The trend away from ICS from Civ 3 through Civ 4 and into Civ 5 was a good thing. In Civ 3 and Civ 4, there was a point where people had to automate cities/workers when the number of cities got too high. In MP in particular, it is both more interesting and more challenging to compete with people who are able to keep track of every city in their empire. I think 3-4 cities is the new Civ sweet spot. I seriously doubt it will evolve into a dedicated OCC Civ based upon somewhat pessimistic slippery slope logic.
Civilization games always were a hybrid of
- game (make interesting choices in context of given set of abstract game rules) and
- historical simulation (recreate human history).
Civ5 supports singleplayer and multiplayer. Civ5 supports mapsizes from duel (40 x 24 = 960) to giant (190 x 94 = 16.920). Civ5 supports different victory conditions as science (space race), domination, culture (tourism) and diplomatic victory. Limiting the game to 3-4 cities per player by designing the game rules so that 3-4 cities (Tradition, National Wonders) for the first half of the game are the optimal strategy for winning the game fast is overemphasizing multiplayer and the peaceful player on standard map size. 3-4 cities doesn't work that good for conquest and on bigger map sizes.
World history is a history of huge empires and nations with military, economically and scientific power, e.g. the Roman Empire, British (Colonial) Empire, Russian Empire, the United States, China, etc.. By constraining yourself to 4 cities you no longer can recreate the huge empires from ancient history and you loose the simulation character of Civ and a lot of fun. On bigger (giant) map sizes it just feels bad and unhistoric to leave all those fertile land and ressources unclaimed and unsettled just because you have reached your strategic 4 cities limit. I think optimal number of cities should scale with world size. People who like to play multiplayer with 3-4 cities could use their (standard) map size while people who like huge ancient empires would play huge or giant with 8-12 cities.
ICS is a strategy based on game rules going back to design of Civ1 exploiting the free ressources per city / building (e.g. free city tile) and the lower food cost for population growth in small cities. (In Civ it is easier to grow 2 cities to size 20 than 1 city to 40.) The strategy is boosted by the fact that research, production and income of the cities in civ games are mostly additive quantities which means more cities are better. In Civ5 there are a number of ICS-counter-measures implemented as global happiness, increased Tech Costs (by 2% per city), increased Social Policy Costs (by 5-15% per city) and increased National Wonders costs (by ca. 25% per City).
By adjusting game rules to real world (history) rules it should be possible to limit the benefit of ICS to make it irelevant. (The game developers had 20+ years time to fix the ICS problem.)
- In real history ICS correlates with claiming unsettled (fertile) land and ressources before your neighbours do until there is no more free land to claim. Claiming as much (useful) land as possible is humanely. Today Antarctica is the only (partly) unclaimed and unsettled continent left on earth.
Until the age of industrialisation (and the associated urbanisation) humans tried to settle every fertile spot on earth (subsistence economy). However density of population varied. (Today it varies from 1,77 inhabitants per km^2 (Mongolia) to 18000 inhabitants per km^2 (Monaco Citystate).)
In Civ games it is not possible to politically or military claim territory. You need to found cities and produce culture / use culture-bomb or you need to conquer cities to claim territory. Population density of non-city-tiles in Civ is binary (0 or 1).
- Population in Civ is restricted to cities. Population Growth in a city is based on surplus food production with increased costs based on city size.
(In real history, fertility was high, especially in the countryside, while mortality was based on available food, disease, health, etc. In unhealthy cities mortality rate often excelled fertility rate and the cities depended on immigrants from the countryside. Population Growth = (Fertility - Mortality) * Population.)
To allow food-based population growth in a new founded, not improved city, the free city tile (in some versions +2 food independant from terrain) was introduced in Civ1. Without the free city tile most cities would have zero growth or would even starve right after foundation of the city.
Implement a realistic population growth in Civ (based on population and health, not only on food) and you can remove the free city tile ...
- Remove all flat boni per city / building.
- Population in Civ5 cannot be moved from one city to another. You can only build/buy settlers who found a new city with size 1 (or higher with certain sp). Building a settler stops population growth in the city but it does not reduce population. If you buy a settler in Civ5, you can even avoid the population growth stop and you get extra population for gold. (In Civ3 a worker cost 1 population and a settler cost 2 population. It was possible to build workers (-1 pop) and later add them to the population of the same or another city (+1 pop) (until a certain city size was reached.) When razing a city in Civ3 the population was transformed into ethnic workers which could be used to build infrastructure, be traded or be added to your cities as population.)
