New Beta Version (3-20b)

[ilteroi]

You can see that she occupies the northeast corner of the Pangaean continent. In order to settle Liverpool, she had to walk across regions controlled by Poland/Russia, India, the Aztecs, and Rome (Me), to get to an area that had been devoid of resources for at least an era. What sort of decision making process could she have gone through to make this decision?

well, it's not a bad spot ... not particularly good but ok. the real problem is that it's hard to defend. for england. question for me is more why none of the neighboring players wanted the spot?

anyway there's a lower limit on the city site score that the AI will accept. it's not hardcoded though but determined at the start of the game according to the map. this threshold is player specific (some like coast, some like hills, some players are more expansionist, etc)

i've got some changes lined up for the next version, try that and report your impressions!

 

[Rekk]

If you have a save (and use reasonnably few mods), I'm sure Ilteroi would be glad to have this on Github to correct AI's logic.

Also, does that zone has late game resources unrevealed ? If the AI is settling for those resources, that's not a bug in the AI settling logic, but that's a bug since the AI is not supposed to know the position of unrevealed resources.

Hmm. The game was won in the Industrial Era, but I guess I could keep hitting end turn and see if Oil/Aluminium/Uranium gets revealed there.

well, it's not a bad spot ... not particularly good but ok. the real problem is that it's hard to defend. for england. question for me is more why none of the neighboring players wanted the spot?

anyway there's a lower limit on the city site score that the AI will respect. it's not hardcoded though but determined at the start of the game according to the map. this threshold is player specific (some like coast, some like hills, some players are more expansionist, etc)

i've got some changes lined up for the next version, try that and report your impressions!

Will do! I just wanted to show an example of weird behaviour. We'll see if it is mitigated with your changes.

 

[Moi Magnus]

Hmm. The game was won in the Industrial Era, but I guess I could keep hitting end turn and see if Oil/Aluminium/Uranium gets revealed there.

You can try loading the game with in game editor, or you can save the map and start a new game on it, at information era.

 

[Basilius]

I really think you are all overthinking this. The AI planning in place is already sufficient, it just needs to be more sensitive to threat and proximity than it is now.

Respectfully Gazebo, I agree with others that it's not just "threat and proximity." The AI consistently views single resources as a viable reason to found a city, regardless of other considerations for growth.

 

[DarkZero]

I had dozens of games where the ai will settle a city on 80% pure desert just for that random 6 iron tile that spawned in there.

 

[Deadstarre]

I had dozens of games where the ai will settle a city on 80% pure desert just for that random 6 iron tile that spawned in there.

they'll do it without the iron as well

 

[pza]

they'll do it without the iron as well

i used to believe they were aware of all the ressources that spawned later in the game, and purposely even grew their border towards those not yet visible ressources. that's a belief i had since civ 3. those damn cheater AIs.

 

[ryanmusante]

i used to believe they were aware of all the ressources that spawned later in the game,

Is that what is actually occurring?

 

[Gazebo]

Respectfully Gazebo, I agree with others that it's not just "threat and proximity." The AI consistently views single resources as a viable reason to found a city, regardless of other considerations for growth.

That's not exactly what is going on. As I noted, it's the criteria for proximity to other cities that is the problem, not yield valuation.

Is that what is actually occurring?

No.

 

[tu_79]

That's not exactly what is going on. As I noted, it's the criteria for proximity to other cities that is the problem, not yield valuation.

I know you are in the case (from Github) and we have faith. Any progress so far?

 

[Heavy Q]

I've seen the most worthless positions settled by the AI too, like 1 or 2 tundra sticking out by the ice caps seems to get settled, even though there isn't even fish and half the possible worked tiles are covered by ice lol. On communitas/immortal if it makes a difference.

 

[RoAdd]

I've seen the most worthless positions settled by the AI too, like 1 or 2 tundra sticking out by the ice caps seems to get settled, even though there isn't even fish and half the possible worked tiles are covered by ice lol. On communitas/immortal if it makes a difference.

That's probably already in the renaissance or industrial, at that point it may be valuable enough probably or he works specialist. I sometimes do it too if I have a colonist/pioneer and an island isn't settled.

 

[Deadstarre]

That's probably already in the renaissance or industrial, at that point it may be valuable enough probably or he works specialist. I sometimes do it too if I have a colonist/pioneer and an island isn't settled.

theres no question they will take the better land first. im actually slightly concerned any tampering with settle logic will hamper where theyre at now which is expansionist monsters, they play very well in this regard.

but, invariably, they do run out of good land to settle, and then will continue to take -any- land seemingly without regard to reasonable base yield / workable tiles or resources, or any other factor that should be making them think twice. its only a matter of time before every possible spot on the map that could be settled will be settled.

 

[tu_79]

but, invariably, they do run out of good land to settle, and then will continue to take -any- land seemingly without regard to reasonable base yield / workable tiles or resources, or any other factor that should be making them think twice. its only a matter of time before every possible spot on the map that could be settled will be settled.

This won't be bad if better settler units were used. As ilteroi claims, a city can be benefitial just in any location with enough buildings, which colonist may provide. So, if AI is using a colonist, I don't care that it takes the weirdest spots. After all, when colonists are online, city borders are on the third ring for the most part.

I don't know, maybe it's easier to code that settlers have to settle at least at 5 tiles to a foreign city, pioneers to 4 tiles and no restriction for colonists. But this is too hard coded, surely they don't like it.

 

[ryanmusante]

Is the reason why "http://www.picknmixmods.com/mods/CivV/UI/Military Log.html" isn't functional because this value is 0 by default?

<Row Class="6" Name="BALANCE_CORE_MILITARY_LOGGING" Value="0"/>

 

[lalalaprout]

Has there been a change to pantheons that I missed? I've picked up God of Commerce twice now and it doesn't give the +2 f/g in capital anymore.

 

[BiteInTheMark]

Has there been a change to pantheons that I missed? I've picked up God of Commerce twice now and it doesn't give the +2 f/g in capital anymore.

If I remember correctly, you have to found atleast one another city, and the effect will come instantly in your capitol and in the other cities, if you connect them.

 

[Owlbebach]

0. seen that a couple of times. they don't build settlers. root cause might be as simple as AI not researching pottery in time, which can happen through bad luck, because of random choices. gazebo topic, maybe solved already.

Seen this many times, last time in 2-27 version with Maya. Maya always suffers from it as they want to get to Kunas/Maths asap and then they get obsessed with Roman Forum. Other reasons are like a civ that rushes to Spearmen to fight Barbs and then gets obsessd with Zeus or something similar to this.
I think that all other Settler logic issues discussed here are not that important but this one is really frustrating cause sometimes it is gamebreaking. Leads to snowballing of neighbours of civ that does not want to settle. I feel like for this particular case it is okay to hardcode AI to research Pottery within first 3 techs.
Settlilng in bad locations sometimes is frustrating but considering how much stuff AI gets on higher difficulties - those cites are usually net positive for them.
Forward settling is not even an issue, human does it for and will always do so should the AI.

EDIT: want some challenge? Teach AI to play Venice

 

[ryanmusante]

EDIT: want some challenge? Teach AI to play Venice

No kidding, that is why I use mod "AI - No CS Annexers (v 1)". In my opinion, the AI's usage of Venice and Austria is broken. However, that is simply because of how truly unique those civs are and I believe it's best left to a human.

 

[Owlbebach]

No kidding, that is why I use mod "AI - No CS Annexers (v 1)". In my opinion, the AI's usage of Venice and Austria is broken. However, that is simply because of how truly unique those civs are and I believe it's best left to a human.

Whats the problem with Austria?