Is ICS the only way?

[Hans Lemurson]

I just started looking back into Civ3 when a friend of mine whose computer can't run Civ4 well told me about a mod he was making.

Now, that aside, is there any reason NOT to do an ICS strategy? When I first read about ICS (years ago) I was appalled, since I was the type of player who wanted to have all my cities be perfect 20-square deals, and this is the antithesis of that. Looking back now several years later and with much more Civ experience under my belt I take a more pragmatic approach, but building a massive uniform grid of ultra-dense cities still makes me feel somehow hollow inside, since none of those cities really matter.

So bottom line, I don't really like ICS, but given the mechanics of the game (unit support being based on number of cities, etc.) it seems like the optimal strategy. But still a part of me still cries "Say it ain't so!"

A second question(though not particularly suited to this board):
Is there any mod which would cause there to be some sort of factor that would make ICS less ideal in favor of large "efficient" cities? What would that mechanic be? What limitations are there in implementing such a thing? (Civ3 looks pretty rigid in terms of the things you can modify int it)

 

[Spoonwood]

I believe you've confused civ II with civ III. In civ II ICS almost always came as the way to go. In civ III, due to things like settler cost and corruption, ICS only makes for the optimal choice in certain situations (like a 100k game).

 

[Aabraxan]

Well, first of all, when you refer to ICS, I think of CxC spacing, with C being a city, and x being an open tile. For many games, ICS makes sense, but only after a certain point. I wouldn't ICS the core, because I want those cities to have more room to grow. Once I get out into the hinterlands, which I use as specialist farms, I will ICS things pretty hard.

On the other hand, if you really detest ICS empires, you could shoot for a variant, like an OCC or 5CC. In an OCC, obviously, you won't be ICSing anything. In a 5CC, I would think, you'd want to spread your cities out so that they each get a full 21 tiles to work, and good ones at that. As I haven't played many OCCs or 5CCs, take my advice on these at your own peril.

 

[Hans Lemurson]

Thanks for the replies, I guess I'll need to do some more thorough reading of Civ3 strategies and the mechanics behind corruption.

So you're saying that I should do the standard "optimal" development for cities near my capitol, but then later just fill-in the gaps in my empire and sub-prime outskirts with a bunch of little filler-cities?

Now for the hypothetical part: What alteration in the game-mechanics do you think would make it not quite so optimal to spam a bunch of little cities? Civ4 used city-maintenance (such that new cities were often net losses for a while), but I'm thinking one thing to do would be if military support were changed to be not on a "per-city" model, or to favor larger cities, but I don't really know.

I guess I've got a lot of reading to do before I come back to this.

 

[Aabraxan]

I'm afraid that I'm clueless on editing/modification questions, so I'd rather leave those alone than give you bad info.

As for empire layout, I think of my cities as being in 3 categories: (a) Core; (b) Semi-core; and (c) farms. Core cities are spaced widely enough that each should be able to work 12 tiles (or more if I want some metros). Core cities also get lots of improvements. Semi-core are spaced a little tighter and get fewer improvements, but they'll still get some. These first two categories build offensive & defensive units. The farms get laid out CxC, no improvements. Hire as many specialists as you can and build workers, settlers, and artillery in them.

 

[Spoonwood]

There exist threads on this topic in the archives somewhere. There's also an article on city spacing schemes in general. I'd consider terrain before you consider much else. Settle on hills generally, select carefully whether you want to use coasts or not, settle on rivers and lakes, and make sure that every square within your cultural borders can get used at some point.

 

[Theov]

Like the horse on a chess board. Cxx
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxC

 

[Hans Lemurson]

The various layout patterns to achieve particular results I already know, I just don't really like densely-packed cities. I suppose though that I could settle dense initially, and then as my "core" cities grow and begin to compete for tiles I could simply "prune" the filler cities and send their population to more useful places.

 

[Theryman]

Now for the hypothetical part: What alteration in the game-mechanics do you think would make it not quite so optimal to spam a bunch of little cities? Civ4 used city-maintenance (such that new cities were often net losses for a while), but I'm thinking one thing to do would be if military support were changed to be not on a "per-city" model, or to favor larger cities, but I don't really know.

I guess I've got a lot of reading to do before I come back to this.

Seems to me like you have a couple of options: first, you could definitely cut the corruption rate significantly. This would allow more cities to be useful, meaning you do not start the ICS sprawl until a good bit further out of the core.

