how would you go about making/influencing AIs to create colonies?
Rules
The rules for an AI creating a Colony:
1. The player's Palace must exist on one landmass and the Potential Colony must consist of 2 or more Cities on a second landmass.
Obviously, the game's settings must allow players to create a Colony. If Vassal States are disabled, Colonies cannot be created.
If you look at a City Screen view and hover your mouse over top of the Maintenance area, you will get a breakdown of the Maintenance types.
On a landmass where you are eligible to create a Colony (your Palace is not there and you own at least 2 Cities on that landmass), the Maintenance breakdown will list Colony Maintenance.
The AIs use this Colony Maintenance value in determining whether or not to make a Colony.
I believe that it goes something like the following, consisting of values from that particular player's Cities:
a) Add up all of the Colony Maintenance values on that landmass
b) Add up all of the Hammers, Flasks, and Gold earned from that player's Cities on that landmass
2. If a) is bigger than b), and possibly even if the values are tied with each other, then that AI player is eligible to create a Colony.
3. Next, I think that the rule is that this AI player must not be at war. It could instead be that there are no nearby threatening Units, but I've only had luck with an AI creating a Colony when they are not at war, and thus I believe that the rule is when the AI player is not at war.
I am uncertain how Minor Civs would count--in our recent 2023 game, we only had a single landmass and thus Colonies could not be created, anyway.
This point should make you start to realise that you need at least 2 landmasses. Even a one-square island can be sufficient as one of the 2 landmasses, since that one-square island can be the landmass which holds the Palace.
From your own experience in a multiple-landmass map, you may recall that 2 Cities on a separate landmass are quite affordable, and 3 Cities is about the maximum reasonably affordable amount, with any higher City count getting considerably more expensive for Colony Maintenance in all of the Cities on that landmass.
Strategies
It helps if you have disposable Cities that you are comfortable not being able to use for a period of time.
It helps if your disposable Cities have reached Size 2 at any point in time, while under the control of any player, so that they will not be auto-razed when captured. Replacing Cities with Settlers repeatedly can be quite expensive.
If you plan to reuse Cities, it can help to have 0 of your own Culture in those Cities for 2 possible reasons:
a) Upon recapture, the City will be in City Revolt for the maximum amount of time possible, which is 3 turns for a Size-1 City. That way, when we gift the City to set up the next Potential Colony, the City will contribute maybe roughly 1 to 2 Colony Maintenance (or perhaps still 0 Colony Maintenance; I am uncertain) and definitely 0 Hammers + Flasks + Gold.
b) Upon recapture, it might be possible to get credit for Liberating a City to an AI. If no player has Culture in the City, then distance rules come into play for Liberation credit. Note that distance to a Palace on another landmass is not the same as distance to a Palace on the same landmass, as another landmass is considered to be "much farther away" than our normal distance calculation. If another player has dominant Culture in that City, we might get Liberation credit if that other player's Palace is reasonably close (but does not necessarily have to be the closest).
In many cases, ideally, we want an AI to create a Colony out of 3 Cities. That way, it can be much faster to create successive Colonies. That fact is because an AI player will accept a gifted City on another landmass when that AI player has 3 or less Cities. Thus, a newly formed Colony could be gifted a City on another landmass. I'll leave it to the reader as an exercise to imagine some of the possibilities that these facts set up, although I will end up mentioning some of them.
If we want to have the same AI player create a second Colony on the same landmass that the AI player created its first Colony, all of the Cities of the first Colony must no longer exist on that landmass. It could be that you eliminate the Colony or it could be that said Colony survives in another City on another landmass.
In my experience, the AI may need as many as 7 Cities on the second landmass before that AI will create a Colony. I've been able to do it with the AI having 3 Cities. It may be possible to do it with 2 Cities, but I'm not sure that I have done it with less than 3 Cities. 1 City would be impossible.
Cities in Revolt MAY or MAY NOT contribute to the Colony Maintenance count--I am uncertain. However, they WILL NOT contribute to the Hammers + Flasks + Gold count. Thus, Cities being in revolt at the end of our turn are highly desireable.
If the Cities cannot be in Revolt, can we influence the AI not to earn Hammers, Flasks, and Gold?
a) How about... if the City is Size 1 and there is a G Pig (not a G Riv Pig) within that City's Cultural Borders? It provides 6 Food and thus it's likely to be the square that gets priority to be worked by the City's Citizen. 6 Food provides 0 Hammers, 0 Flasks, and 0 Gold. There is still the City Centre square and there are still Trade Routes, though.
b) What if a City is not Coastal, not on a River, and does not have any Roads or Railroads connected to it? It won't have any Trade Routes. Trade Routes typically become Flasks and/or Gold.
