City placement strategy

[szemek77]

When playing vanilla and then BtS I always tried to place my cities very carefully. First of all, I tried to use all possible resources (which is obvious), avoid tundra and desert withing city cross and keep the distance between them, so the city will use all of tiles.

RoM changed the circumstances, but I'm still mentally locked to these rules. After some games on 2.71 I'm ready to modify the city placement strategy anyway. First of all, RoM offers the possibility to work almost all tiles (even mountains in some modmods). Secondly (more important) the techs and improvements gives you so much food that you don't need to use all the cross to make the city big, great and prosperous.

That gives you the possibility to place more cities in small area, which is good for two reasons: shorter distance makes the maintenance costs smaller, and secondly - it means better communication, and better defense in a result (it is easier to move your units within your borders).

Does 2.8 change anything? What is your strategy in city placement?

 

[Afforess]

You know we have a strategy subforum, correct? Please use that in the future for strategy questions.

As to your question, I weight my city placement on two major factors:

• Resources that I need
• Strategic locations

If I need stone, and there is a spot for it, I'll place a city next to it. Likewise, If I know one or two strategically placed cities will cut off the AI's growth and development, I'll place cities their too. If neither applies, coastal or river spots are really good.

 

[szemek77]

You know we have a strategy subforum, correct? Please use that in the future for strategy questions.

I'm sorry, You are right.

As to your question, I weight my city placement on two major factors:

• Resources that I need
• Strategic locations

If I need stone, and there is a spot for it, I'll place a city next to it. Likewise, If I know one or two strategically placed cities will cut off the AI's growth and development, I'll place cities their too. If neither applies, coastal or river spots are really good.

I try to do the same. Lately, I tried to cut AI off, but then I found out that Hathepsut (and she is a woman, so I will not say what I would like to say about this move ) sent a settler on a boat just to place a city thousands of kilometers from her capital just under my belly.

Anyway I just wanted to say that RoM makes the city placement not as important as it was i.e. in Vanilla, where you had a lot of non-workable tiles, little resources and bad placement could make the city to stay as a poor village for ever.

 

[RobO]

I usually try to make sure that the tiles they can work don't overlap, but it never works out that way in practice due to other constraints (resources, non-workable tiles, coastal or river placement, defense considerations). And with 2.8 the cities grow much slower (which is good), so there's an incentive to space cities closer.

 

[Grunthex]

As RobO pointed out above, the food growth is much slower in 2.8. On Marathon (or Snail I imagine!), this means my cities are always growing until at least the mid-industrial age... I just can't reach my happy/health caps before then!

Because the cities stay smaller longer, there's more temptation than ever to overlap cities. Sometimes the pure builder in me cringes, but I point out that those 7 squares the new city gets are probably enough to get me a size 15-20 city come modern ages!

 

[szemek77]

Very good point!

This change (slower city growth) has a fundamental meaning for city placement. It seems that in 2.8 the cities don't need to use their full cross. As you guys mentioned, my hand is also still shaking when I need to place next city which would overlap the previous one, but it seems now that it become wise to conduct much more tight city placement.

 

[MisterBenn]

I agree completely with not worrying about all city tiles and to scoop up as many resources and you want and need.

I also try to collect as many flood plains squares as possible at the start of the game, there's a huge difference to your starting growth if you can bag a couple of these and get cottages on them. A grassland river square with farm or cottage brings 3 food (not enough to push more growth) and coasts with lighthouse also 3, whereas the right type of floodplain brings 5 (or 6 with Agricultural). It's a huge difference to your early growth and worker / settler production and you can keep up with Emperor AI players right from the start if you do this (or outpace them from the start if you have the agricultural trait too). Sometimes you can use that +10 or +15 food surplus for troops with Despotism to great effect too!

