The ultimate ICS thread: analysis and solutions

[korn469]

The ultimate ICS thread: analysis and solutions

1. What is ICS?

ICS also know as infinite city sprawl, is a strategy in civ games where a person seeks to create an endless amount of cities radiating out from a central hub. these cities generally lack the overall infrastructure of a so called perfectionist strategy where cities are spread out and are highly developed. the perfectionist strategy calls for maximizing each city's production. ICS is about optimizing your civ's overall production.

2. Why is ICS a problem?

ICS is a problem for a multitude of reasons, but the biggest problems arise from the current civ models for city production, unit support, the happiness model, and the growth model. i will address each model one at a time

city production:

the civ model for city production is base squares worked equal number of workers per city plus one multiplied by the total number of cities
or
(W+1)xC where w=workers per city and C=total number of cities

additionally each city is a production center

for example we have two civs: the green civ is a perfectionist civ with one size 10 city, the yellow civ is an ICS with ten size 1 cities

the yellow civ started with one colony pod and spent 270 minerals building nine more, each of the yellow civ's city have no infrastructure...

the green civ's city has the following alpha centauri infrastructure...a recyling tank, a recreation commons, a hologram theater, a tree farm, and one 1-1-1 police units costing 260 minerals...

all citizens of both civs are workers, and all citizens are working a forest square; there are no special resources or economy or industry bonuses

the yellow civ's total output is:
base square: 20-10-11 (+1 for HQ)
workers: 10-20-10
total: 30-30-21
surplus: 20-30-21 (minus people eating)

the green civ's total output is
base square: 3-2-3 (+1 for HQ)
workers: 20-20-10
total: 23-22-13
surplus: 13-22-6 (minus people eating and maint.)

additionally the yellow civ has ten production centers while the green civ only has one

unit support:

using the same stats for the yellow and green civs, if both civs have zero support ratings in the social engineering table, gives us

the yellow civ can support 20 units for free
(2x the number of cities)

the green civ can support 2 units for free
(2x the number of cities)

happiness model

using the same stats for the yellow and the green civs we get the following happiness scores

on librarian on a standard map at zero effic.

the green civ has a total of 10 workers seven of them are drones before base facilities and garrison units
four are taken care of by base facilities
three are taken care of by the police garrison unit (assuming the green civ has a police rating of +3)

the yellow civ has no drones

the growth model

with adequate food and a pop boom the green civ could only grow one citizens per turn, while under the same conditions the yellow civ could grow ten citizens per turn

also it cost ten food per worker to grow a base by one citizens so at size one the yellow civ's bases would only need one tenth of the food to increase by one citizen compared to the green civ

korn469

[korn469]

Solutions and analysis:

this is only the first installment of proposed solutions, and only a brief analysis for now...i intend to further explore causes, effects, and solutions for ICS plus add in more analysis

Re: city production:

although i am unsure of what to do about the (W+1)xC problem, i do have one proposed fix for this...advanced cities after building some facility should be able to produce more than one unit per turn...in Alpha Centauri i would suggest that each factory facility (genejack, robotic, nanoreplicator, quantum converter...) should add a new building slot...so if a city had all five of those facilities it could produce a maximum of six units at once

Re: Unit Support:

this is quite simple to fix. add in global unit support. give one point of support for each population unit. a side note to this is unis get more advanced and costly they should cost more support.

Re: Happiness Model:

though not as easy as a fix as unit support (which could be implemented in various ways but would still have global unit support as its crux) civs with a roughly equal population should have roughly equal drones reguardless of cities. Also, and this though not completely related to overall happiness, does have direct consequence to the problem. A city should not be able to have more specialists of any kind (i'm mostly refering to doctors/empaths/transcendi) than it has workers (until it's of a huge size like 20+), and a city should always have to have at least one worker.

