The ultimate ICS thread: analysis and solutions

[MadWoodster]

I still think it is unrealistic to work two tiles with a size one city (that is 2tiles per citizen) and only 21 tiles with a size 20 city (that is 1.05 tiles per citizen)!

[korn469]

MadWoodster

that is one of the problems that must be addressed before ICS will go away...what to do about the base square...we know that...

the base square is evil, stupid, unrealistic, and causes ICS

that without the base square most small cities don't have a chance of growing or building anything

ok so if we do go with the diplomat or madwoodster ideas about the first citizen works the base square, then what are we going to do about city growth?

maybe have the land under the base square count as fully upgraded and then possibly give a 1.5 times food in the square

so if you built in a desert you'd get 1 food +1 times the bounus or if you built on a food special you might get 4 food plus a 2 food bonus...this small bounus amout of food would certainly be better than the current ICS situation...especially if all cities require the same amount o food for growth...do you think that is a good compromise? the one positive is that cities would get to work up to 21 citizens now

youngsun

i like your idea about barbarians limiting civ expansion...i think this thread is spawning ideas about three different areas of civ gameplay

1. ICS (Infinite City Sprawl): a technical study of how the rules of growth, production, and support make it better to have 10 size one cities compared to 1 size ten city

2. ICS (Infinite China Syndrome): a study of how empires last from 4000bce to 2000ce virtually intact, and that over the course of hundreds of games less than ten cities will declare indepedence or revolt out of the thousands of cities you build

3. ICS (Infinite Civ Stratofication): ok i was reaching on that one but what i mean is that once a civ gets to be a certain size about mid way through the game then that civ will win in almost all cases...the fact about civ is that a small civ has no advantages over a large one...this makes for boring games

youngsun i am mostly just concerned with the technical issues of ICS in this thread but i think i am going to open a thread for the other two ICS problems...as for barbarians, maybe they could work like this...

when civ starts off each civ has borders like in SMAC, however most of the land outside of those borders are divided into different barbarian territories...there would be no cities in barbarian territories but barbarian units would roam this territory with a number of units...barbarians would start out with just horsemen but they would steal military technology by osmosis (kind of like the great library, when it comes to a military technology the barbarians could build any unit that three civs could build)...when you sent units in barbarian territory barbarian units would head out to attack unit units, and if you built cities in barbarian territory, barbarian units would come and try to attack it...if you had a city in their territory for long enough it would assimilate the territory, and push the barbarian borders back...barbarian units would rarely leave their territory but they sometimes would go on raids into the nearest civilized territory...you could sign treaties with the barbarians and ask them to attack your neighbors...barbarians would have around a fixed number of units...though massive settlement of their ands would cause them to spawn even more units, after a barbarian unit died a new one would respawn in a few years (5-15) time...there would be a random chance of a barbarian area becoming civilized...if it did, cities would pop up in that barbarian's territory and then they would be a civ

how does that sound?

korn469

[The Mad Viking]

Great thread!
I agree completely that the problem lies in the giving a new city founded by one settler, two squares to work. It's illogical and non-sensical, and it's the main reason why ICS works.
As discussed, an additional 1 food bonus would suffice to provide growth. You could also reduce the number of food required to grow from size "0" to size 1.
I think, however, that the notion of food surplus as the population growth is another problem. Cities grow because of trade and jobs (prodution) Food is needed, and may be quite important in ancient times, but a modern city with tons of trade and production not growing because it doesn't generate enough food? Please!

Also, I think small "cities" should suffer from limitations on square production like Despotism. Up to a certain size, can't get more than two trade or production from a square. Because only larger cities become centres for Trade and Industry. That will kick the s*** out of ICS.

O/T? A lot of people complain about the 21 square model and the relative scale in miles it represents. I think the base square represents the actual urban centre, with all the improvements, and each square worked is a nearby village or suburb, economically linked the the city.

Korn - your Barbarians ideas are great - but maybe take it a step further - and all rival CIVs start as "Barbarians". Have 30+ tribes and you don't know which ones will become major civs, minor tribes or stay nomads until they're wiped out.

Megalopolis - I think once cities are connected by rail, they should automatically link their windows, and pool all their resources. And yer, if production is high enough, you should be able to build a spy, an Armor and a libary in the same turn. (This doesn't solve ICS)

[Xin Yu]

Korn: Paul has answered your question for me. You play SMAC so you may not be familiar with double production.

Now I have another proposal.

