Proposals to Fix ICS in Civ3: Firaxis, please stop by...

[Ralf]

Originally posted by Admiral PJ
I like the idea of needing roads connecting land before you settle a city...

I believe this road-connected newbie-cities idea could be easy to implement as an option. After all; its not that big difference, game-mechanically speaking, if a colony must be road-connected in order to produce any output, compared with a city that must be road-connected in order to produce any output, is it? Of course, the AI must be able to adjust.

A compromize would be: Lets say you could found a city anywhere, but unless its road-connected (or harbour-connected) it wont grow beyond 1 pop.

I think this idea would solve Player3:s argument as well:
"Consider the following situation: you start a game where there are two areas rich in resources separated by a natural barrier of mountains and jungle. Many players would rush to build citys on both sides of the barrier and later link them with roads. This would be impossible with CCBs"

[Ralf]

Originally posted by Admiral PJ
I like the idea of needing roads connecting land before you settle a city...

I believe this road-connected newbie-cities idea could be easy to implement as an option. After all; its not that big difference, game-mechanically speaking, if a colony must be road-connected in order to produce any output, compared with a city that must be road-connected in order to produce any output, is it? Of course, the AI must be able to adjust/exploit these rules.

A compromize would be: Lets say you could found a city anywhere, but unless its road-connected (or harbour-connected) it wont grow beyond 1 pop.

I think this idea would solve Player3:s argument as well:
"Consider the following situation: you start a game where there are two areas rich in resources separated by a natural barrier of mountains and jungle. Many players would rush to build citys on both sides of the barrier and later link them with roads. This would be impossible with CCBs"

[Ralf]

Originally posted by Admiral PJ
I like the idea of needing roads connecting land before you settle a city...

I believe this road-connected newbie-cities idea could be easy to implement as an option. After all; its not that big difference, game-mechanically speaking, if a colony must be road-connected in order to produce any output, compared with a city that must be road-connected in order to produce any output, is it? Of course, the AI must be able to adjust/exploit these rules.

A compromize would be: Lets say you could found a city anywhere, but unless its road-connected (or harbour-connected) it wont grow beyond 1 pop.

I think this idea would solve Player3:s argument as well:

"Consider the following situation: you start a game where there are two areas rich in resources separated by a natural barrier of mountains and jungle. Many players would rush to build citys on both sides of the barrier and later link them with roads. This would be impossible with CCBs"

[Ralf]

Originally posted by Admiral PJ
I like the idea of needing roads connecting land before you settle a city...

I believe this road-connected newbie-cities idea could be easy to implement as an option. After all; its not that big difference, game-mechanically speaking, if a colony must be road-connected in order to produce any output, compared with a city that must be road-connected in order to produce any output, is it? Of course, the AI must be able to adjust/exploit these rules.
A compromize would be: Lets say you could found a city anywhere, but unless its road-connected (or harbour-connected) it wont grow beyond 1 pop.

I think this idea would solve Player3:s argument as well:

"Consider the following situation: you start a game where there are two areas rich in resources separated by a natural barrier of mountains and jungle. Many players would rush to build citys on both sides of the barrier and later link them with roads. This would be impossible with CCBs"

[Ralf]

Arrrgh! This slow server is a nightmare. MarkG - if you read this; delete above copies.

[Drake Tungsten]

Originally posted by Arrian
Simple solution: same thing w/borders. "Supreme leader, such a move will violate the sovereign territory of China, and will lead to a major international incident." 1) Oops. 2) Consequenses Shmonsequences!

This would obviously have to apply to the AI as well as the human player. Thus, borders would act as barriers, unless you are willing to invade.

This is the best idea I've heard so far in this thread. If borders don't act as a barrier to enemy units, then what is the point of borders? It sounds like there should be much tougher penalties for AI units that violate your territory.

[Howling Chip]

Originally posted by yin26
A more flexible solution like SITS mentioned would be better. But I would also add: What are colonies for? As the game stands, they are next to useless when you can so easily just send out cities.