Another thought regarding the very start of the game, it seemed a little odd to me that at the start of the game a farm improvement brings +1 food and a cottage brings +1 food, +1 gold. I never build farms at the start, I cottage all food squares for the commerce and then consider replacing with farms later on once the technology increases their food output. I would have thought a dedicated farm at the start of the game should produce more food than a cottage... has a +2 food per farm at the start been considered? You'd then have a growth or commerce decision to take with your food squares in the earliest turns.

 

[szemek77]

Another thought regarding the very start of the game, it seemed a little odd to me that at the start of the game a farm improvement brings +1 food and a cottage brings +1 food, +1 gold. I never build farms at the start, I cottage all food squares for the commerce and then consider replacing with farms later on once the technology increases their food output. I would have thought a dedicated farm at the start of the game should produce more food than a cottage... has a +2 food per farm at the start been considered? You'd then have a growth or commerce decision to take with your food squares in the earliest turns.

I've noticed that too. In result I stopped to build farms at the beginning of the game, especially since I prohibited myself to build caravans (cheat, cheat, cheat) so I need to pay special attention to my economy.

 

[DRJ]

The caravan now evened out as I mentioned earlier I don't know what speed you play but if the gain in snail is 170 it's in lower speeds lower, and btw. why don't you build caravans if the AI still does?

In history, several civilizations even had a whole economy depending on caravaning, like the Kingdom of Saba or the powerful Samarkand at the silk route or Timbuktu in the sahara...

 

[zappara]

I would have thought a dedicated farm at the start of the game should produce more food than a cottage... has a +2 food per farm at the start been considered? You'd then have a growth or commerce decision to take with your food squares in the earliest turns.

Caste System civic...

 

[MisterBenn]

That's certainly true once you have a couple of cities and techs, but for your first city site that's not available yet. I was thinking more in terms of mitigating the luck element of your start position. The ability to build under forests in Rise of Mankind was a fantastic decision, but getting your early food is still a bit of a lottery sometimes. If you are unlucky not to have food resources or flood plains in your first city cross then you are at an immediate growth disadvantage even if you have access to fresh water farms (and coasts with lighthouse). On the higher skill levels I'll generally regenerate the map unless there are clearly good spots waiting for further cities.

On another note, seeing Trade Caravans being discussed again, I always thought it was a little strange that trade caravans paid out lump sums once they are deployed (being a derivative of the great merchant). How about an alternative implementation?

When I deploy a TC in a foreign city, how about a building is spawned in the foreign city which is "English trade caravan" (where I am the English player). It would grant a bonus to the foreign city (perhaps 5% of the city's gold output) and also deliver the same bonus gold to the person who sent it, via the same mechanism as the privateer gold plundering. After a certain number of turns the building will expire. Sending further trade caravans merely replace the building and reset its timer.

I like a few things about this, i.e. trade caravans benefit both the sender and the recipient (so pick who you trade with), the income for both people is a stream rather than a lump sum (mor realistic, easier to balance) and it seems less exploitable (sending countless TCs to the same city won't yield extra gold). Thoughts?

 

[RobO]

On spacing cities: After playing into the 1900's I can see that there are several buildings providing food. This gives you an advantage when you place cities close together, as you can build more of these buildings with more cities and thus have more citizens and more specialists.

 

[DRJ]

That's certainly true once you have a couple of cities and techs, but for your first city site that's not available yet. I was thinking more in terms of mitigating the luck element of your start position. The ability to build under forests in Rise of Mankind was a fantastic decision, but getting your early food is still a bit of a lottery sometimes. If you are unlucky not to have food resources or flood plains in your first city cross then you are at an immediate growth disadvantage even if you have access to fresh water farms (and coasts with lighthouse). On the higher skill levels I'll generally regenerate the map unless there are clearly good spots waiting for further cities.

On another note, seeing Trade Caravans being discussed again, I always thought it was a little strange that trade caravans paid out lump sums once they are deployed (being a derivative of the great merchant). How about an alternative implementation?