Let me start with a population model:

dP/dt=A*P-B*P*P

where P is the current population, dP/dt is the rate of change in population at the current time. A and B are positive numbers: A is normal in scale but B is quite small.

The model basically says that, when population is small (thus B*P*P is neglectable), the population growth rate is proportional to itself (dP/dt=A*P); but if it grows to a certain size, then due to 'competition' (the term -B*P*P), the growing rate will decline and finally halt.

We can impliment the model to the game. Let A be proportional to the total resource available in the city radius, and let B be proportional to the number of cities in an area surrounding the city (say a 10*10 square). The size of the food box is proportional to 1/(A*P-B*P*P). Thus if there is more resources (can use 'food + 1/2 trade') in the city radius, A is bigger hence the food box is smaller; on the other hand, if there are more cities in the surrounding area, B is bigger and the food box is bigger.

An example: Suppose A=100 and B=#of cities in surrounding area (including the city itself). Food box=1000/(A*P-B*P*P).

A perfectionist puts 4 cities in the area, hence B=4. A size 10 city will have a food box size 1000/(100*10-4*10*10)=1.33.

An ICS player puts 40 cities in the surrounding area, B=40. A size 1 city will have a food box size 1000/(100*1-40*1*1)=13.33, or ten fold the perfectionist's food box at size 10.

[wheathin]

Awesome thread! 2 Observations:

City growth is indeed totally unrealistic - in fact, Civ/SMAC/CtP gets it backwards! Until modern improvements in public health, most cities only maintained size thru immigration - the city was a totally unhealthy environment, with disease, fire, and crime, and deaths greatly outpaced births. Rural areas had to continually replace the losses of cities with new workers. And those workers only came to find jobs. So if there were no jobs in an town, it would not grow. In fact, if the employment base dried up, the city would shrink very rapidly as many died and more fled.

It follows that the presence of so many people creates a demand for food that must be met. Rome imported vast supplies of grain from Egypt to feed its people, and most of the large Medieval cities maintained population with food imports, usually from central europe via the Baltic Sea, or thru the Med.
Jobs draw people, who demand food. Not Food creates people, who work jobs.


As to ICS, the problem is indeed based on the free base-square. The 2-for-one deal creates the exponential ICS growth rate. To combat it, you need to reduce it to an arithmetic growth rate (or very close to it). At that point, the natural economies of scale that infrastructure improvements afford will tilt the balance to non-ICS strategies.

But giving, say, +50% to the base square only gets you back where you started: all those additional base squares get the same +50% bonus. I think the answer is three fold:
1. eliminate the free base square, (you can work it like any other square, and can be improved - a city is never the size of one of the map squares)
2. make city growth at least flat: same food for each pop point, and
3. provide a limit on the size required to build settlers - maybe 3? No community of 20,000 is going to send half the pop away to found a new city.

One obvious result is that it takes a while to get cities up to speed because of slower growth - so then just make food/nutrients more abundant!

...or just recognize the underlying flaws and abandon the model altogether.

[korn469]

wheathin

if you adopt the the upgraded base square and give it +50% to food production, this means that the base square would produce between one to three extra food, which with an arithmetic growth system would make that small amount of extra food per base insignificant...fourty size 1 cities compared to one size 40 city would produce between 37 and 119 extra food a turn compared to 80 food 40 minerals and 40 energy a turn (in SMAC)

in SMAC it would take 800 food for those size fourty cities to grow from size 1 to size 2, and 800 food for the size 40 city to grow to size 41...under our flat food growth system it would take 1600 food for those 40 cities to grow from size 1 to size 2, while it would only take 40 food for the size 40 city to grow to size 41

i like your idea that a city can't produce a settler until it reaches size three

Xin Yu

i think your growth model might tip the scales back to far in favor of Super Cities...but please tell me more about it ok?

korn469
[This message has been edited by korn469 (edited March 02, 2000).]

[korn469]

The technical ICS Issues Summary:

Growth: the nutrient(food) box should be the same size for all cities, and once that food box gets filled the city grows in size by one population. I propose that this nutrient box be size 40 nutrients(food). This is the growth box.

In a side note, after building a granery, cities should be able to direct some of their food into another box. This box can hold up to ten food per citizen. This represents the stored food in your city. If for any reason you run out of food in the growth box, then your city will drawn on food stockpiled in this granery box. This will allow you to stockpile food in case of sieges, droughts, famines, ect. Until the discovery of refrigeration, a very small percentage of the food in this box will be lost to decay. This loss should be around one unit of nutrients per turn.