Another option would be to change the unit support rate. In republic, for example, I believe you get one free unit for a town, 2 for a city and 3 for a metro. You could make it so you get NONE for a town, and something like 3 for a city and 5 for a metro. And maybe jack up the price of units from 2 gold per turn to 3 gold per turn. You would have to do serious balancing with the other governments too, though. This would encourage you to space the cities a bit further apart to try and get them up to cities. While they would still be corrupt beyond use, they can still grow!

 

[timerover51]

The various layout patterns to achieve particular results I already know, I just don't really like densely-packed cities. I suppose though that I could settle dense initially, and then as my "core" cities grow and begin to compete for tiles I could simply "prune" the filler cities and send their population to more useful places.

I intensely dislike densely packed cities too, and do not do it. What you do for a mod is boost the resource yields, yields of all of the resources, and boost the effects of mines and irrigation. All that makes a large, efficient city produce units are a rate of one per turn, at most two, and reduces the effects of corruption. The AI is still going to pack the cities in, and then hits the optimal city limit faster, boosting the corruption there. You do have to do this on a continents or archipelago map. I have a couple of mods that I use to keep corruption down on me.

 

[Spoonwood]

Timerover51,

Do you space your cities wider than the AIs?

 

[splunge the 2nd]

My 2 cents- I use CxC spacing in the corrupt hinterlands but always take terain into account. To get a working farm I need food so irrigated and railed grassland is wonderful as is desert if I'm playing an agricultural civ. I'll still do CxC in plains but mountains, hills, small islands and tundra teh only benefit from lots of cities is unit support as specialists starve! I'll use looser spacing and maybe even build a cheap temple if I'm looking for territory in a domination win or to control a wanted resource (eg. CxxxxC in tundra to control oil)

 

[timerover51]

Timerover51,

Do you space your cities wider than the AIs?

The closest I will ever put a city is CxxxxC, where neither city overlaps the other, and that is only when the resources justify it. Otherwise, I spread my cities out, and like to have undeveloped land around them. I think that the most cities I have ever built directly, not counting ones taken from the AI is about 25. With my boosted resources, each city is normally a unit per turn producer. I have also increased the number of citizens in a city to 20, and a town to 8, so as to fill all of the production tiles.

Even playing the standard game, I do not crowd cities together.

Edit: I do not like to see cities packed together, nor the landscape a dreary succession of roads, railroads, and mines. I specifically modify the game to avoid having to do that.

 

[Hans Lemurson]

I suppose that doing "less than optimal" play by not building myriads of cities is just akin to raising the difficulty level by assigning aestheticly based restrictions to myself. I do not seek Demi-god world conquest.

 

[timerover51]

I suppose that doing "less than optimal" play by not building myriads of cities is just akin to raising the difficulty level by assigning aestheticly based restrictions to myself. I do not seek Demi-god world conquest.

That essentially is correct. The only difference between the lower and higher levels is the AI trading rate, the bonus in combatting barbarians, and the number of optimal cities. By going with what would be considered by some as "less than optimal" play effectively allows the AI to expand at the lower levels as fast as the higher levels with would some view as "optimal play." My response it to modify the game to even things out.

 

[Spoonwood]

The AIs have a faster build rate at higher levels also. I hardly see a wider spacing than CxxxxC coming out as "less than optimal" as an opinion. There exist plenty of tiles that never do or can get used, the more distant cities experience more distance corruption, workers have to travel further to start working tiles, commerce and production start up later and come in weaker than with a tighter spacing, military needs to travel farther to defend cities, you need plenty of culture (really both a temple and a library to get to 1000 culture) to close off your borders, etc. The only possible advantage comes as that the AIs have less territory, but that often doesn't matter too much. It comes as difficult to imagine any sort of game where such ultra-wide spacing has a real advantage over CxxxxC, CxxxC, CxxC, or really even CxC spacing. It seems closer to a fact that wider than CxxxxC comes out as "less than optimal". By all means play as you like, but please realize there exists more than mere opinion here. Note: I don't mean to say that one of CxxxxC, CxxxC, CxxC, or CxC necessarily has better spacing than another here, but that any of those comes as better than wider than CxxxxC in almost all cases for any sort of objective yardstick such as victory finish date or score.

 

[CKS]

Of course, the yardstick that really matters is subjective - that is, what is the most fun.

 

[Bowsling]

Now for the hypothetical part: What alteration in the game-mechanics do you think would make it not quite so optimal to spam a bunch of little cities? Civ4 used city-maintenance (such that new cities were often net losses for a while), but I'm thinking one thing to do would be if military support were changed to be not on a "per-city" model, or to favor larger cities, but I don't really know.

In the editor, it is quite simple to change unit support numbers under each government tab to greatly favor cities and metropolises, simply by making unit support something like 2, 6, 10.