a nice luxury like elephants on jungle hills. A lux you should be able to sell for 6 GPT so only take 7 turns to get your return and with the extra production/food it was worth buying rather than settling for.
If playing a warlike game I will place cities where they are best suited (my current default activity). If the AI wants to settle between them it saved me a settler.
If I am playing a peaceful game I will clump cities more although I will often forward settle my second city to stop the enemy taking all my expansion space. Forward settling will definitely entice war but playing peacefully you still need an army and its doing nothing much and fending off a neighbour without taking cities often means you can get great rewards from a later peace deal, especially if you manage to pillage their lands. If you forward settle and promise not to do it again it takes about 45 turns before you get a positive modifier with that civ for keeping a promise so I consider forward settling sensible peaceful activity as long as you promise not to do it again. 8 tiles is the trigger point for settling too close.
 
Overlaping tiles for me is a huge mess since i play long games. -.- In my opinion they should have made it so that cities start with normal oval shape and 6 tiles around it but can expand where ever they want up to 36 tiles.Example. You have 5 tiles from your city center wheat so you buy tile after tile to get it.And the first 36 tiles you get/buy are the ones that will be only workable. The problem is that the cities are oval shaped and because of that overlap. With a game mechanic like this cities could be near each other,share effects but not overlap.
 
The problem is that the cities are oval shaped and because of that overlap.

Workable tiles are 3 so you can set them up so the do not overlap... they are hex shaped not oval. If you extend it to 5 they also do not overlap.... of course when you take terrain into consideration is more complex but fundamentally neither oval nor do not have to overlap

In reality they do not often get to 36 without some good culture buildings but I guess with your long game they do... but do they get to 36 population?

Sorry for the bad diagram
upload_2017-11-26_15-34-39.png
 
Overlaping tiles for me is a huge mess since i play long games. -.- In my opinion they should have made it so that cities start with normal oval shape and 6 tiles around it but can expand where ever they want up to 36 tiles.Example. You have 5 tiles from your city center wheat so you buy tile after tile to get it.And the first 36 tiles you get/buy are the ones that will be only workable. The problem is that the cities are oval shaped and because of that overlap. With a game mechanic like this cities could be near each other,share effects but not overlap.

Even in a marathon game, how many of your cities use all their tiles? My answer is 'none' :D
 
As a general rule I hate placing my cities too close together. Maybe it's about the aesthetics of it or older civ instincts, or just the need to keep district costs and amenities in check. That doesn't mean that you should avoid overlap. The outer ring is always fair game. Beyond that, it comes down to available terrain and resources and what you can get out of each city. Adjacnecy bonuses from districts are always nice, not just for Japan.
Bonuses from the Collosseum and entertainment and industrial buildings should apply to at least four cities.
But as I said, it all depends on available terrain. Even a city that will hardly contribute to your economy on its own is worth it if it gives you access to a new luxury that will boost your four largest cities.
 
Even in a marathon game, how many of your cities use all their tiles? My answer is 'none' :D
And my answer is that in my epic games, it is TOO MANY!
I prefer 5 or 6 vacant tiles between cities. Not all of them will be desirable to work, many of my cities will grow into the mid/high-20's or beyond by turn 600 or so, with the Colosseum (and particularly if I can get suzerenity of Toronto) amenities and war weariness is only a minor concern.
 
I avoid/minimize overlap in cities that are clearly going to become population/production powerhouses in the late game. Usually that includes the capital. For any other city I try to overlap 3rd and sometimes even 2nd rings - there's a huge advantage in being able to transfer high production hexes from one city to another whenever you like - a feature I'm sure many people don't even know about. And then there's the regional area of effect buildings.

As a general rule I hate placing my cities too close together. Maybe it's about the aesthetics of it or older civ instincts

At least in Civ III, the most optimal way to play involved jam-packing cities extremely close together. I never did it though, looked really unfun.
 
At least in Civ III, the most optimal way to play involved jam-packing cities extremely close together. I never did it though, looked really unfun.

Yeah, call it stubbornness but I often play Civ games the way I think they should be balanced. Even if it's suboptimal, I want each city to have character, and the district system really amplifies this obsession.
 
I hate the way the AI crams cities into every available space. I use a mod that restricts the minimum distance between cities to 5. It works wonders.
 