Except when that resource is one or two hexes away from your border. Don't want a city there. I always play (so far) with culture as a prime objective. I colonize, then absorb the resource as my borders grow. Used them for remote parts of the world also, when workers are plentiful and quick/cheap to build vs a colonist.

[Howling Chip]

Originally posted by yin26
A more flexible solution like SITS mentioned would be better. But I would also add: What are colonies for? As the game stands, they are next to useless when you can so easily just send out cities.

Except when that resource is one hex away from your border or in the middle of the mountains or desert. Don't want a city there. I always play (so far) with culture as a prime objective. I colonize, then absorb the resource as my borders grow. Used them for remote parts of the world also, when workers are plentiful and quick/cheap to build vs a colonist.

[HalfLotus]

Since when is ICS a problem in civ3? More cities is not neccessarily better anymore. Corruption makes sure of that. And with the two pop for settler you no longer get a free worked square.
When you found a city, two squares are worked. Since it took two pop to found it, there's not the advantage of a one pop. settler founding a city.
Why do you guys think ICs ia a problem?

[RobC]

Originally posted by yin26
ICS = Infinite City Sleaze. Basically the idea is simple: By sending out settlers at the earliest possible moment in all your cities, you can literally grow exponentially.

This isn't ICS. ICS has to do with exploiting the free/extra production that the city square gets so that the citizen in a size-one city is more productive than a citizen working any other square in a bigger city. You should call this something else, like 'AAE - Annoying AI Expansion'

Originally posted by yin26
Frankly, I think there's a better way. I offer the following 'easy' solutions that I think will bring back a huge amount of fun and challenge.

1) "Continguous City Borders" -- CCBs
[snip]
2) "Settler Deporation"
[snip]
I honeslty believe that those two solutions working together will make Civ3 vastly superior. Any thoughts? Thanks for listening.

I think that requiring contiguous city borders would negate the need to deport settlers, wouldn't it? After all, they can't build INSIDE your borders without going to war, and if they walk through your borders to the other side, then a new city wouldn't be contiguous with their other cities.

Personally, I think this isn't really necessary. Although I agree it's annoying, you really can deal with it through war for the impatient like me, or cultural expansion for the really patient folks. In the map you provided, you could eventually grab all those little cities along the eastern coast through eventual cultural assimilation. It's not aesthetically pleasing, and it *is* annoying, but I think it should be an allowable strategy.

[Ray K]

Originally posted by libredr
I like these two ideas to prevent ICS. There are other solutions however.

There is an Open Source/Free version of Civ which is called Freeciv (http://www.freeciv.org), which is an excellent clone of civII for Win, Linux, and other OSes. They have also worked on ways to prevent ICS (they call it "smallbox") by increasing the number of unhappiness citizens when the number of cities is too important and/or they are too close. They have completely solved the ICS issue. The whole thing is explained on http://www.freeciv.org/tutorials/nopox.html. However, I like your idea to use borders.

I read this page and it seems little different from the Civ3 solution. You still have a cap on the number of cities, but you are affecting happiness instead of shields. That may be more effective at stopping ICS, but it still does not address the root problem.

THE ROOT PROBLEM:
--The city tile is always worked for free, no matter the city size. This gives a proportional advantage to smaller cities.

Related problems:

--For each point of population, the city "consumes" two food for each population point. Early in the game with no irrigated squares, the city needs that extra tile to grow.

--In addition, the starting Despotic government cannot benefit from the extra food provided by the irrigated squares.

--City tiles are automatically irrigated and laid with roads


How to solve the problem:
-- The city tile is the only tile worked in a size-1 city.
-- The city consumes only 1.5 food for each population
-- Despotic governments can benefit from irrigation
-- Force workers to irrigate and lay roads in the city tiles


What would happen:
If you founded a city on a grassland square, you would produce (after irrigation) 3 food and consume 1.5 for a net of 1.5 food per turn. In the current game, you would net either 1 or 2 depending on the second tile you worked.

