A Guide to Settling Cities

[Stalker0]

Thought I might take a crack at a strategy guide for once, and would love others take on this topic, as I'm sure there are lots of tips and tricks. This is a guide about settling new cities.

Part 1: Why build a city?

When planning to build a city, the first key question to ask is...should I build a new city? Even for a player going a wide strategy, its important to understand what a new city will gain you....and what it won't.

What you Lose: Science, Culture, Tourism, Happiness:Past your first 4 cities (sometimes even past your first 3), the penalty to Science and Culture per new city (12%) basically ensures that your newer cities will NOT be positive science or culture generators.

First of all, the second you build that new city, the penalty begins. You have to build up a lot of science and culture infrastructure to match your older cities. All of that time spent building means you are losing science and culture. And realistically you are never going to match your capital, unless its been ignored.

But even if you managed to do it, you still lose out on the bonuses from Great Scientists, Great Writers, trade routes, bonuses from pop growth, etc. These static values are more useful to a TALL civ than a WIDE one, even if a WIDE one can generate them a little more frequently.

For happiness, especially early in the game, you are going to lose 2-3 happiness from that new city in the majority of cases. So unless your happiness is +12 or better, you are losing more percentages to all yields in your empire, and it weakens your Golden Ages.

For tourism, early on you can get tourism through the extra arenas you could build, but the "smaller civ" multiplier to tourism ultimately gets you more in the long run, so you don't get more cities for tourism.

What you Gain: Hammers, Faith, and "Maybe" gold: Religion is a big civs game, and you can get a lot more faith with more cities, which also strengths their internal religious pressure.

Gold is a maybe. On the one hand, you can in theory get more gold from new cities, especially if they give you luxury resources you can trade, and you go progress which gets you a lot more gold. However, if you spend a lot of money on new building maintenance, or investing in new buildings to help those cities grow, new cities will often be money neutral, or even negative! So normally you gain gold, but you have to consider how much of that gold simply goes into city investment vs other uses.

However, the thing you really get....and ultimately the real reason to go Wide.....is hammers, hammers, hammers. There is no TALL substitute for hammers, more cities will always net you more hammers than a small one, and that is ultimately the key yield to utilize with a big empire.


So....here's why you build a new city: Now knowing what we know, here are some of the key reasons to build a new city:

a) Strategic Resources. Iron and Horses for most of the game is often one of the key factors in a new city. These are key to your military, and they provide more of those hammers (a key reason for new cities!)

b) Multiple Luxury Resources / Monopoly. One luxury by itself is not enough, even if you don't have that luxury. The happiness you gain is a wash with the new unhappiness you occur, and now you have more penalties to other yields.

If you get multiple luxuries, than that's a solid happiness contributor, or can be flipped for a good amount of GPT.

Monopolies are also a good thing to look for with new cities, but keep in mind what you are getting. A bonus to science or culture is often wasted due to the new penalty, a bonus to happiness may not be worth the new unhappiness plus the extra penalties. If your thinking about a corporation, its often better to plan a late game settle or conquest to get it, then to get stuck with an unproductive city most of the game just to snag a monopoly.

c) Key Territory: This is rare, but there are times when controlling a narrow strip or land or a thin waterway gives you a major military advantage, and forces others to either go a very far away around or deal with your city directly. This is a rare consideration...the city also needs to have good productivity, as decreasing the value of your entire empire is not usually worth it, even if it is a good chokepoint

d) Military Unit / Diplomatic Unit Production Center: Barring the other reasons, this is the normal reason to set up a new city. Hammers are your advantage with more cities, and this what you use them for. Build up a strong army for conquest or for defense....or get diplomatic buildings and then build diplomatic units to secure CS allies. While TALL civs often gain as much from a CS as you do....they simply cannot keep up with the hammers need to crank out those units like you can if you set up a few cities to focus on it.

Domination and Diplomatic Victories are generally easier for WIDE civs, so factor that in when you get a new city.


