I the idea of a stability system may make of the best additions to civ games that I have seen in the last few years. But I am a little afraid that the AI wont play competitively, accounting for the stability system. Could that be the cause of the high number of collapses we’re seeing?
To me it seems that civs are collapsing too easily. Making stability dependant of continued growth may only worsen this. And make the mod less historically accurate, as there have been many empires/civilizations that were stable with little growth (at least in area), even with isolationist policies that went on for centuries (China and Japan come to mind). Other very aggressive empires have collapsed as its founder or his immediate heirs died (a number of indian and central asian empires). Others yet, like the Byzantine Empire, managed to drag its decline for centuries, without any kind of sudden collapse - what territory they lost was lost to outside enemies.
Regarding the fall of empires "from within", I believe it happened when the local elites (and all agrarian empires depended on local elites for any kind of effective rule) decided to break away from the centre. They would eventually attempt that if they realized they could extract more wealth or power for themselves (not necessarily that their region would become more wealthy) after becoming independent from the centre that they did as part of the empire, and if they believed they could get away with it. Mere economical theories err in not accounting for the fact that the fate of agrarian empires was decided by the will of the local elites (a small percentage of the population) to cooperate with or oppose the centre. Their beliefs and those of the local population, and the way they established their legitimacy to rule the local population, might be much more important than the wealth of the region and trade relations they had with the rest of the empire.
Often the idea of the empire, and its law, survived in the conscience of the population the disappearance of central rule (western roman empire, China in a number of occasions, the arab caliphate, etc.). Separation was decided and carried out “from above” (not above as in the political centre of the empire, bus as in the political regional centres).
A historically accurate stability system would have to account for this (local economic and social conditions in the various regions of the empire), not just aggregate growth. This is, of course, impossible. So, as a best approach, I like the suggestion, by Guest01 in post #43, that differences in score between the core (capital ?) and the provinces (other cities? Newer cities? Cities farther from the capital than a certain average?) might be taken into account (if that is possible, of course!). Richer regions are more likely to secede successfully. Perhaps the existence of different religions could also be taken into account – religious controversies at least hastened the demise of many empires (HRE, Byzantine).
Another interesting think to take into account might be size. Distance from the capital, by itself, was not a big problem for empires. The problems distance posed were in the exercise of “cultural influence” by the centre over the more distant regions (needed to maintain some cultural cohesion within the empire) and in sending (and controlling) bureaucrats and military units from the capital to these regions. Both were much, much, harder to do in overseas territories than in contiguous lands. That’s the reason Russia and the US kept and annexed the empires they built up to the XIX century while western european nations eventually lost theirs. So territories not connected by land could have a higher likeliness to become independent that those connected.
I realize most of this would only be useful in deciding what cities should become independent in the event of a collapse, not the collapse itself, and I don’t know if it is possible at all to implement any of it.