With a size-2 city, you would enjoy a food surplus of either 1 or 2 depending on the second tile you work. This is about the same as the current game.

Allowing despotism to benefit from irrigation while reducing consumption to 1.5 per population increases the growth rate of larger cities, offsetting the benefit that small cities get from the smaller food box.

With a free worker at the start of the game, forcing him to work on the city tile is not that big of a deal. Also, this would place a larger premium on a good starting location for cities.


In my opinion, fixing the fundamental causes of ICS would be much better than simply setting a hard limit on the number of cities.

[yin26]

Soren: Thanks for the explanation. But in my game, I kicked the units off. Literally just a few turns later he came back in and ignored me. Could you please help us understand the conditions more clearly?

Thanks.

[Skanky Burns]

Originally posted by RobC
This isn't ICS. ICS has to do with exploiting the free/extra production that the city square gets so that the citizen in a size-one city is more productive than a citizen working any other square in a bigger city. You should call this something else, like 'AAE - Annoying AI Expansion'

Lawrence (sp?) has come up with the defacto new acronym for this situation:
REX - Rapid Early eXpansion

Other than that, i dont have much more to add to this converstion, not having the game yet (bloody infogames), but id just like to say @ Ralf... a quadriple post!! Thats a first

[TechWins]

Define "normal".

First I'll tell you what 'normal' isn't. 'Normal' isn't building a ton cities just for the sake of building a ton of cities. Continously expanding to a point where it starts to hurt your Civ is not 'normal'. Having over ten cities with the highest pop size being 2 in the before BC age is over is not 'normal'. Coming into another Civ's borders and placing a city in between all of their great culture when very well the AI should know that the city will assimilate over into the other Civs because of culture is not 'normal'. The list could go on and on but now I will tell you what normal is.

'Normal' is the appropriate manner the AI should perform. This playing manner of the AI should always vary a bit for gameplay purposes. Expand to a point where it will still benefit your empire. Take a little time to build some infastructure and let your cities grow. Trade a little in a reasonable fashion. Strengthen your military, so you don't get taken over. My point is that the AI needs to play in a way that will continously beneift their Civ. Spreading all over the map, for the sake of expansion, and not paying attention to other aspects of your empire is not beneficial. In all reality I find the AI to be much easier and boring than Civ2 because they are predictable, too spread out, and have too low of a culture always. At least in Civ2 I didn't always know that every AI Civ would be completely compiled of mega amounts of useless cities.

Is 'normal' simply sitting there with only two or three cities waiting placidly for another AI or Human player to destroy your civilization? I think not.

I would much rather the AI have two or three extremely powerful cities than them to have ten to twenty weak cities. Now that's now what the AI should do but it is still more efficient for the AI. The AI should play 'normal' (refer to my previous two paragraphs).

'Normal' would be trying to ensure the survival of your civilization by grabbing as much land, and hence resources, as soon as possible.

No, that is not the correct way to ensure survival. In the game I'm playing right now the Romans would have been able to dominate me, but instead they went to any means neccesary to expand and plopped a few cities in the middle of the empire, while neglecting growth and infastructure in all of their cities. Now those few cities have assimilated into my culture and I am much stronger than them with many 6-12 pop cities with very good culture. Where as the Romans have about the same amount of cities but with very little to no infastructure and having their cities at 1-2 pop size. If the Romans would have backed off of expansion, started to build infastructure in their cities, and let their cities grow some then they would be much stronger than, but now it is quite the contrary.

Yes, this does mean that the AI will expand like crazy, which is fine because as a player I expand as fast as possible myself.

Expand yes but don't do it in a crazy way. Do it so in an effective way (refer to 'normal').

All these 'solutions' are very narrow minded.

I fail to see your logic in this comment, sorry.

1. Have a 3 pop cost settler. Say what? So now the expansionistic phase lasts longer?

For this proposal to become effective my second proposal will have to be taken into account as well.

The only problem I have seen is that the AI settles in some of the worst possible places like the middle of a desert or in an area with all tundra, hills & mountians.