Part 2: Selecting a Site (Coming Soon!)

 

[jma22tb]

Something to add for Great People on tall / wide:

Tall is a boost for Policies and while you don't get more GP Points per turn, your Culture Guilds will lead to faster policy gains and Wonder access than Wide because the Policy costs are fixed over time, while your culture per turn will grow. Because you can only have six writers, artists, and musicians from guilds and never have more, there's no point expanding past the amount that can handle that number if you're looking to win Wonder spam and to get to Ideology first. Notable exception being the Celts, who get an additional musician per city from Celidh Hall.

Wide is a boost for Scientists, Engineers, Civil Servants, and Merchants because the more cities you get, the more points per turn you can gain from those buildings and slots. You can only have three writers, artists, and musicians from guilds, but as many of these specialists as you can get.

 

[Strigvir]

First, Science and Culture penalty is dependent off the map size.
Second, your math is wrong. Penalties get noticeable around 30-50 cities (depending on the map size).
You get roughly +100% increment in science output per new city with 6-16% increment in costs.
So five cities will result in 500% science output versus 180% costs in the worst case scenario. 50 cities result in 5k% Science output versus 800% costs. Even then it only get noticeable because the stage you can support that amount without breaking down is near the end of the game and later cities don't have time to build up to normal Science production.
Culture is a different matter, since its production per city heavily depends on the playstyle and generally is lower than Science output. So the decision between Tall and Wide comes down to Culture output.

 

[Stalker0]

First, Science and Culture penalty is dependent off the map size.
Second, your math is wrong. Penalties get noticeable around 30-50 cities (depending on the map size).

Your first point is completely valid, I don't get to play large maps very often (my computer is too crappy) so this point is valid.

However, I will argue your second point. A few notes:

1) In order for a new city to contribute positively, it has to build all of its infrastructure, which takes a significant amount of time. If I build all the science buildings, I have not built the culture buildings yet, and vice versa. It also takes time to get all of the population you need to build specialists (which are the strongest source of science and culture.) All of those turns spent getting your city up to speed means turns that you are losing culture and science.

2) The capital in most cases gets significant bonuses other cities do not get, so satellite cities never get as high a bonus as the capital. So a 4 city civ does not get 300% more science than a 1 city civ, just because the capital generates significantly more.

3) Vox Populi relies a lot more on immediate bonuses to science and culture than vanilla does (such as the university or council bump, GS science, culture bonuses for expended GP, etc).

These bonuses do not scale with number of cities. That means a 1 city civ gets as much bonus from a GS bulb as a 50 city civ does....yet the 50 city civ has a much higher science penalty.


All of these contribute to the fact that WIDE civs just don't keep up in science and culture like a TALL civ does, but they gain superior bonus to faith, hammers, and gold.

 

[tu_79]

A semantic note:

I think the terms tall and wide come from how populated a city becomes. Tall evokes a skyscraper, while wide evokes a suburb. In gameplay, building tall means leaving enough space between cities so they can work all the tiles, and this way, getting superpopulated cities. Building wide means settling at a shorter distance, overlapping the workable tiles.
The advantage of tall is evident, more available tiles in the city allows for the biggest cities possible. But you don't get to work all those tiles until the very late game, while a wide layout allows for faster hammer production in the early game.

Brief.
Tall = Bigger and more distant cities
Wide = Smaller and closer cities

Then you get to settle an map area. If the area you can call your own is smaller than most other civs, you have a SMALL empire, or a LARGE empire if the opposite is true. Typically you can fit more cities in the same land if you settled wide. We have a misconception that building wide equals to more cities, but we could as well settle wide and stay small with only 3-4 cities, or we could settle tall and have many cities to get a very large empire.

The last thing to consider is the expansion rate. It varies from OCC (no expansion at all) to very aggressive. Even if I want to settle a small empire of just 4 cities, I'd say I expand very aggressively if I set my capital to produce 3 settlers on row, without any building or unit. Those cities are going to be undefended and underdeveloped, and probably affected by unhappiness, but have a better chance to get good spots. I think it is suicidal, though.

