Recently upgrade my CPU to 12th gen Intel and played some late games. It seems that Canada under AI's control always to fail colonize western Canada. Winnipeg, maybe, but much later than actual. Alberta, barely. Vancouver, worse than barely.
Wondering if it's the case in everyone's game play.
Also I think there's some mechanism that factually discourage AI's Canada from colonizing the west. Even if you assign it a few more settlers, or remove those useless tundra plots' settler value, AI would still rather let the settlers sleep in Montreal than expand to those plots of 700 settler value like Vancouver.
Is it that the way to the west is blocked by Americans？If American cities are placed too north then Canada may not settle Winnepeg etc. because they sit too close.
Is it that the way to the west is blocked by Americans？If American cities are placed too north then Canada may not settle Winnepeg etc. because they sit too close.
geographically blocked? not at all. Does Grand Rapid/Sioux Fall's expansion makes the algorithms feel blocked? very likely. It's annoying that we have no way to know inside the black box.
Yes. AI will avoid settle cities that are too close to already existing cities. Grand Rapid is too close to Canadian territory that it make Canada avoid settle a city there. Go north further there's ice and tundra where Canada don't want to settle either. So Canada just park their settlers at home and do nothing.
Not sure if the AI cares at all about this but when I play as Canada I do feel a bit of pressure to avoid founding cities because their economy at the beginning is atrocious.
Not sure if the AI cares at all about this but when I play as Canada I do feel a bit of pressure to avoid founding cities because their economy at the beginning is atrocious.
You should. I notice that whenever I spam settlers first, the hit on economy feels really bad, but as long as you also spam workers and quickly improve the BFCs and build infrastructure, it pays off rather easily. I agree though that Canada has a rather slow expansion. Not sure if giving them more starting settlers will help.
Recently upgrade my CPU to 12th gen Intel and played some late games. It seems that Canada under AI's control always to fail colonize western Canada. Winnipeg, maybe, but much later than actual. Alberta, barely. Vancouver, worse than barely.
Wondering if it's the case in everyone's game play.
Also I think there's some mechanism that factually discourage AI's Canada from colonizing the west. Even if you assign it a few more settlers, or remove those useless tundra plots' settler value, AI would still rather let the settlers sleep in Montreal than expand to those plots of 700 settler value like Vancouver.
Regarding giving them more settlers, in one game I actually tried giving them settlers to make them keep America from getting too big, but they founded them all in northern Quebec lol. I did only try it once, but I don't think giving them more settlers to begin with would fix it because of that. I know when I gave them settlers, the first couple came in far apart and they settled westwards, but when 3 or 4 arrived at once while the first ones were traveling, they mostly went north into the useless tundra. I don't know though if that was because the AI couldn't think ahead thought that the west coast was too far away from its cities or if it just thought the tundra was better to settle than anything over there for some reason lol.
Another thing is that I wonder if the AI is able to see that tiles in Seattle's outer area can be stolen with culture if you found Vancouver. That could possibly explain it as there aren't many appealing tiles on the west coast area except for that and maybe Victoria, but it's usually pretty crowded by Seattle by the time Canada spawns anyway, so I always assume that I'm gonna have to go crazy on the culture there, which doesn't sound like a thing the AI would think of in advance.
I wonder if it's possible and would work to give it an "AI only" UP of a bonus to settler production? That way it might end up actually settling a good chunk of the area controlled by Canada irl both west-east and north-south, kind of like how Russia always settles the useless tundra tiles anyways which makes it look more like it does in the real world. If it's worried about Seattle's culture, maybe making Prince Rupert a nicer place with more settlervalue could at least make it try to go all the way west, though changing the tiles there too much could unbalance it for human players.
I dunno, just throwing out some ideas to see if any are on to something
The target that an AI settler may choose is a black box. Sometimes you can predict or intervene it but sometimes you can't. I'm confused on that because I always expect cities to be found in historical sites, while AI always not. Like in this example that Canada don't settle westward, or England don't send settler-Galleons to overseas, or Egypt never settle the world-famous Alexandria. It cost me quite long time to find a possibility that fits the real world best lol.
What's more, I don't think AI is smart enough to build a city and steal the plots. They just avoid close settlements in most cases, even if they can get enough plots from culture competition. But to speak twice, that's a black box and I can't say a clear conclusion.
The target that an AI settler may choose is a black box. Sometimes you can predict or intervene it but sometimes you can't. I'm confused on that because I always expect cities to be found in historical sites, while AI always not. Like in this example that Canada don't settle westward, or England don't send settler-Galleons to overseas, or Egypt never settle the world-famous Alexandria. It cost me quite long time to find a possibility that fits the real world best lol.
What's more, I don't think AI is smart enough to build a city and steal the plots. They just avoid close settlements in most cases, even if they can get enough plots from culture competition. But to speak twice, that's a black box and I can't say a clear conclusion.
They provide weights for specific locations and prevent some entirely. However, a lot of other factors determine city locations. For example, if the AI decides that forever reason a tile should not be settled (e.g. due to proximity to another city) then the settler value does not even come into effect and has no impact on that decision.
I would second the suggestion of Canada getting a couple of settler events as it's really quite sad to see half of their territory unclaimed in the year 2000.
I would second the suggestion of Canada getting a couple of settler events as it's really quite sad to see half of their territory unclaimed in the year 2000.
Yeah, I don't think I've ever seen Canada settle further West than Winnipeg. They always prioritize settling unproductive cities in Northern Quebec which end up crippling their economy. I don't even think it's proximity to American cities because then they would settle cities in northern Manitoba, Saskatchewan, BC and the territories. If Canada can't be made to settle more aggressively then it should probably get settler events.
They provide weights for specific locations and prevent some entirely. However, a lot of other factors determine city locations. For example, if the AI decides that forever reason a tile should not be settled (e.g. due to proximity to another city) then the settler value does not even come into effect and has no impact on that decision.
I noticed that in 1.16.7 Britain sometimes (say 60%) colonize Vancouver around 1800AD, which I had never seen before, and pass it to Canada. Is it out of an intentional minor change, or is it just the side affect of other changes?
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