Maybe you got a different version of Civ3 than me.

[Jolard]

We cannot have the contiguous solution (only allowed a new city if it is contiguous or close to another). This makes no sense historically (Australia, America anyone) or gameplay wise. Why should you be forced to settle a desert when there is good land on the other side? Think of Australia, you have Perth way out on the other side of the continent, pretty much alone. Or think of the American West, with huge areas that were unpopulated while the West Coast was populated. Gameplay wise this would restrict your options, and the argument so many have is that they feel restricted to having to expand rapidly now!

As far as the units setting a city down in the middle of my civ, I have had this happen in every game, and it is never a problem. Every single time, I just wait long enough and it is absorbed into my civ. Not a problem. If you are impatient, take it quickly and sue for peace. But please don't restrict me in where I can settle. The whole distance/corruption equation already makes that difficult enough, you don't want it to be even worse.

I love this game, and I think a big part of the problem is that it is just different, and different strategies are needed. I loved and was good at civ II, but my first game of civ III was harsh and eye opening, I was dead before 200 AD. My second was on the large world map, and was a long hard slog, where I only won through points, (I consider that a minor victory, major would have been one of the victory conditions). My third game is almost over, random map, medium size, and it has been wonderful! Changing the settlement options is not an option, as they work well now.

However, that said, I can see the value of possible having more of an ability to expel foreign settlers. Unless they have a right of passage, they should have to leave or declare war on me if I ask. I think that is reasonable.

[Comrade Tribune]

Originally posted by korn469
ICS isn't the same thing as expansion, ICS was basically a way to exploit the game mechanics in order to achieve an advantage, in the same way a player would station a bomber over a stack of units to keep them from getting killed

ICS existed because of the following reasons

*by building a settler 1 pop essentially became 2 pop with the free settler
this is fixed in Civ3

*a large number of size one cities supported far more units than a few large cities with equal pop
this is partially fixed in Civ3

*size one cities grew exponentially faster than really large cities
this has pretty much been taken care of in civ3

civ3 has done a fairly good job of breaking the mechanics that made ICS work but what's left is a situation where it's either expand or die, and you can say the exact same thing about starcraft...plus larger cities are more valuable in civ3 than what they were in civ2, so exceptions to the expand or die rule will exist

Korn, I totally agree with this post: 'Expand or Die' is not as bad as ICS.

Yin: Your solution is too restrictive. Furthermore, I don´t quite get this: What´s the problem with enslaving his settlers? You get 2 workers, he gets nothing: Great deal? Then threaten his cities, make peace with him, if he sends settlers again, enslave them again!

[Meneer Popken]

I skipped over some parts of this thread, so I hope this hasn't been mentioned yet, but my ideas about this problem are:

Borders should be respected always. You only cross borders to attack or if you have agreed that it is allowed to cross them, this would make the game more realistic and would solve the problem of the AI building cities in enclosed areas of another civ's empire. Bad side might be that it would be possible to cordon off an area and stop all growth of an enemy, but the AI is handy enough with boats, so that shouldn't be too much of a problem.

Second thing is to make borders of high culture cities not retract when someone else builds a city next to the other city. Now an enemy gets all 8 surrounding squares if they lie in 3rd or beyond culture range of a city. It's not realistic and allows for stealing resources etc without causing a war.

[Comrade Tribune]

Originally posted by yin26
It's not that I can't beat the AI that does these things. I'm killing India in my game, for example. It's just that it's not very satisfying. I was bored after the first 30 minutes playing land grab and 'watch the AI walk through my territory.'

Yin, think of it as 'Illegal Immigration'. That problem has continued to persist to this day.

[Comrade Tribune]

Originally posted by Skanky Burns
Yin, you need a new acronym to describe this, since it isnt ICS.

What about 'Infinite Expansion Madness', IEM ? (I am not yet fully convinced that it is a problem, though.)

