Permanent war

[Thalassicus]

What's the difference between these variables in GlobalAIDefines?

permanent_war_other_chance_close
permanent_war_other_warmonger_chance_close

 

[Spatzimaus]

I'd assume that the first is the chance that an AI goes into "Permanent War" status against a major empire they're at war with, while the second is the chance of the same thing happening after that particular major empire has been branded a "Warmonger" for conquering too many city-states (4 or more, I think?).

 

[Thalassicus]

I'm basically searching for variables which control the chance of citystates declaring permanent war on attackers. I haven't ever seen a major civ declare permanent war, so these variables seem likely candidates. The ambiguous part is that "warmonger" reference. That word is also used in the diplomacy defines, which would hint at major-civ usage, but I don't know how these would be used for major civs.

Perhaps "warmonger" refers to the chance if the player attacks the citystate, while the non-warmonger variables refer to when the citystate's ally attacks the player?

 

[Spatzimaus]

I should have been clearer. These variables are ENTIRELY about city-state-vs-empire relationships, and there's not that directionality issue; the difference between the two is in what YOU have been categorized as by the AI.

My guess is that these variables' chance values are used to decide whether a city-state you're already at war with (for any reason, whether they're allied to someone you declared on or whether you declared on them directly) decides to make that state permanent instead of just allowing you to go back to a peace state (albeit at -60 Influence) the moment you make peace with their main ally.
The first variable was the chance that this would occur under normal conditions, while the second was the chance that this would occur after you'd been labeled "Warmonger", which has a very specific meaning to the diplomacy AI. Basically, if you've conquered (or just declared war on, I forget) two city-states you get labeled "Aggressor", which causes a few specific variables to increase, and if you conquer four you get labeled "Warmonger" and another set of values get used. Obviously, a city-state's going to be less likely to forgive and forget if you've got a history of conquering other city-states, hence the separation between these two variables. And once you get that status you never lose it, so you can't conquer the local city-states and then later on hope to ally with the ones you discover on another continent.

 

[PawelS]

Hmm... interesting question...

I think I understand the "permanent", "war", "warmonger" and "chance" parts of these variable names, but I still don't understand "other" and "close"...

Edit: I checked the other variables, and now I see that there are 4 categories of distance between you and the CS: distant, far, close, neighbors. And "other" can mean a situation where you didn't directly declare war on that CS, but on the major civ allied with it...

I also noticed the thresholds for "aggressor" (2) and "warmonger" (4).

But I still don't understand this one:

PERMANENT_WAR_OTHER_AT_WAR

 

[Spatzimaus]

Edit: I checked the other variables, and now I see that there are 4 categories of distance between you and the CS: distant, far, close, neighbors.

Exactly. Proximity is a big factor in this; presumably they didn't want there to be as big a possibility of a permanent war lockout for a city-state on the other side of the planet, which was presumably never in danger from your war with their ally. But if you're right next to them and were at war with them, there's a chance of permanent war.

Frankly, proximity should have been a major factor in the Influence gain/loss already, to where distant city-states are harder to gain or lose major Influence. And the ease at which you "prime the pump" should depend on your history with that city-state; one you've just met recently should be harder to raise through gifts than one you've been a neighbor of for several millenia.
But they didn't do that, so right now it's in your best interests to be friendly with whichever city-state gives you the best effect (including resources), regardless of where it's located. And that's why this "close" part seems so out-of-place.

But I still don't understand this one:

PERMANENT_WAR_OTHER_AT_WAR

From the context, I'd assume that this is the chance of a city-state declaring permanent war on you for being at war with their main ally, in situations where none of the other variables apply. (Example being that other-side-of-the-world city-state I mentioned above.) But it should be easy enough to test.