Real Strategy (AI)

Real Strategy (AI) 2.3.1

@alexu4326 Wars depend on diplomacy, military strength, geo-situation, and many other things. There is no parameter „minimum number of wars per era”.
You can have peace almost all the time and you can have wars and slaughter almost all the time. Anything in-between is ok.
 
I play against AI Canada on a duel map, Canada has 15 cities - and makes ca. 1100 tourism/turn in the atomic age - how can he do so much...?
(OK I play with mods, but still...)
 
I play against AI Canada on a duel map, Canada has 15 cities - and makes ca. 1100 tourism/turn in the atomic age - how can he do so much...?
He is just playing well...?
RST doesn't add any bonus, doesn't change any civ's ability, etc.
 
Hello all. First, great thanks to Infixo - great idea, and a very decent realization! From my point of view, AI has become much smarter and less dependent on just quantity, added by high in-game difficulty.
Second, I got a question. What exact parameters AI looks at, when choosing a diplomatic victory strategy? Except the number of diplomatic VP?
 
He is just playing well...?
RST doesn't add any bonus, doesn't change any civ's ability, etc.
Perhaps I am not expert on how to make much tourism... I have lots of wonders, walls, I have theaters, yet I make 230 tourism...
Canada also makes 900 culture/turn, I make 210...
(Again, it may be the many mods I have on...)
 
Second, I got a question. What exact parameters AI looks at, when choosing a diplomatic victory strategy? Except the number of diplomatic VP?
- leader's "affinity" to go diplo -> important ealy game
- current gov, policies and suzerained city states
- items acquired during the game - each has a flavor that says how much it helps - wonders, great people, beliefs
- number of diplo points
- favor per turn
 
Thank you. And one more question - is it possible to make the AI more predisposed to certain government form, based on victory strategy, era and so on? I've seen in your code, that you tried prevent AI from studying civics related to some governments, but is it the only solution?
 
And one more question - is it possible to make the AI more predisposed to certain government form, based on victory strategy, era and so on?
Not directly. See below.

I've seen in your code, that you tried prevent AI from studying civics related to some governments, but is it the only solution?
The only way I know :) AI will eventually culturvate all civics and choose the government for its liking. I just try to push it into the first one most suitable for a victory by giving hints about civics.
It is working (to some extent) because AI almost always changes government to the one with more slots once it is available. The only exception I've seen is Theocracy which has such strong bonuses that civs going for religious victory keep it a bit longer.
So, making it culturvate a specific civic with a gov is an indirect way of getting that gov.
 
Thanks again. But I have one more question - how do the pseudoyields, connected with city attack (like PSEUDOYIELD_CITY_BASE) actually work? Do they mean the general AI's desire to declare war and capture the city, even when AI is at peace? Or it's a tactical preference - which city should be attacked first during the war?
In my games I had situations, when AI had a large army, declared war and then didn't attack any enemy city, basically walking around and dying from cities' and encampments' ranged attacks. Maybe it was because he lacked siege units but I'm not sure. Is there a way to make him take objectives any better?
I am playing at King/Emperor difficulty if it matters.
 
how do the pseudoyields, connected with city attack (like PSEUDOYIELD_CITY_BASE) actually work?
There is a post at the begining of the thread that explains that in details.
Do they mean the general AI's desire to declare war and capture the city, even when AI is at peace? Or it's a tactical preference - which city should be attacked first during the war?
They are used to caclulate the value of the city, which is then used as a base for things like what city to capture, how quickly the units will be produced. The desire for war is probably somewhat affected, as one of factors.
In my games I had situations, when AI had a large army, declared war and then didn't attack any enemy city, basically walking around and dying from cities' and encampments' ranged attacks. Maybe it was because he lacked siege units but I'm not sure.
AI certainly attacks the city once it gets its units in close proximity. If you see behavior like you describe, then this could mean:
- still producing units and e.g. lack of siege units could stop the op as they are needed to capture a walled city
- still moving to a target city i.e. the target city is not always the closest one
- not close enough to the target - if the op is not close enough, it will only defend itself and attack enemies’ units
- unit deadlocked (may happen if there are really many units in the op)
- too many units were lost and ai is withdrawing and looking for a new target
You would need to check behavior tree and operations logs to actually find out what happens. Without that, just speculations.
Is there a way to make him take objectives any better?
I would like to know myself :) any ideas?
 
So funny, because the usual modification on the AI is better than the one presented by the game developers (Firaxis) I really don't understand why so few people use it.
 
After 160 turns of peace i have to ask government diference is not anymore a way to anger the AI ?
I am with autocracy and all my neighbors have oligarhy but nobody have a negative reputation agaist that.
Or that diference will be visible betwen marchet republic and monarhy and so on ?
 
So funny, because the usual modification on the AI is better than the one presented by the game developers (Firaxis) I really don't understand why so few people use it.
It has 6000+ subs on Steam, so it is not bad. CF downloads will always be much smaller :)
But spreading the word always helps (e.g. reddit). More users, more feedback, more ideas.
 
After 160 turns of peace i have to ask government diference is not anymore a way to anger the AI ?
I am with autocracy and all my neighbors have oligarhy but nobody have a negative reputation agaist that.
Or that diference will be visible betwen marchet republic and monarhy and so on ?
Isn’t it only for Tier3+ govs?
 
It has 6000+ subs on Steam, so it is not bad. CF downloads will always be much smaller :)
But spreading the word always helps (e.g. reddit). More users, more feedback, more ideas.
No, I think that there should be MUCH MORE. At least several tens of millions. Apparently you haven't played much with AI. Do you understand? This is too little. I do not in any way exaggerate, In all seriousness. I can share with the Russian people who do not know about this mode. You will not mind? Only when and which version. I have long wanted to ask a question. Will you update this resource in the future? Or you can say with confidence - At the moment it is fully operational (does not need to be improved at the moment) There is such a system (for the general public) in which it is better to understand everything at once. Apparently, you perceive yourself as a regular developer of modifications, but this completely changes the game! (I am afraid of the inscription "Test version") You should have more followers.Better try to perceive yourself .... as a developer firaxis.
 
Last edited:
@Exletrum Thank you for such an encouraging opinion :)

As for my plans. I've been waiting for a patch to be released to make a fine-tuning for the Version 2.0, sometime mid-April most likely.
After this, I suppose I will be releasing some balancing updates, depending on comments and feedback I will get, also from my own test games. RST is very much about balancing and this takes a lot of time and needs a lot of feedback.

RST is language-independent. There is not a single text that would require translation or anything because it is entirely behind-the-scenes type of mod. So yes, you can definetely recommend it in the Russian community.
 
Just a few words on my game I had in the last few days, with the usual many mods:


Duel map with 4 civs, I am Cree, AI: Canada, Germany and USA, smoother immortal level, quick speed (330 turns)

Canada is the big guy. I got 7 cities, Canada got 8 and then another 8 in wars…
Me and Germany had religion, I have won RV in the industrial era. I played on after that.

I tried to make culture and tourism, but even with many wonders and a healthy number of culture districts, I can make only ca. 200/turn in the endgame, while Canada makes 1100, and the same is with tourism (same rates). So Canada won CV in ca. turn 260.
Maybe I should have started a war against Canada earlier…

Edit:
Canada of course leads in tech too: ha has 62, I have 52... so war is not easy :)
 
Last edited:
Since you have made best UI mods there is, maybe you have tinkered with the music that plays in the background too? It no longer plays the music native to the civilization, but all kinds of tracks at random. Maybe you could have any idea why that is?
 
Back
Top Bottom