How are you defining steamrolling? If I plan and expand smartly and then decide to crush a smaller kingdom, is that steamrolling?
I can only speak for myself:
Currently it gets very costly if you try to build a lot of cities by your own. If you have a decent income because you were lucky and have a good amount of gold mines this is not an issue and gives you an advantage towards player which were not so lucky. Secondly, if you conquer others cities you don't have any downfall and get more and more cities. Both supports city spamming which normally everyone don't want to have.
With my proposal there would be a downside and on the other hand provides some additional tactical features.
Some starts are better than others, and I'm okay with that, but I understand what you're getting at. The primary issue that I am seeing in your writing is that it is too easy to expand quickly and consolidate conquests. This means that a faction which springs out to an early lead in cities will be able to easily overpower the rest of the players, ending the game before it begins.
So there are three points at which you want to address the system:
- Founding New Cities
- Maintenance
- Integrating Captured Cities
I'm going to quickly run through each of these points, talking about how some previous games (CIV IV, FfH2, EU3) have addressed them, what the general effects of a particular tweak were, and then afterward, rephrase your proposal and sum up my thoughts.
Founding New Cities
By increasing the cost of new cities, either through pioneers or on founding, you curtail early expansion. A good example of a game which took this approach is Fall From Heaven, where settlers cost significantly ( x 2, iirc) more than they did in standard Civilization.
In the case of super-starts, it means that the difference between a super-start and a normal-start is smaller, perhaps only a 1 city difference, instead of 2 or 3 cities.
What increasing the cost of founding cities does not do, is curtail mid and late game expansion by established powers. This is made even more so in Elemental since Unit and Building queues are separated from each other.
Maintenance Systems
A maintenance system is designed to slow growth by forcing players to do some form of "building up" between additional cities, either through research of maintenance reducing technologies (part of your proposed system) or the construction of particular buildings in a city (courthouses in CIV IV). It is supposed to reflect the "overhead" associated with having control of a city.
It can reflect the need to improve the imperial bureaucracy (building up your palace in the capital), establish a civil code (technology), or the appointment of local magistrates (building courthouses in the new cities). Hopefully it also reflects how "connected" your empire is (I continue to hold out hope for such a system).
A game which used a particularly aggressive maintenance system was European Universalis 3. In EU3, the amount of research necessary to advance a level in technology was directly dependent upon the number of provinces that you controlled. A many province power would need to accumulate many many more research points before it advanced a technology level. This meant that grabbing many provinces quickly, before you could improve them, would stagnate you - leaving you exposed to smaller, more technologically advanced opponents. However, there were certain provinces which would never be able to produce enough to justify owning them if you were primarily concerned with technology, no matter how much you improved them, and it was a very frustrating feeling for me.
Integrating Captured Cities
Integrating new conquests normally has additional mechanics associated with it beyond the ever-present maintenance system. In CIV IV this was accomplished by having cities fall into civil disorder for a period of time after conquest, before the city produced anything for the conquering power. These mechanics exist to slow down a conquering power, such that it cannot immediately turn its new conquests to use.
EU3 forced integration slow-down by having the player wait a number of years for a territory to become fully integrated into their nation, achieving core status. Until a territory was considered a core, it would only produce on the order of 10% of its actual value. This meant that while new territories would immediately provide you with more manpower and trade-goods, they would not provide you with the means of supporting such expenses. If an empire acquired too many non-core territories as a ratio of cores/non-cores, it could gain the "Over Extended" debuff, which reflected the over-strained administration's attempt to integrate so many new territories.
Integration mechanics force an empire to conquer at a pace which allows for consolidation phases. These are phases where the empire is not ready to take on new territories.
Solution On Table
- Number of Cities controlled without penalty is a function of Palace "Level" and technology.
- Going beyond an integer limit imposes malus on all resource generation.
- City cost is baked into pioneer kit - pay your Gildar up front.
- Conqueror is slowed by additional cities hitting him with the "Over the limit" penalty.
I like the idea of building up the palace. I'm pretty sure I put forward a similar idea in the distant past, but I'm not going to look for it. I think it's both flavorful and creates a weak point in an empire to attack. If you give the palace multiple levels, it also becomes progressively more expensive to replace, and a capital strike on a large power becomes incredibly powerful.
Making pioneers cost more than they currently do is a twiddle value, I'd want to see how it plays. I agree that it is the correct knob to tweak if you want to make the lucky and normal start more similar in the number of cities they create early on.
The universal malus is something to be careful with, and I think having the "allowed number of cities" is a bit clunky. Here's how I would implement the same idea:
Imperial Efficiency:
Take two point pools, one called "Administrative Load" and the other called "Bureaucratic Power", shortened to AL and BP respectively. Imperial Efficiency is the ratio BL/AL. AL points are generated by cities, so a Lv. 1 City "costs" 3 points of AL, a Lv. 2 costs 5, etc. BP points are generated by "government" buildings, so things like courthouses, road connections, palaces, city centers, a technology of two, etc.
The entire production of the empire is modified by Imperial Efficiency. Eg. if your empire would normally produce 10 materials per turn, and your Imperial Efficiency is 150%, you would instead produce 15 materials.
I'd also put this ratio (as a percentage) somewhere nice and prominent on the UI (next to the resource bar perhaps?) and kingdom report.
The end result of such a system would be to encourage both small and large empires to research this particular part of a civics tree, since a small empire with lots of things improving is BP would be gaining a multiplier, while a large empire that didn't expand smartly would suffer the penalty. This makes your previous discrete system a full sliding scale, and lets the player see the immediate effects of their builds on the efficiency number.
Personally, I think there should be some period of time where the conqueror is integrating his new conquests into the empire. I prefer the EU3 form of this mechanic over the CIV mechanic. For elemental, I would make it so that a newly conquered city has an administrative control timer based on the level of the city, the level of the imperial palace in the conqueror's capital or perhaps the conqueror's Imperial Efficiency, with a penalty for the number of cities already in administrative transition. While that control timer is ticking down, the city does not harvest resources. So the new conquest can be used to build troops and structures, but won't generate the resources to support such things itself. Effectively, when you conquer a city, you immediately deny its resources and territory to your opponent, but do not gain them for yourself until you've negotiated a consolidation period.
... and that ended up being longer than I thought it would be when I started typing.