[Comrade Tribune]

Originally posted by Talenn
Even something so simple as making it so that a city must have a pop of at least 4-5(?) before it can build a settler would cure many of the problems. This would cause those little worthlessly placed cities to be effectively dead-ends and would slow up early spreading in favor of building up some cities, which would in turn, encourage some real cultural development to keep those cities functioning properly.

At any rate restricting where you can build limits a lot of strategies IMO and takes a lot out of the decision making in the early game. I'd much rather see a system where it was far harder to colonize instead so that you have to prioritize your locations more and thus, you'd have less worthless placeholder cities.

This seems the best solution. Simple and elegant.

[yin26]

TechWins has highlighted the issue nicely: I predict that in a few weeks time, more and more players will begin to complain that AI's 'expand at all costs' is cranked up too high and is actually at some point HURTING the AI's empire.

Yes, yes, yes. I know that I can easily swallow up his cities. The less thoughtful people in this thread think I am arguing that those cities somehow hurt ME. They don't! They hurt the AI by being senselessly eaten up by my culture expansion.

I want the AI expansion to curb back at a more reasonable moment to focus on its own infrastructure.

Sure, it works when the AI has lots of land with which to work. It should STOP doing this stuff once it has to resort to plopping down size 2 cities between mountains in the middle of your empire.

This gets the AI nowhere, people. Consider that point. The expansion needs to be more refined and the player should still have more consistent control over how trespassers get delt with.

[player3]

As I said before, I consider the AI's wanton expansion into my territory to be a clever strategy at disrupting my own expansion. It doesn't necessarily harm me per se, but it doesn't make things any easier either. The AI actually positioned itself within my territory so as to sever my trade route on one occasion (much to my dismay), so it can be effective. And the worst that can happen is that they lose one worthless city to culture influence. No big loss really, since settlers are cheap to build resource-wise.

With regards to discouraging ICS, I think a more effective approach would be to give players more incentive to favor large citys over numerous smaller ones. You could give bonus resources, commerce and culture to citys based on population, for instance. There are plenty of possibilities without having to severely restrict expansion.

[Bosh]

"AI expansion is great! Awesome, in fact! But the AI will expand at the expense of his own empire and common sense regarding borders."
"Sure, I could. But that just means the expansion AI is forcing me into wars every game. It shouldn't have to play that way."

Hmmm Yin you seem to be contradicting yourself is the strategy self-destructive for the AI, or so effective that it forces you to emulate it?
Ray K really hits the nail on the head, the solution isn't to impose artificial solutions like population caps, rules about where to place cities etc. (many of which would do just as much to hurt us bloodthirsty conquerers, such as the current corruption in Civ 3). What needs to be done is make cities in piss-poor locations be unable to grow until quite late in the game (at least without huge investments in workers) and generally be just a drain on resources. Having cities grow slower without granaries would also help since it would slow things down a bit to have all of the expansionists have to pause a bit to build granaries.

[yin26]

What I mean is that the AI needs just a bit more tweaking as regards to where it sets its cities. If, for example, it can slip behind me and start grabbing great tracts of land ... well, that makes sense, especially if the AI can't expand backwards because it is next to an ocean or something.

Those kinds of expansions are awesome and a great lift to the game IMO. Yes, I do the same in that situation.

But: Even if we argue that settlers 'don't cost all that much,' we have to agree that yes, in fact, settlers and cities that are destined to do nothing but get swallowed up are a bad idea for the AI and a bad idea for the player. That same settler either A) put in a better position or B) left to help grow its orginal city would produce better results.

Now, I understand that for both the AI and the human player, this is always a luck and risk factor. That is cool. And I'm not saying the AI should never try to found a city unless it has some magical 100% assurance of success.

But there clearly are a number of times that the AI cities could never get him anywhere. I suppose this is a bit of "Throw enough at the wall, and some of it will stick." It's a solution of raw force that will have increasingly dimished returns as players learn to counter it better.

The one way to stop ICS is ask Firaxis to recode the AI not to build on Tundra at all, can only build in Jungle after clearing, and on desert tiles only if 6 of the 8 surrounding tiles are grasslands or plains. Another is stop the trespassing unless right of passage is given.