The "ease" of creating multiple Colonies comes from an AI only having 3 Cities in its Colony. If we need to provide at least 4 Cities, then there's not much concern with having to give away more Cities so that the Potential Colony's City count is 5, 6, or 7. Basically, if the AI isn't making a Colony with only 3 Cities, and I have extra disposable Cities on hand, and that AI's tech level is low enough that the garrison Units won't be too difficult to kill, I might just go ahead and gift up to 7 Cities, so that the creation of the Colony is almost guaranteed to happen on the following turn.
As for which garrison Units get created, they might be: Warriors if you are lucky, Muskets or Riflemen if it's later in the game, Infantry if you are unlucky, and Axemen/Macemen if is still earlier in the game but you give away a City settled on Copper/Iron and/or have not Pillaged all sources of Copper/Iron that are within the Potential Colony's Cultural Borders.
Tactics for an Existing AI's Cities on another Landmass
Privateers can help considerably with an AI's Cities which are not in Revolt. A Privateer can be used to set up a Blockade that cuts off Trade Routes. Trade Routes are one source of Flasks and Gold. Note that when you are at war, other boats can be used to set up a Blockade, but then the war would prevent the AI from creating a Colony.
If an AI has multiple Cities on another reasonably large landmass, you can gift/Liberate additional Cities to that AI on the same landmass to encourage that AI to create a Colony.
Tactics when using your own Gift Cities
Try to capture 3 Cities and while those Cities are still in Revolt, gift them to an AI which is eligible to form a Colony on that landmass. Eligible, as we recall, means their Palace on another landmass, that AI not being at war with any player, and that player's Vassals (which include Colonies) not having any Cities on the landmass where you would like to have a Colony created.
When moving a player's Palace to another landmass, you may get lucky with a Palace's movements, but the most reliable way of moving another player's Palace to another landmass is to first ensure that said player actually owns a City on another landmass and then capture all of that player's other Cities. With only 1 City remaining in the hands of that AI, said City will automatically become that AI's Palace City.
Be careful of Diplomacy. You'll need a lot of warring. It may not be wise to do this stuff at a point in time where you still need reliable trading partners, as your war declarations on multiple AIs are likely to anger all of the other AIs. Do remember that you have one potential weapon in your arsenal to counter this diplomatic effect: an AI will only count your war target as a Friend if the average relationship is Pleased or higher with every member of that Team, rounded down.
A Team will consist of a Master, its Vassals (and hence Colonies), and its Permanent Allies.
Summary
Overall, it can be a lot of work to influence an AI to make a Colony. The investment in resources can be pretty high.
Consider that by being less greedy, one could keep less Cities, could leave the original AIs in place with their large Cultural Borders, and could simply focus on making one's existing Cities be productive. Strategically, this approach is probably better on most maps.
That said, it can be quite a lot of fun to cause AIs to create Colonies.
It can also offer peace-of-mind on a map where you are likely to be ever-checking the Domination Land Limit, although I will caution you that you will want to plan to increase the Domination Land Limit much earlier than nearly reaching it, probably when you have about 35% to 50% of the world's Land Area. If you get too much more Land Area before you start causing the AIs to create Colonies, you might accidentally trip Domination when the changing Cultural Borders cause your other Cities to expand more culturally than you had planned.
Further Research Needed
As you can see, there is some code-diving that could help, here.
Does Colony Maintenance count at 0 or as a greater value than 0 when a City is in Revolt?
When there is a tie between of a) and b), can an AI create a Colony?
What XML factors may contribute to an AI being willing or unwilling to create a Colony? I have noted anecdotally that some AIs appear to be more willing to create a Colony than other AIs, and it could perhaps be that an XML value affects one of or both of a) or b) as a multiplying value.
As I'd sort of planned at the 1 AD mark, I soon went after Ragnar, who was rather stifled by the Dutch, with trebs and elephants, and then attacked the Dutch from two sides, vassalizing him. He'd bribed Victoria into our tiff, however, so I went after her next, and also vassalized. From there it was just a matter of time since I had at least twice as many cities as anyone else, though was still behind some AIs in techs. Germany and France, both very small, were wiped out next, and then Joao became a vassal and source of several techs via trading down the road. By that time I had grens and cannon, and sent everything south to go after Spain. She was in a war with America at the time and pretty easy pickings, but decided to become an American vassal after I'd captured 3 cities and was poised to take another two. Fortunately, there was no 10 turns of peace so after waiting a couple turns to heal up, reposition, and upgrade cannon to artillery, I attacked the America/Spain combo and had pretty overwhelming numbers by that time, plus those sweet sweet artillery. Unfortunately, both of them had frigates so I couldn't do much in the way of seaborne attacks, but cavalry moved things along pretty quickly by land.
In 400 AD, I resumed the war against Victoria with 20 units including 7 catas. I wanted to take 1 city and feudalism, but she was ready to give only the alphabet, so in 560 AD, England ceased to exist. The same thing happened with Spain, which I attacked in 475 AD and destroyed in 680 AD.
.