When I deploy a TC in a foreign city, how about a building is spawned in the foreign city which is "English trade caravan" (where I am the English player). It would grant a bonus to the foreign city (perhaps 5% of the city's gold output) and also deliver the same bonus gold to the person who sent it, via the same mechanism as the privateer gold plundering. After a certain number of turns the building will expire. Sending further trade caravans merely replace the building and reset its timer.

I like a few things about this, i.e. trade caravans benefit both the sender and the recipient (so pick who you trade with), the income for both people is a stream rather than a lump sum (mor realistic, easier to balance) and it seems less exploitable (sending countless TCs to the same city won't yield extra gold). Thoughts?

Interesting thoughts.

I quoted your ideas in the mod-mod-request-thread and replied to it here (post #405).

This can become something great!

 

[Bezhukov]

I tend to space cities closely in the early game, then optimally later when they can grow quickly, as your initial cities will have used up their wonder allotment already and don't need the extra production that a full-fatcross can produce.

 

[civ_king]

I caravans, useful for solving money issues, but my favorite use is building them in my city to transfer production to new cities, god for speeding up that second-third-etc city

I tend to space cities closely in the early game, then optimally later when they can grow quickly, as your initial cities will have used up their wonder allotment already and don't need the extra production that a full-fatcross can produce.

with industry/mines you don't have production issues, mines for hills, else industry

 

[szemek77]

That's certainly true once you have a couple of cities and techs, but for your first city site that's not available yet. I was thinking more in terms of mitigating the luck element of your start position. The ability to build under forests in Rise of Mankind was a fantastic decision, but getting your early food is still a bit of a lottery sometimes. If you are unlucky not to have food resources or flood plains in your first city cross then you are at an immediate growth disadvantage even if you have access to fresh water farms (and coasts with lighthouse). On the higher skill levels I'll generally regenerate the map unless there are clearly good spots waiting for further cities.

The problem is that it is hard to tell when to regenerate a map (just leaving the question "if" on a side). I saw starting location which seemed to be bad, but after discovering few technologies it appeared that they were great (5-6 different resources around the capital). But I got also games, when I was happy to get i.e. two corns close to my first city, and then - nothing more. Three months ago I played a game (it was 2.71 I guess) in which I had at least 5 much better cities than my capital.

 

[mussbu]

I'm just wondering if there's any consequences to building cities closer together (3-4 squares away) much later in the game. Does a city eventually need the bigger space in later periods.....?

If not, then:
1) mores cities in less space, then
2) civs are more compacted, then
3) more civs could fit on a smaller map, then
4) we can start more AI civs at beginning of a game. Maybe 10 -12 on a Large map instead of the standard 8 civs.?

DOES THE AI KNOW HOW TO SETTLE CITIES IN CLOSER PROXIMITIES? If so, then this is doable.

 

[Afforess]

DOES THE AI KNOW HOW TO SETTLE CITIES IN CLOSER PROXIMITIES? If so, then this is doable.

In The bottom half of this thread, I note that, at least in my modmod's, the AI is looking at and knows that city radius's are 3 apart, instead of 2. Now, I can change that in the SDK, and I also believe there is a global define for it too.

The AI doesn't really "know" anything. It's really a conglomeration of functions that determine which action the code says is best for it to do at that particular moment. ATM, it isn't always as "wise" as it could be.

 

[mussbu]

In The bottom half of this thread, I note that, at least in my modmod's, the AI is looking at and knows that city radius's are 3 apart, instead of 2. Now, I can change that in the SDK, and I also believe there is a global define for it too.

The AI doesn't really "know" anything. It's really a conglomeration of functions that determine which action the code says is best for it to do at that particular moment. ATM, it isn't always as "wise" as it could be.

Most excellent! Is there a way to code in the 'wisdom'. We can start taking up a special collection for better AI wisdom. Silly Friday!

 

[CiverDan1]

Farming at the start of the game is rather pointless, as both the farm and cottage are +1 food. Another thing is if you are not near a river and do have a food resource your city is simply not going to grow early in the game.