[player3]

Yin, I'll grant that the AI doesn't seem to have any strategy behind infesting your terrritory with worthless citys. As you noted it's a sound strategy for the AI to take advantage of neglected land within your borders, but there are other times where their intrusion seems to serve no purpose other than disrupting your expansion.

Still, there seems to be some intelligence behind it. In the aformentioned situation where an intruding AI city cut off my supply route (I was exporting incense to Greece) I was forced to build an alternate route through mountains and jungle to re-establish trade, only to learn that the intruding AI began exporting THEIR incense to Greece as soon as they cut me off!!! I was both pissed and delighted (in that the AI was clever enough to think of something so sleazy). It almost seemed like a HUMAN strategy and not the product of a befuddled AI.

Now, I'm skeptical as to whether that was really their intention all along or the whole situation was just a coincidence of the AI dropping a city in the right place at the right time. But you have to wonder...

But as to whether the AI benefits from building worthless citys in your territory; in the scenario above they definately benifetted, but this is rare. I think if they've already filled their territory then it only makes sense to colonize whatever open space they can find, even if it only serves to disrupt YOUR empire.

[yin26]

But as to whether the AI benefits from building worthless citys in your territory; in the scenario above they definately benifetted, but this is rare. I think if they've already filled their territory then it only makes sense to colonize whatever open space they can find, even if it only serves to disrupt YOUR empire.

Well, I'm almost certain it was just lucky. Again, throw enough and something will stick. But I take issue with the second part of the quote: Actually, if their territory is filled and if there is no reasonably good territory to grab they should INVEST IN INFRASTRUCTURE! [sorry, I'm not shouting at you ].

Again, this is particularly so since they don't disrupt your empire in any significant way with those cities but they DO disrupte themselves. Even in the spectacular case where your trade route was severed (something I'm sure you'll simply watch out for next game), I bet you could easily have built roads or just waited until you absorbed his city with your superior culture.

Basically, I think there is a 'sweet spot' that the AI needs to hit, and I feel as it stands the AI was turned to 'maniacal' on the expansion setting perhaps to hit us all hard as we first learn the game, in which case it is somewhat disruptive.

Two months from now, though, I think we'll be begging for a smater AI, and one of the most important things we'll analyze then is better placement of AI cities and its know when to stop expanding and invest all the saved resources in its culture and army.

[player3]

I noticed that the 'AI not spending enough on infrastructure' issue has been brought up by other people, and I agree that their citys are often woefully undeveloped. Then again, MY citys don't develop as easily as they did in Civ2, either. This still doesn't explain the 'senseless' expansion that we've all noticed though, so my assumption is that this was intended to disrupt player expansion. I honestly don't think Firaxis would have missed something like this, and it must be intentional (then again, I've been wrong before about erronius 'features' in SMAC)

One other thing I've noticed is that the AI likes to keep spare settlers in case a space suddenly opens up. I remember one instance where I raized an enemy city, and the very next turn a settler (escorted by a spearman) from another nearby civ comes running in and plops down a city right where the old one was! Now that's some aggressive expansion!

[yin26]

Well, it's a fact that game developers who don't invest in an open beta/alpha pretty much have to try to guess how the general public will adapt to a given AI strategy. Actually, even an open beta/alpha won't always tell you.

So I think Firaxis is looking at how the public is playing since we are the 'real world' environment for which the game was intended. It is only my humble opinion that some tweaking will have to be done to this AI expansion code.

Keep settlers on stand-by just to claim an empty square 20 some odd turns later isn't quite efficient. It's certainly an interesting plan and can have payoffs under the right conditions, of course.

I will wait now and see how public opinion develops on AI expansion and AI efficiency.

[player3]

On a side note I think the AI in general is quite intelligent. After reading your account of your first game and the 'British Invasion' it sounds like you would agree.

Naturally some tweaking might be needed on the expansion issue as well as others (corruption perhaps) but like you said it's going to be several weeks before any of us figure out the AI completely.