This is certainly a thought provoking question!

Unfortunately, Civ5 has trained many veterans that you want big, developed cities that "pull their weight;" cities with only a few population should be avoided because of the mechanics of Civ5's happiness system. In Civ6, however, there is no per city penalty, and amenities effectively become a "maximum population limit." But you even get the first 2 pop in a city amenity free.

**The trade off of all this is that you have to sink more hammers and population into settlers, which is not always easy to do, even with the card.

Currently, with everything being flat bonuses, every instance of a district is what matters for several yields (gold, faith, science, somewhat production) and so more cities=more instances of those districts. Not to mention trade routes.
With most yields, the aggregate sum is what matters: science, culture, faith, tourism, gold, great person points, and so on. This naturally tends to wide empires, but not necessarily packing.
production is a little different in that there is a time based aspect to it; you want to be able to produce certain things fast enough to be competitive. But this usually comes into play for building wonders; being terrain bounded, this often has a secondary effect of making select cities the ones that want the high production as opposed to all of them. A group of "Non-core" cities all producing "commercial hub investment" is unbelievably powerful. Getting density with constrained territory, incentivizes packing. As does the Coliseum, which is so good that you should almost always put as many cities around it as you can. (note that you cannot swap a city's first ring of tiles to another, excepting already built districts; you don't want to accidentally mess up a powerhouse city!) As a side note, any city within range of the coliseum can have 8 population for free. Denser cities also means more value from other regional effects.

Further, city state envoy bonuses also put many districts on steroids.

With all that in mind, I think it whether to try to pack 'em in or not also has a huge civilization choice component. I think Australia, because of Outback Station adjacency, does better when they have some breathing room between cities; there's less 'loss' of tiles being eaten up by districts.

On the flip side, there are 4 civs which I think have phenomenal bonuses towards packing:

Japan

Their district adjacency ability is much stronger when they have more districts close together, which is exactly why you would pack 'em in.
England
The ability to have 2 trade routes per city is huge. A very strong strategy for marginal coastal cities is to make a triangle of city center-commercial hub-harbor; this assures 3 gold for the harbor, which via the harbor adjacency card guarantees 6 production from a shipyard. (This works for any civ; japan gets extra yield out of this. Plus you get nice gold from the harbor) England's two trade routes are important because you can put them anywhere; you can actually support your core with your "colonies" unlike other bonuses that stay local.
Aztecs
Builder-rushing districts means even snow outposts can get a district up in 5 turns once you have serfdom. Monty's leader ability means you get extra amenities from luxuries, helping support all this settling.

These other civs are really not in the same league as Germany, though.
Germany
IMO, they have the best dense city game out there. This stems from 2 factors:
Their Civ ability to support an extra district means they can place a Hansa+CH the turn they settle, and lock in the price on both.
The Hansa's adjacency bonus with CH applies to each CH. This means you can get really crazy. See illustration below for classic templates:

upload_2017-11-26_18-56-44.png


I will assume the craftsmen (+100% IZ adj. bonus) card is being run.

The diamond is the classic 2 city configuration. Each Hansa will produce at least 10 production, thanks to both CH's plus being adjacent to two districts. You can almost always pull this off, even if cities are not very close.
The triangle also guarantees 10 production per Hansa, but requires only 3 cities instead of 4 (for two diamonds.)
The ring is handy in case you have a resource you can put in the center. In this case, that would imply 12 production per Hansa.
The crescent is a variant on the ring. This allows one Hansa to get an extra CH adj. (+4 prod,) while the others are unchanged. This is useful if there's a resource in the way.
The 4 city cluster is much more efficient than 2 diamonds. The three core Hansas yield 14 production minimum; this bumps up to 16 if you place another district next to one of the core hansas (or the city center works for this too.) The fourth hansa can go wherever the resources make the most sense. (If you're really good, you can line up another set of cities so that it drops into the center of a Ring layout, granting a nice +18.) You can also protect all those hansas from saboteurs with just one spy in the middle, which is nice.

Obviously, these shapes can be rotated and translated to fit the terrain and fit where resources are. Usually you can configure one or two resources for each Hansa, and often another point of district adjacencies; this can push yields up to 20+, PER CITY. Since these clusters go between cities, the land around is leftover for working normally.