 

[merill]

Second, your math is wrong. Penalties get noticeable around 30-50 cities (depending on the map size).
You get roughly +100% increment in science output per new city with 6-16% increment in costs.
So five cities will result in 500% science output versus 180% costs in the worst case scenario. 50 cities result in 5k% Science output versus 800% costs. Even then it only get noticeable because the stage you can support that amount without breaking down is near the end of the game and later cities don't have time to build up to normal Science production.

And i think your math is also wrong
The penalty is additive, not multiplicative.

iNumCitiesMod = iNumCitiesMod * GET_PLAYER(ePlayer).GetMaxEffectiveCities(/*bIncludePuppets*/ true);
iResearchCost = iResearchCost * (100 + iNumCitiesMod) / 100;

with a penalty of 10% per city:
1 city: cost 110% (+10%)
5 cities: cost 150% (+50%)
10 cities: cost 200% (+100%)
50 cities: cost 600% (+500%)


edit:

These bonuses do not scale with number of cities. That means a 1 city civ gets as much bonus from a GS bulb as a 50 city civ does....yet the 50 city civ has a much higher science penalty.

scientist bulb is a % of your "current" science output => wide/tall neutral.
scientist plant (academy) is a flat value => tall advantage.
source:

// Beakers boost based on previous turns
int iPreviousTurnsToCount = m_pUnitInfo->GetBaseBeakersTurnsToCount();
iValue = pPlayer->GetScienceYieldFromPreviousTurns(GC.getGame().getGameTurn(), iPreviousTurnsToCount);

 

[Txurce]

Most importantly, this is fertile territory for a guide. Settling cities is something most players do on auto-pilot, regardless of how many factors they process. It's good to review the numbers behind the intuition.

And it's also good to argue about the numbers.

I was struck by the implication that Authority and Progress are the way to go for Domination, that progress is also optimal for Diplomatic Victory, and that Tradition is best for the other two. Those are pretty major, definitive statements, and they warp how one looks at the number of cities to be settled.

 

[tu_79]

Most importantly, this is fertile territory for a guide. Settling cities is something most players do on auto-pilot, regardless of how many factors they process. It's good to review the numbers behind the intuition.

And it's also good to argue about the numbers.

I was struck by the implication that Authority and Progress are the way to go for Domination, that progress is also optimal for Diplomatic Victory, and that Tradition is best for the other two. Those are pretty major, definitive statements, and they warp how one looks at the number of cities to be settled.

I don't see why you couldn't begin with Tradition, and go domination later. Tradition has its best in GP generation, but it also has a fast start and a shot to some early wonders. Later, its ability to grow population faster can help to recover faster those cities you are going to conquer. I'm not completely sure, but I'd say that GP generation is per city based, so the more specialists in cities the more GP are obtained (the Capital gives more GP thanks to the specialist slots in the Palace). The bonus to food is good to produce GP in every city. And, with the exception of the GE, every GP strength comes from global yields, so they scale with number of cities too. Perhaps they don't get money and science from connections, but they can plant a Town and an Academy in every city, while saving some worker turns. They are great too in tundra and desert, with its +1 food in GP Tile Improvement, making those tiles workable since the beggining.
Just take Imperialism later, when your army is ready, and some piety/religious beliefs to counter the unhappiness.

 

[merill]

I'm not completely sure, but I'd say that GP generation is per city based, so the more specialists in cities the more GP are obtained (the Capital gives more GP thanks to the specialist slots in the Palace).

Yes it's city-based. But the cost is empire-wide. You can have a city that work a scientist and never ever create a GS because the cap is inflating too quickly (useless GS points).

 

[tu_79]

Yes it's city-based. But the cost is empire-wide. You can have a city that work a scientist and never ever create a GS because the cap is inflating too quickly (useless GS points).

