As I said in my second post on this thread, I think too that it would need some serious modification of the mechanics of Civ. We could try to reduce those changes in a minimum way, like for example conquered cities flipping back to its former owner pretty quickly if the king is still alive. If the king is well perceived by his population, this last one could rebel to its newer ruler. You would have to keep a good army in the conquered cities if you don't want them to flip back (that would work for short border wars) to its former owner, that reducing you power in a long war (to be continued with a lot less troops) . But as soon as the king is killed/emprisonned, the cities cease to rebel or rebel a lot more less.
Late reply as promised.

This would work, as another modifier to city rebellion. I think that it would have to have a lot to do with happiness throughout your empire, also. And perhaps a city in which the king unit is receives either a happiness bonus or penalty depending on what the populace think of them.
I don't really see where you go there. Why would it need a serious change of the tile system?
Well, currently tiles are mainly populated with improvements. But if castles and forts were to necessitate having drastically more importance, then these would be needed on nearly every tile, almost making void the concept of tile improvements, without some other modification to the system.
With the nowadays Civ system that is: no ZOCs, impossibility to build a castle on a farm or cottages, etc... I think that first we should be able to build castles on improved tiles.
Well, there you go. That is a solution. Although this would need a nerf to stop absolute and total castle spamming. Perhaps the tile would have -1




Second, we should be able to intercept any enemy unit, with a system a little like ZOCs, but with more range. For example, if the movement of a unit is 2 (on roads), it should be able to intercept any enemy unit in a range of 2 tiles around. The problem being: would this interception be an attack or a defense? If it's an attack, that should be problematic if the enemy units are on a forest hill, for example. If it's a defense, that would mean that you couldn't enter an unoccupied enemy forest hill if a castle or city or any unit being able to intercept is at 2 tiles of it.
That goes against the basics of the turn based nature of civ. Ergo, I don't like it. It would be very confusing and detrimental, IMHO, to seriously mess with the turn based system like this. Sure, fighters do it, but air units are a completely different kettle of fish.
The best thing may be to nullify the defense/attack system, it is to say that defense would not be prevalent to attack anymore, unless we've time to build serious defenses, like walls. Walls should be able to be built by any kind of military units, and it should not be variable with time like the Fortify command in Civ4 (when we have to wait only 1 turn in order to increase our defense), it should be 0 or 100%, only when entirely completed.
I think that attack is much more dangerous than defence as it is now anyway, in that those that attack a city generally win. Sure, it may be at great cost, but making attacking even easier for the player would put out game balance. Then again, you could probably say the same thing for defence.