This would stop people being to escape drone riots at will by simply turning workers into specialists to avoid drones. currently it is possible to ALWAYS avoid growth related drones only using specialists, without units or facilities if you are paying attention. This model would also help enfore the pacifism penalty. Currently a size one city with the only citizen turned into a doctor could support an entire army and suffer no pacifism penalty. With supply crawlers ferryingminerals back it could easily support 10 needle jets which normally would incure 20 drones due to pacifism. Also, pacifism drones like unit support should be global (working in the same way as ineffic. drones do)

Re Growth model:

this is fairly simple...do not base the growth model entirely on food surpluses

korn469

[Xin Yu]

I have proposed an idea a month ago but the thread was buried deep inside the forum somewhere. Basically I proposed an idea that, after cities reaches certain size and after certain tech is obtained, several cities can be combined into a megalopolis. Let's say 4 cities can become a megalopolis, which has 1/4 more productivity than the summation of the 4 cities. Also they share city improvements: one city wall works for all at 1/2 effect, two city walls double the wall effect, 3 redouble, etc. More important, a megalopolis can build a (very expensive) 'group unit' which has the strength of 5 units altogether--it must be killed 5 times before being eliminated. And only megalopolis can heal a group unit. I bet ICS will disappear with this setup.

[korn469]

Xin Yu,

have you thought of the implications of your idea? i feel that is not a good fix for ICS. in fact i feel that your idea would encourage ICS even more than what civ does now. merging a number of cities into a megopolis, which is better than a city, encourages a player to create even more cities. a player will want to have a great number of megopolises in the later stages of the game to feel competitive, so they build a huge number of cities now. ICS lets you grow much faster than having just a few cities. an empire with ten size one cities with adequate food will grow 100 times faster than having one size ten city. a megopolis probably needs to be composed of cities that are close together. ICS does this, perfectionism doesn't. this idea encourges ICS and offers little to stop it, however you could be onto something that i am not. if there is something i am missing tell me.

[MadWoodster]

There is one easy way to solve ICS and that is to do what CTP has done and create unhappiness when your empire gets above a certain size, unfortunately Activision forgot to have this size figure dependent on the map size and not just a standard for all maps, which meant on a small map ICS was still a problem but on a large map it is difficult to maintain ICS due to unhappiness.

To prove this by example say the size is 10 cities and you have built 12 size one cities. Each of these would then have a -2 happiness modifier in CTP, this figure could be made higher to make it more effective of course.

[The Joker]

About a month ago I posted ideas in the thread "How to portray the rise and fall of great powers in Civ3" to work out the same problem: That large civs in Civ games are always more powerful than smaller ones, and that the most powerful civ at one point in the game was more or less sure to be that at the end too. This way the game was more or less set by 1500AD, meaning that the rest of the game would only be working towards the inevitable.

I think this is an appropriate place to put these ideas back.

My ideas are not as well defined as yours. They are more overall guidelines to what should be done. I am aware that including all the ideas will be too much. I am merely putting them all in so we can find the best ones and work on them.

-Huge corruption and waste for cities far away from the capital (I think SMAC did ok here).

-Less control over cities far away. This would especcially work before the industrial revolution, and could mean that you were not in control of what was being built in a distant city, and that you might only get a percentage of it's tax revenue. You could enforce your dominion by sending troops built in your core area to the city. This could be pretty frequent for cities far away.

-I like the concept of some sort of bureaucracy points. These would be an expence on your national budget. The more bureaucracy points you have the lower the chance is for revolt, and the more control you can have over cities far away. But the more cities you have the more of these points would be required. So a large empire wouldn't just mean that the periphery cities were more or less useless. It would be a very large expence on your national budget, forcing you to raise taxes, which could create even more unhappyness plus it could reduce your research rate. This concept is far from perfect, but an advanced version of it could be implemented in Civ3, making the game more exciting. This could also help setting back very large civs in research, which could help smaller powers to take over.

-The chance of civil wars raised if you were a large civ. The chance would be even higher if you had a very polarised civ (like great difference between the individual cities in wealth etc), which would also be more likely in a large civ.

-Being a large civ would require more units around the world. If the cost of units would be both raw materials, food, credits and perhabs labour a large army could be really expensive, also in your national budget. This could cause unhappyness and less research due to raised taxes.

-CTP's State of Alert could be implemented (this is among the things I like about CTP), but better. There should be 3 or more levels of alert - low, medium and high. If set on low your units would get their attack/defence rates halfed. This would make them pretty useless. If set on medium their rates would be 3/4 of usual, and on high they would be normal. However, on high an average amount of units could cost perhabs 40-50% of your country's labour and gross income plus loads of raw materials (of cause decided by the amount of units you have)! This would make that REALLY expensive, and therefor only workable in a serious war. The major downside for large civs should be, that these would be significantly slower at moving between these states of alert. If you have 150 cities the time to go from low to high could be 20-30 turns! For a small civ with 15 cities this could be reduced to 1-2 turns. This way large civs could be not only unstabile, but unflexible, and would make it possible for a small civ to win a war against a large one before the large can set it's SoA to the high level.