Ok. You still get more GP with more cities anyway. And if you pick the religious belief that grants yields for every GP spent, you are getting tons of everything when going Tradition, even with the extra cost. Not saying this is ideal to conquer everything in sight, just that it isn't so bad even if it doesn't scale so well with number of cities.

 

[Strigvir]

The cost also increases with each non-puppet.

 

[Lifariz]

Sorry to necro this post but I can't help but add some more opinions and questions to this topic since it's incomplete.

First, how far should your new cities from your capital? How far should they be from your other cities? Should you expand towards or away from other civ's? And I'm pretty sure it's going to be dependent on social policies. In my opinion 4-5 tiles depending on the resources the roads are going to cross over would be great. Tile improvement chaining is an issue I don't understand well though.

Second, should we build roads/rails towards the tile we plan to settle on before your settle is even made? Assuming we have the gold and other resources to.

Thirdly, should we raze down enemy cities if we feel like we could place a city in a better position?

 

[tu_79]

Sorry to necro this post but I can't help but add some more opinions and questions to this topic since it's incomplete.

First, how far should your new cities from your capital? How far should they be from your other cities? Should you expand towards or away from other civ's? And I'm pretty sure it's going to be dependent on social policies. In my opinion 4-5 tiles depending on the resources the roads are going to cross over would be great. Tile improvement chaining is an issue I don't understand well though.

Second, should we build roads/rails towards the tile we plan to settle on before your settle is even made? Assuming we have the gold and other resources to.

Thirdly, should we raze down enemy cities if we feel like we could place a city in a better position?

1. Totally dependent on policies, you're right.
1a. For Tradition, I'd settle my first two cities at 3 tiles from capital. The reason is that a Tradition capital with Hanging Gardens uses very few tiles, and the first cities are going to be populated soon, and you can connect them for an early gold rush. The following cities can be settled at 4-5 tiles apart. Your strength is population, Tradition cities will have many specialists, but not so many as the capital. You probably want to stop at 5-6 cities, then develop and expand only trough conquest, when your neighbors reach your borders (beware unit limit cap).
1b. For Progress, you want as many cities as posible, so you're better forward settling. This is: sending your first settled cities near the neighbors borders, then fill the void. Your strength is in buildings, so settling each 3 tiles is not too bad an idea, but 4 tiles is ok. It depends on the land resources. Try to settle 10 cities, then defend, develop and you're ready to follow as you please.
1c For Authority, rush a big army, and settle every 4-5 tiles. You're going to be working a lot of tiles (happiness issues), so be sure they are improved.

2. Don't build roads in Ancient Era, except for tactical reasons. Progress has a policy that let you connect your cities faster (+2 gold for every connected city), but the others need 5-6 pop cities before they can be connected with profit. Roads are a good tactical asset, though, so if have spare gold, use them to reach faster to sensible points, spread your religion and expand traders range. The benefits for connecting a 1 pop city (except the mentioned Progress policy) are minimal.

3. Is the city you captured smaller than 5 pop and very underdeveloped? Raze it. But a well developed and populated city is good enough to keep even in a bad location. Other reason to raze cities is to lessen unhappiness for heavy warmongers.

 

[southblade]

There's another reason to build cities: you could be more close to an enemy capital. I tend to play warmonger and go for domination no matter wich open policy I choose. If no bias because of civ, map or situation like chance of Stonehenge sometimes I choose my policy according to my UU, a good early UU allows you to expand fast and protect you before you complete the branch, so progress could be ok (persian expand rush). If your UU comes in medieval you can go autorithy and timing all (mongol skirmishers total war). If your UU comes really late, you can go tradition and expand through opportunistic wars.

 

[SaberSix]

I am at tutorial turn 72 with 3 settlements of 6, 7, & 8. I think I recall there were things to buy/build or things I would lose if/when I convert them to cities. I have 1.000 gold. Can anyone refresh my hazy oldstimer's memory?