Consider that a normal IZ is considered "good" if it can achieve 6 production. Six. Those are rookie numbers!
Contrast with 10-20 production; that's several mined hills worth of production, without having to work those tiles at all. You also aren't constrained by tiles that can be mined, so it's even easier to set up Ruhr valley in a good spot. There's no need to have a dense blanket of packed in cities either, just clusters of them around your empire.

And, of course, the Hansa is half price. Now that's German Efficiency!
 
To answer your question, what are you seeking for in the game and how long do you expect your games to last?

Are you seeking an immersive experience or just trying to win the game as fast as possible? Do you like your cities or are they just a means to and end? Do you want more Wonders and Great People?

If you are just concerned about speed then ideally you would not be concerned if cities overlap because each city will never need to grow large so they'll never work beyond the 2nd ring. The proximity of cities also means you get the benefits of area of affect buildings from the Industrial Zone and Entertainment Districts. You would want to avoid emotional attachments to anything in the game and avoid doing anything that hinders speed even though it may be enjoyable. That means no favorite cities, no "turtling", no peace play and avoiding most wonders and great people like the plague.

The downside of this that the maximum potential of your cities will be severely hindered and that will become an issue if your opponent manages to drag out the game because small cities don't scale well with time. Also, small cities can't build wonders well or complete city projects quickly because they lack the yield potential of larger cities. Then again, people only concerned about speed don't see most Wonders or Great People as significant so that's usually irrelevant.

To sum up, if you want a Civilization with unlimited potential that grows ever stronger to stand the test of time; If you desire Wonders, Great People and all else that adds to the glory of your Civilization, then space out your core cities 6 tiles from each other so each one has space to grow to its fullest potential.

This is the true reason for the "just one more turn syndrome" actually. It's really just a matter of which scales better with time.
 
Last edited:
The ability to have 2 trade routes per city is huge. A very strong strategy for marginal coastal cities is to make a triangle of city center-commercial hub-harbor; this assures 3 gold for the harbor, which via the harbor adjacency card guarantees 6 production from a shipyard.
Nice write up... but for England not accurate enough, you cannot say 3 gold and leave it at that and the clustering is not required.

The triangles are great but do not require clustering cities. The benefit gained is too small.... it's half a gold per district adjacent.

The triangle bonus is not well described especially for England.
The commercial hub will get +5 gold if the triangle includes a river
The harbour is+5 if it is off continent and +1 for every adjacent sea resource. With a shipyard and adjacency the harbour is giving 10-14 gold and 10-14 production per turn while the commercial hub provides 5 gold.

An off continent rivermouth with 2 bonus resources I find hard to turn down with a half price harbour. But playing an efficient game I will still mostly bypass and get a Campus there first... in place of the CH.

An on continent triangle with 0 sea resources nets 8 gold and 2 trade routes before buildings and at half price harbour is quite strong still.

Playing an SP efficiency game you just cannot beat the agressive small city campus chop strategy as far as I can see.
 
Last edited:
I hate the way the AI crams cities into every available space. I use a mod that restricts the minimum distance between cities to 5. It works wonders.
I should probably get this mod, because I don't like building too close because I like to have room for districts and farms, but I also hate when the AI walks a Settler in and plops it down in that one valid tile exactly 4 tiles from your four cities. :mad:
 
I should probably get this mod, because I don't like building too close because I like to have room for districts and farms, but I also hate when the AI walks a Settler in and plops it down in that one valid tile exactly 4 tiles from your four cities. :mad:

Yep, that got so irritating recently that I ragequit due to this... I wonder why do the devs insist on programming the AI to do it though. I would understand if it was their 5-th city or something, but they'll often pop their second city in the crappy spot next to my capital even with plenty of nice places on their end of the continent...
 
Yep, that got so irritating recently that I ragequit due to this... I wonder why do the devs insist on programming the AI to do it though. I would understand if it was their 5-th city or something, but they'll often pop their second city in the crappy spot next to my capital even with plenty of nice places on their end of the continent...
To be fair, I sometimes do this to the AI knowing that a) they're likely to take the land otherwise, and b) I can settle all the nice places on my end of the continent later. Only if there is actually a nice spot near them of course, I wouldn't plonk a shoddy city down right next to them just to be annoying. Well, unless maybe I want them to DoW me instead of the other way round. :smug:
 
Back
Top Bottom