So theres this weird exploit ive been using; civs that start in cave_detailed sites on mountains but have no settlement biomes will never create another site, theyre bound to their initial fortress unless they successfully conquer someone elses site and take over. Which os awesome for civs that arent supposed to sprawl out indefinitely.
But if you change any of that, from letting them start in any other biomes or try to do the same with cities or forest retreats, it just doesnt work. Other site types can found new sites around their mountain range even if they dont have any settlement biomes described. And if you try to set their biome preference for mountains to 0 the civ wont show up, at all. If they start in any other biomes theyll build as many hillocks as possible in a dense cluster around their fortress, but never mountain halls or new fortresses.
This behavior is really weird, and it drives me nuts that i cant find any other ways to get similar behavior. Ive spent all day running worldgen trying to find other reliable ways to stall civs development. The next best thing is to make their biome preferences so restrictive there are only a miniscule quantity of locales in a given world that suit their needs, but even that doesnt seem to be reliable, and of course you cant guarantee a worldgen where those sites are truly scarce unless you use perfect world or something where youre generating the same terrain repeatedly, which gets boring frankly.