-Better AI. Small civs should make alliances with each other if they were attacked by a large one.

-Cities should grow faster in modern times. This way cities made in 1800 wouldn't be useless.

-Every civilization should spend part of its research points on education, just to preserve the knowledge it has: a larger civ should always spend/pay more on education just to ensure that no knowledge disappears; if it spends too little, doesn't have enough libraries, advances/knowledge will disappear (like a substantial part of the knowledge of the Romans after the Great Migration); as it has more people in it, it needs more administrators, more priests, more lawyers, more scientists just to run the empire!

-When building units conscription of the troops and manufacturing of the weapons should be divided (units would require real people from the cities, so cities should consist of people, not heads). But would it be wise policy for the Mongols in China to arm the subjected native population? Of course not! So you would have to conscript people that were loyal to you. This would make it hard to build a large army in a civ based on conquest, as you could only build loyal units in assimilated cities. Assimilation should take a lot of time, determined by the happyness of the city and how you treat it.

-When researching advances should spread slowly across your empire (with a speed of maybe 3). Cities who didn't have an advance couldn't produce things requiring this, and it couldn't help researching an advance which required one they didn't have. This would mean that a large part of your civ could end up not helping you with research. With the discovery of advanced flight this would no longer be, as all cities would get an advance immediately.

-There should be far more advances, so many that one civ could not research them all. This would mean that you would have to have good relations with some civs, and trade advances with these. This would also mean that large isolated civs would stagnate, as they couldn't trade advances with anyone (this would solve the China-stagnation problem).

So what do you think? I hope this could finally put this major flaw in focus, so it wont return in Civ3.

[Pythagoras]

wouldnt just decreasing the amount of support a city square provides fix this?

[Xin Yu]

City square is crucial in growing. Without the city square's food, a city may stuck at size 1 forever; without the city square's science, it takes 100 turns for the first tech; without the city square's shield, one needs 10 turns to build a warrior.

[VetLegion]

Joker, great ideas!

Civ is trying to resemble real world mechanics as much as possible, and your Ideas are on the right track. Xin Yu's idea also has support in real world(Tokyo-Yokohama etc.).

The Idea about keeping the existing level of knowledge I hear first time and it is one of the best civ3 suggestions I heard!
It would force people to bild improvements...excellent! No ICS, and I agree we have to keep something like base city square. It represents local population joining the settlement.

[korn469]

Xin Yu,

i just typed for 45 minutes in responce to your megalopolis model. my internet messed up on me and lost my message. needless to say i am extreamly mad...however after running the numbers, the 40 size one cities still beats a size 40 megalopolis.

40 cities still produce more food and minerals than a size fourty megalopolis, they also produce more labs, the megalopolis produces more energy but not enough to make up for the superior minerals of the 40 cities. also the fourty cities still grow ten times faster than the megalopolis. and they still support ten times more units for free than the megalopolis does.

also if a perfectionist strategy is pursued by a player of equal skill against a player of equal skill who is pursuing an ISC strategy, then the ICS player will win, especially if megalopolis are added into the mix. an ICS player might not be able to turn all of his cities into a megalopolis but he could definantly cobble one or two together in the core of his empire with lots of little cities left over, compared to one or two megalopolis representing the entire perfectionist empire.

another fallacy of the megalopolis idea is that it doesn't stop ICS or offer any solutions for it during the earliest stages of the game when ICS is by far the strongest. by the time your megalopolis idea comes into play the person pursuing the ICS strategy is already firmly in control of the game, and can probably take advantage of megalopolis first.

going ICS doesn't mean that you purposely try to keep your cities at size one. it means early game you just keep on building colonly pods and expanding your empire like crazy...soon the first colonies you established stop building colony pods. they are the core of your empire. outside of the core you have your provinces churing out more colony pods. all of your cities are bringing in money, which you use to finance rapid infrastructure construction. so they can quickly catch the perfectionist in infrastructure. also small cities grow quickly especially with pop booms and can once again catch the perfectionist strategy. ICS gives a player the tech advantage early game to the perfectionist player. megalopolis does nothing to address this.

also is there one square on the map that represents your megalopolis? if so then a player could quickly capture your size 40 megalopolis with a sneak chop and drop attack.

so in summary, i think your idea does not stop ICS, in fact i think it encourages ICS. i also think it adds in new exploits and abuses.

korn469

[Pythagoras]

quote: Originally posted by Xin Yu on 02-24-2000 01:36 PM City square is crucial in growing. Without the city square's food, a city may stuck at size 1 forever; without the city square's science, it takes 100 turns for the first tech; without the city square's shield, one needs 10 turns to build a warrior.