Food and production in Civ are also restricted to cities cannot be moved from one city to another (except for the caravan in Civ1). (The Civ5-BNW-Traderoute does not move production or food between cities but instead gives a flat bonus. The player can choose between an income-bonus or a food-bonus or a production-bonus.)
In real world, working population, production ressources, luxuries and food are movable commodities.
Change population to a global movable population which can move from city to city and the player can decide on his own if he wants to found 100 size-1-cities with lots of ressources (ICS), 10 size-10-cities or 1 size-100-city (if he can employ all 100 workers in that city). Add tile-improvements / buildings which allow multiple workers (pop) to work highly productive tiles / buildings and the benefit of founding a new city becomes a tradeoff with the highly productive worker you loose in your major city.
(This population and employment model is inspired by Colonization but population growth should not be based on food production.)
It also should be possible to move or delete cities if a city-location no longer is beneficial.
- Research in Civ is simplified. Research in Civ is an additive number. Research roughly scales with Empire-size. If you have a city which generates 100 Research Points per turn and a technology costs 3.000 Research Points, it takes 30 turns to research the technology. If you have 30 cities with 100 Research Points, it takes you 1 turn and if you have 300 cities you would make 10 technologies that turn. (Civ5 has a cap of around 1 tech per turn but research overflow is stored.) This Research modelling is unrealistic, e.g. communication effort between your research teams in different cities is completely ignored, there are no minimum research times, etc. (In Civ3, minimum research time was 4 turns, maximum research time was 40 turns, techs could be traded for Gold per missing Research-Point.)
Compare it with game development :
If a Firaxis-team of 40 developers might develop Civ6 in 30 months (?), it is not said that a team of 400 developers can do it in 3 months or 4000 developers can do it in 9 days. It is more likely that 4.000 developers in 20 teams would produce 20 different Civ6 games in 6 months.
So in real live science output of a strong 100 city empire might be slightly better than that of a strong 10 city nation but it probably won't be 10 times better.
ICS in terms of research would become completely irrelevant if every city would have its own research programm (= tech). The tech-tree usually only features 6-8 different techs at a time and so the player would only need to build 6-8 Research-Cities or less if there is a dominant Super-Science-City. It would not matter if there are 100 minor cities in the background since they would not contribute to research. Research progress would become more predictable if it would not scale with number of cities.
- If you play Civ games with a 100+ city Empire and compare your empire with historical empires, you quickly notice that elements like administration and bureaucracy are missing and you cannot establish even a primitive hierarchy, e.g. distinguish cities between military / colonial / trade outpost, city, provincial capital, capital. In Civ games you just distinguish between capital and normal cities, in Civ5 also between puppets and non-puppets. There is no administrative hierarchy, no complex structure. Every city can build most of the available buildings.
In a more complex model, there would be separate buildings for capitals, provincial capitals and cities, e.g. build schools everywhere but universities only in capitals and provincial capitals. Minor cities would expand the area of their provincial capitals and would provide food, ressources and working population for the provincial capital.
- A wide empire might have higher administration costs. In Civ1-Civ3 the cost of a wide empire were called corruption and reduced income and production in peripheral cities. Courthouses could reduce corruption. In Civ3 there was also a specialist to further reduce corruption in a city. Communism in Civ1 provided a flat corruption in all cities. Democracy in Civ1 eliminated corruption at the cost of unhappiness per military units and limited belligerence. In Civ4 the upkeep for buildings in cities, corruption and empire administration costs were replaced by a new empire upkeep cost based on number of cities, city size, distance, colonies, etc....
In Civ5 the upkeep for buildings in cities came back but global administration costs were removed. Instead global happiness and unhappiness per city was introduced, which can be neutralized by certain social policies. (So Civ5 supports late game ICS / Conquest more than previous versions of Civ, but usually it means war and if you are winning anyway late game expansion becomes optional.)
Administration of real empires usually increase bureaucracy. In a civ game bureaucracy could be implemented by having a special administrative building in capital were officers (specialists) control the minor cities. If your empire expands, the capital will instead control a set of provincial capitals while the provincial capitals will control the minor cities etc. Cities which are not controlled would not contribute to empire income. Officers would consume food and gold and do not produce any ressources on their own.
That are just some ideas how to make ICS less relevant.