Cities should have to be size three before they can build a settler.

Support: The amount of free support generated by your empire should be based on three things. One is population. Two is support generating facilities. Three is your civs Social Engineering Support level. Your SE Support level would act as a modifier to the first two categories. We would use the following formula to calculate support.

S(P+I)

Where S equals SE support level, P equals population and I equals support increasing infrastructure. Units would not be supported by cities. Instead your civ would generate a number of support units based on the three categories. Each unit would cost so many support points. After your civ had exceeded the number of support points it generated, then shields from random bases would be converted into support points. SE Support leves would determine exactly how many support points each shield would generate. This process would happen automatically and the player would not be involved in it. Support generating facilites should be military structures (barracks) and advanced manufactoring structures (factories).

Production: Take away the free base square. Instead have it where the base square can be worked. The base square counts as an irragated square and produces 50% extra food.

Also when a city builds an advanced manufactoring facility it should add in another production slot. So that a civ that had a nanoreplicator could make two units at the same time. A city that had a nanoreplicator and a quantum converter could produce three units at the same time.

Happiness: For cities size 1-4 there should have to be one worker for each pysch specialist, with there always having to be one worker. If the worker was discontent(a drone) then the city would riot and revolt. For larger cities, size 5-20 there would have to be one worker for every two specialist, with a minimum number of three workers. If these workers were discontent they would riot and revolt. For the largest cities, size 21+ there should be a minimum number of seven workers but an unlimited number of specialists. If the workers were discontent the city would riot and revolt. Now only police, happiness spending, or happiness facilities could pacify this minimum number of workers. They could not be turned into happiness or any other kind of specialists.

Additionally all specialists should take three food to support instead of the normal two food for workers.

korn469
[This message has been edited by korn469 (edited March 02, 2000).]

[korn469]

The Infinite China Syndrome Summary:

-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.

-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!

Assimilation: This should definately be included, but I think it should be done by including nationalism, and not just have it so that after x turns a city is assimilated. Assimilation should be gradual, and determined by your actions against the ethnicity the city belongs to, the happyness in the city, the use of police, the amount of improvements you build in the city etc. A city should be able to concist of 20% americans, 70% Germans and 10% Russians.

you have alot of good ideas! they have got me thinking...how about this...use a modified prototype system like in alpha centauri...

when your civ discovers a new technology
then a check is made by checking your sphere of influence

a sphere of influence is this, 5 times the speed of your fastest unit. then apply that distance like how borders are applied in SMAC

the check follows this preliminary flow chart

1. does your civ have the strongest military?
if no prototype cost is normal

2. does the second strongest military have at least 80% of your strength?
if yes prototype cost is normal

3. is your civ twice as strong as the second place civ?
if yes add 50% to cost of prototype

4. is your civ twice as powerful than the second and third place civ?
if yes add 100% to cost of prototype

5. is your civ more power than all of the other civs in your sphere of influence?
if yes add 150% to cost of prototype

6. is your civ twice as powerful as all of the other civs combined in your SoI?
if yes add 200% to cost of prototype

7. do no other civs exist in your sphere of influence?
if yes add 300% to cost of prototype

in addition i think that prototype costs should be higher than in SMAC, somewhere between +100 to +200 percent

so in that model your first unit could cost up to six times the cost of a normal unit...the same system could be applied to technology and everything

going to edit this

[korn469]

David versus Goliath: Small states triumphing over large

summary up soon

[Youngsun]

Korn

I was waiting to see your last summary and it never came up. what happened? I know you are very busy and have so many posts to take care of but I hope you did not forget this one "David vs Goliath"

[korn469]

Youngsun

no i haven't forgot, i have been busy lately, and i am waiting for the EC3 final drafts to come in so i can pull ideas together from it into the eternal china and david versus goliath sections

korn469

[supremus]