Start a city at size 2, lower the food in the home square. Also as citys get bigger they get less and less from their own square
or just make it illegal to build two adjacent citys or have a limit on the amount of adjacent citys you can build, or even better, you try to build a city next to one thats reached its adjancentcy max, and it just adds to the pop of the original nearby city, like pressing b in the city.
[This message has been edited by Pythagoras (edited February 24, 2000).]

[korn469]

The Joker,

in responce to your ideas...

corruption: leaving it alone like you suggest sounds pretty good

less control in farflung territories: i do not like this idea because it clashes with gameplay...it is like the computer purposely puts your bases on govenor, that would make me want to break the game

bureaucracy points: could you give me more examples of this idea? it sounds really intriguing. i think this could be an innovative solution that solves the ICS problem

civil wars: i am an ARDENT supporter of this idea. that is eactly why i made suggestions about the proportions of workers to entertainers. if my idea is implemented it would make it harder to control "incubating cities" which are an invariable aspect of ICS...incubating cities are cities which have everybody turned into doctors until you have enough money to rush build a recreation commons.

quote: this is my workers to specialist idea. cities size 1-4 should have to have a worker for each specialist and should have to have at least one worker. cities size 5-20 should have to have a minimum of 3 workers and at least one worker for every two specialists. cities size 21+ should have to have a minimum of seven workers and that is the only restriction on them.

additionally losing your HQ should make a large number of cities revolt.

more units for a large city: i do not agree with this idea. for one thing it is easier for a large ICS civ to have and support lots of units. for another thing it would give them a large army to crush the little guys with.

State of Alert: this idea could work well, or could totally disrupt the balance of the game. I like the way it gives a small civ flexablity over a larger civ. however the high tech civs could stay on medium and never have to go high to high alert. or they could build cheap garrison units that cost little supply as the mainstay of their force and use a small high-tech multi-attack (nerve gas copters) armored spearhead to crush their enemies.

better AI: GREAT IDEA! this is one idea i always support

faster growth for current era: encourages ICS unfortunantly

losing tech: GREAT IDEA! i love this idea!

assimilation: i agree that assimilation should take a while, but i don't entirely agree with your ideas...but they could work

research spreads slowly: it's a novel idea but i think it would interfer with game play and wouldn't hurt ICS too much

more advances: it is not a solution to ICS

in summary, i think these ideas that you suggested hold the most promise to ending ICS.

1. bureaucracy points
2. civil wars
3. State of Alert
4. losing tech
5. assimilation

add in some of my ideas

global unit support
unhapiness not determined just by city
global pacifism penalties
minimum number of specialists

and one new idea

military units, except for the most basic military units, cannot be built unless a city has the appropriate facilities

and you have gone a long way to Fix ICS

please The Joker, could you discuss more about your five ideas? i feel that those ideas hold the most promise in killing ICS

also here is something to consider, in Alpha Centauri there is an assimilation period that lasts 50 turns, during that period their is increased drone riots and the city looks like the enemy you captured it from, there are also other things about it like the enemy can buy it back cheaper. now i think taking your assimilation idea and adding it to SMACs already established model. basically you couldn't build military units out of those cities for at least 25 turns (as time passes in SMAC the penalties go down) does that sound good?

korn469

[Xin Yu]

korn469:

A group unit=5 units but only costs 4 times of a single unit, and needs only 1 shield support. Moreover, if an ordinary unit is killed, it is out, but a group unit can be killed 4 times and recover as new (reduce the chance of stack-kill). The superiority of the group unit will render all units built from ordinary cities obselete. Hence, even the shield production of 40 small cities is larger than a size 40 megalopolis, they are of no use at all if building units.

I wouldn't worry about defensing a megalopolis against ordinary units. With a barracks a group unit can be fully recovered in one turn even after being killed 4 times. If I have 3 vet fortified group pikemen I should be able to hold the city against 15 crusaders.

