Things known for sure:
1. There must be an 'open slot' - i.e. less than 7 active civs in the game.
2. The capital must be captured.
3. The civ that lost the capital must be 'bigger' than the one that captured the capital.

What is in doubt:
Basically what the definition of 'bigger' means. From testing I have done and what I have experienced from past games, bigger means more cities. Others believe that population may play a part. Either way, since most human players usually have both more cities and a larger population, civil wars are pretty rare.

Question 2, grounding ai space ships: It shouldn't matter how the capital is captured, the ship should come back. However, if the AI has 1000 or more gold, it can relocate its capital to another city if it is captured and thus the spaceship continues to soar. The human player can also purchase a new capital if the old one is captured for the same 1000 cost. If this is the case, you should receive a message telling you that the capital has moved. I don't think that you need an embassy to get the message but you might (I usually have the UN and the automatic embassies)

I hope this helps a little,

quote: 3. The civ that lost the capital must be 'bigger' than the one that captured the capital.

In my most recent came I lead in pop and cities, nuked a capital and paradropped in there and split them in two.

quote: Originally posted by SMACed on 02-26-2001 11:19 AM In my most recent came I lead in pop and cities, nuked a capital and paradropped in there and split them in two. Now I'm confused, maybe it's broken again ?

Did you capture more than just the capital city in your turn? From other's experiences in the past, it appears that the calculations are done at the beginning of a turn rather than at the time of capture. Maybe not, but just a thought...

but for anyone who is interested in testing- the scenario: battle of the sexes (in best of the net2 scenario folder with the fantastic worlds expansion) is a good place to run a test. I only played that scenario once as the men. This was a while ago, and I think I was only playing on Warlord or Prince (it seems higher levels the ai moves his capitol when it is captured more easily). But I found the scenario fairly difficult at first. I had no hope of conquering all their cities quickly. So I went for their capitol. And they split (the scenario designer should have made this no schizm but they didn't). It was cake after that (I made piece with one civ and conquered the other- I even allied with the one civ )

quote: Originally posted by the_orange1 on 02-26-2001 11:58 AM That's a bit confusing. Has anyone else managed to cause a civil war even if a bigger civ? Maybe it's a probability thing, where being smaller increases the likelihood of a civil war, but it's still possible (however remote) when your civ is bigger. Any other thoughts?

I haven't done it myself, but I've seen a bigger AI civ -- bigger in every way -- split a smaller AI civ. And, fo what it's worth, neither of them was biggest (I was). This event put to rest, for me, both the idea that the split civ had to be bigger and the idea that the biggest civ had to be involved. Now I'm as confused as everyone else.

In one case, the huge Mongol empire was squashing the once-grand Spaniards. The Spanish were reduced to about ten cities when the Mongols (1st by far in pop and cities and might) captured Madrid. Civil War!

In case two, the English lost London to somebody, and a civil war resulted, despite them being middle of the pack in all ways. The only difference I could think of is that they were vastly spread out across infertile lands.