I respect all the posters in this thread that once indentified a problem worked hard and with creativity trying contribute to solve it. But I would like to assert again I SEE NO WRONG WHIT ICS. So the problem is not a problem, or, at least, the problem is in another area.
A large state that made ICS a lot in its history is more powerfull than smaller ones. That's what history tell us. In the WWII USA and Soviet Union won because they had a immense territory and it gives them much more critical mass than Germany or Japan. (Of course it is a simplification, for several other important nations won too among the allies.)
The overall production is always higher when you dominate larger areas, mainly in agriculture and simpler industries where technology is not essential.
Otherwise, Japan and Germany are two economical superpowers and it gives them critical mass in the economical battlefield.
So, If we eliminate the advantages of a great terrytory (ICS) in CIV 3 we are going to be inaccurate historically. 20 cities of size 1 are obviously much more powerfull (even in a military strategical sense) than one city size 20, because 20 smaller cities gives you control over a much larger territory. Economically a size 20 city is more powerfull than 20 size 1 , and that's what can be improved in CIV 3, because that is the point where the real problem is. The solution should be giving to larger cities some additional financial advantage.
With this economical plus, players who prefer a perfeccionist strategy can use the extra money to import goods that he/she does not produce, to buy military units, to influence the global diplomacy, etc. etc. , exactly like in real life. And of course they can win, exactly like in real life.

[korn469]

The Joker

thanks! so besides having SE influence the growth rate here is what else should...

commerce (money from trade routes)
hospitals (maybe have this building take the food box from 40 to thirty to represent how much faster a population can grow without disease)

the prototype system is completely hammered out but i am going to work on it more...there are about five EC threads that i am looking at for ideas right now in my last two so look for new summaries on thursday or friday!

Xin Yu

i am going to have to look over your growth model more and see exactly how it counter balances large versus small but it sounds like you have been doing some research!

supremus

quote: But I would like to assert again I SEE NO WRONG WHIT ICS. So the problem is not a problem, or, at least, the problem is in another area. A large state that made ICS a lot in its history is more powerfull than smaller ones. That's what history tell us

this is not a history thread it is a gameplay thread...but i do dispute your "historical" evidence...how about great britian throughout the 19th century? i could give more examples but it is pointless to argue over history which is subject in it's interpretation

ICS is a problem because in the highly unrealistic game of Civ it is almost always better to expand that to build up your territory...the same amount of population spread out through more cities is always better than having it concentrated into a few cities, and that strategy is simply unbalanced

plus civ is incredably boring because strong states always grow more powerful and weak states always get weaker, once a human player who has beat the AI before gets on top they stay on top...it is just a matter of time before the human grinds the AI into dust with no challenge at all

there needs to be a better balance between building up and spreading out (the first summary) and there needs to be a better balance in the rise and fall of civs (my second summary) and there needs to be a way to simulate a relatively new civ (the US) who is weak, growing to be a world Super Power in less than two hundred years instead of once the romans got on top, they stay on top and send a space ship to Alpha Centauri in 1600 because they crush their competition

we are trying to fix ICS and make civ gameplay better, i would also like to make it where it is virtually impossible to win with a One City Challenge also but i think ICS is more of a problem than a OOC victory right now

summary: to have a large happy empire that stands for centuries and remains a super power throughout the game should be hard but not impossible to do. there should be a better balance between building up and spreading out with both strategies being valid

korn469

[The Joker]

The idea of having all cities grow the same way (with a 40 food box that needs filling for the city to grow) is great. If we include, that this box's size can be increased/decreased due to the SE growth rate, the happyness and the amount of trade per pop in the city this system could truly revolutionize growth in Civ games.

I also like the idea of a city having to be size 3 to grow. It not only solves some ICS problems, it is also realistic, as people don't move out of cities untill they become overpopulated.

I like the prototype system, all though it would propably be most useful for research. It would make you need other civs in your area. Great idea!

Looking forward to more summaries!!

BTW go check out my thread on the x10 model. It could really be a great addition to civ3.

[Xin Yu]

Korn: The basic model of Malthusian theory is assuming that population growth rate is proportional to the current population:

dP/dt=A*P

This means, suppose a city grow from size 1 to 2 in 5 turns, then it will grow to size 4 in another 5 turns (the speed of growing at size 2 is twice of that at size 1), and to size 8 in another 5 turns, and so on. This is called a geometric growth. According to the original theory, since consumption goods increases at a much lower rate, people will become starving and standard of living will decline. War and epidemic desease will be the only things to reduce population.

A later theory adds a term of 'competition' to the equation:

dP/dt=A*P-B*P*P

The competition term has little effect when population is small but will be more and more significat in reducing the population growth rate when the population is larger.

Carefully chose A and B will give a balanced growing for large and small cities. My previous example was, eh, just an example. If you feel it was an overkill, you can scale it back. You can add government type as a factor in deciding A and B, thus allow population grow faster in better governments.