For the food problem, a megalopolis can build a group settler unit, which costs 4 times an ordinary settler's shields, but only deduce the entire population by 1 and eat 1 or 2 food per turn. When building a city, the unit settler will give an instant size 5 city, which may have a temple and a marketplace already built.

If an ICS player keeps on expanding without building towards megalopolis he will face superb units which will wipe out his entire empire easily. Thus, sooner or later he must build city improvements.

Another virtue of the megalopolis idea is that it reduces micro management. Moving a group unit is also faster than moving 5 independent units.

[korn469]

Xen Yu,

for one thing you are wrong about the megalopolis making all other units obsolete. there is an attack to defend ratio advantage in Civ2 and SMAC. in civ2 there are some times when it doesn't exist but in SMAC the 2-1 attack to defend ratio always exists and many times its 13 to 3 attack defend ratio (shard against plasma). so basically your megalopolis unit is just five ordinary units, which a copter could kill in one turn. and once again chop and drop rules the late game.

with each new post of yours you increase the power of the megalopolis and you never address any of my concerns. the megalopolis idea does not stop ICS. i use ICS in SMAC, it is my strategy. to play any other way is foolish. there comes a certain point when i stop expanding because i can't take the micromanagement. i know how to use ICS to full advantage, so i try to limit what i know works.

your idea just keeps on making ICS look more and more attractive. you can easily expand early and then use the power of population booms (we love the presidents day) to quickly build the industrial might and population size of your ICS empire. a competant player with 4 size two cites can easily catch up to a player with 2 size four cities, if the size four city person doesn't go on a mass ICS expansion streak of their own. building lots of early colony pods then going for a megalopolis in your core cities won't slow you down very much. then who ever gets megalopolis first rule the game...with the cheap super settlers in four turns your could build four large cities and be well on the way to starting a new megalopolis.

it would work like this. i ICS and get a megalopolis first...if i have a space to build a new megalopolis on and it is connected by railroads then in eight or nine turns i can have a new megalopolis. all the while your idea does not stop ICS it encourages it. with the super settlers it would easily make the IMS (infinite megalopolis sprawl) a new term that civ4 would have to fix.

korn469

[raingoon]

Tare a couple things missing -- or I missed -- in this discussion:

First, I would add that I for one am against any artificial rule that penalizes a player for practicicing ICS. ICS needs to be treated at the root, not the stem. Some of the good ideas, like bureaucracy points are suspect for that reason. Once you start inventing intangible commodities like "bureaucracy points" it feels artificial. Not saying I know for sure it's wrong, I just know it's a step in the wrong direction.

State of alert -- again, artificial. Why should a huge nation take 20 to 30 turns to mobilize when the US did it in about a year in mid-twentieth century? It's another good idea that for me approaches being a false penalty.

Civil wars, as always, a must. Great idea.
Losing tech -- this is a genius idea, so long as there is an advance that will make you immune to it. Otherwise it becomes a false penalty well before you approach the modern.

So that's an idea of what I mean by not imposing any false penalties on the problem of ICS. The perfect Civ 3 would allow for ICS to happen in theory, but it would require a great deal of skill.

Second, and importantly, ICS is not in and of itself the problem. The main problem of ICS is the units ICS allows the player to spawn at below the average support cost to those not practicing ICS.

Other issues such as the happiness model are secondary to the unit support problem, in that the latter can make you unfairly lose the game. It's just too easy to mass large armies for less, when one practices ICS.

Third thing I want to introduce into the discussion is what I've been talking about elsewhere -- the fact the energy resources have not been adequately modeled in the game. I've said this in the Energy thread, and in the Energy Barrel model I proposed. The gist of it is, if you want more it must cost you more, not less.

For me ICS exists because Civ 2 does not model resource dependency adequately enough. Make no mistake, World War II was a war fought over resources. The control of oil fields was a fundamental element in turning the tide against the Germans. So, to agree with Korn469 in part, unit supply is really the kicker. In the energy resource model I posited there is a "sliding scale" -- meaning as unit capability increases, the required amount of energy to support them increases proportionately. Or even geometrically, if needed. The point is to link not only unit supply, but unit production, transportation, and the Civ 2 trade stream, into an energy commodity that will serve as a natural check and balance against unrealistic, no-cost expansion. See my Energy model in the EC3 Fixes or in the Energy Thread for The List. The details of it are not what I'm advocating, I'm sure they're wrong -- it's the concept in macro that I want to add to this discussion.