Do something, anything for the AI in the 867 start date - even by giving the AI the ability to outright cheat!

[AlipheeseXV]

Well, well.

After playing many games in the 867 start date with the 1.4.X updates, where I first believed nothing changed for the AI, I slowly changed my mind... up to the point that I really have to pity the AI!
Like seriously, pdox, what is your goal with the 867 start date!?
As it is now, I can start as a count and in like 2/3 of the starting locations on the map have an empire around the year 900... and then end the game, as aside from the byzantine Empire like no threat at all is left! Not even crusades as how "dumb" the crusade AI is.
Before the 1.4.X updates the AI managed to create at least some threatening realms:
- Have a Muwalladi Hispania? No chance now. Muwalladi may still spread, but only due to all the Ummayyad Lords ally each other
- Have Francia form? Was like 50% before 1.4.X, now I've seen it once, thats all
- Have at least a Germania as Empire? No chance in hell with all the pressure going on in that region
- Have a Scandinavia by the AI? Also no chance in hell
The current state of the game in the 867 start date is: "I want to be an independent count!!!111elf"
That's it.
Aside from the Byzantine Empire with in built Promigeniture and to some degree the Abbasids (Arabian Empire), nothing will be built... not even medium sized Kingdoms, as vassals in any newly formed AI Kingdoms just want their independence or at least do some "King/Queen Roulette" until the kingdom collapses...

So, pdox, I question you: What is your goal with the 867 start date?

I personally love the 867 start date for two things:
- I like to form something as independent count/duke and not have to start inside of Kindoms/Empires
- I like to revive "old things", like Hellenic Roman Empire, which the 867 start date fits much, much more
But with the poor, like really poor handling of the AI and realms ever since the 1.4.X updates, like mentioned above, almost nothing forms and most already formed realms will sooner then later crumble.
So, buff the AI in the 867 start date!
I don't even care if you enable the AI to outright cheat - as long as the AI will finally be able to create and most importantly "hold those created realms together"!

 

[Deshiba]

I'm reading a, quite detailed, description of what problem you are facing since a particular update. And though I don't disagree with the problem you are voicing, I am seeing a lack of actual suggestion. It's more of a request to 'do something, anything, to fix this problem (please)'. Which would fit better in the forums main-space (imo)

---

That aside, let's get in to the 2 things you like

- I like to form something as independent count/duke and not have to start inside of Kindoms/Empires

There's loads of those counts and dukes around if you look at the map. There's also

1. The most prominent one I have enjoyed is any of the irish dukes and forming Ireland, quite partial to Ormond myself.
2. Then there's the Russian and Baltic ones, a lot of those are single counties or dukedoms without formed kingdoms (yet)
3. There's also most of africa. Which could form africa
4. Scandinavia has a couple of counts/dukes. which could form scandinavia
5. In the Caucasus there's some. (not use what they could form)
6. And finally there's india/asia . (don't really care what they could form)

- I like to revive "old things", like Hellenic Roman Empire, which the 867 start date fits much, much more

This is just flat out possible.

---

AI-behaviour changed during patches so something broke in regards to forming and retaining titles. Seems like a bug.

 

[AlipheeseXV]

I'm reading a, quite detailed, description of what problem you are facing since a particular update. And though I don't disagree with the problem you are voicing, I am seeing a lack of actual suggestion. It's more of a request to 'do something, anything, to fix this problem (please)'. Which would fit better in the forums main-space (imo)

---

That aside, let's get in to the 2 things you like

There's loads of those counts and dukes around if you look at the map. There's also

1. The most prominent one I have enjoyed is any of the irish dukes and forming Ireland, quite partial to Ormond myself.
2. Then there's the Russian and Baltic ones, a lot of those are single counties or dukedoms without formed kingdoms (yet)
3. There's also most of africa. Which could form africa
4. Scandinavia has a couple of counts/dukes. which could form scandinavia
5. In the Caucasus there's some. (not use what they could form)
6. And finally there's india/asia . (don't really care what they could form)

This is just flat out possible.

---

AI-behaviour changed during patches so something broke in regards to forming and retaining titles. Seems like a bug.

I know that - with a custom ruler even far more things are possible.
I stated why I like the start date and play it exclusively. Though should had mentioned that. Also, maybe I became "too good/accustmed" to that start date? Who knows...

My gripe is, the AI suffers badly, like really badly ever sine 1.4 hit the game. When there are only 2 AI realms on the entire map which can somehow manage all this factions, while the rest breaks down to dukes or even single county counts, it is just flat bad game design.
That's why I'm mostly asking for "anything" - but you can boil it down to "buff the AI in 867, even by letting the AI cheat".

 

[Deshiba]

That's what I figured so far. All I'm saying is that it isn't a suggestion as much as it is a request to fix a problem (bug). I'd file it as such if you did not already do so.

 

[AlipheeseXV]

That's what I figured so far. All I'm saying is that it isn't a suggestion as much as it is a request to fix a problem (bug). I'd file it as such if you did not already do so.

Don't get me wrong, I understand what you suggest.
I even thought about it before going here - but, well, reporting the "dumbness and pity state of the AI in the 867 starting scenario" as a "bug" seemed more wrong then coming here.
At least from my point of view.

 

[Deshiba]

Don't get me wrong, I understand what you suggest.
I even thought about it before going here - but, well, reporting the "dumbness and pity state of the AI in the 867 starting scenario" as a "bug" seemed more wrong then coming here.
At least from my point of view.

Fair enough.
I consider this more of a bug since it's functionality that broke through updates, rather than a suggestion for new or different functionality.
In case of doubt I'd do both and then cross reference them to each other in the comments. As for a bug report title "AI rarely (re)forms or holds realms since 1.4"

I've had a game (yesterday) where France fell apart and split into France, Aquitaine and Bourbon (Burgundy was part of Italy). 100 years later they were still there, france never reformed. Not sure if me playing as Brittany was the cause of this, but they seem to not use the casus belli to go to war and reform the realm when it falls apart. So it's probably something in that AI behaviour that broke.

 

[AlipheeseXV]

Fair enough.
I consider this more of a bug since it's functionality that broke through updates, rather than a suggestion for new or different functionality.
In case of doubt I'd do both and then cross reference them to each other in the comments. As for a bug report title "AI rarely (re)forms or holds realms since 1.4"

I've had a game (yesterday) where France fell apart and split into France, Aquitaine and Bourbon (Burgundy was part of Italy). 100 years later they were still there, france never reformed. Not sure if me playing as Brittany was the cause of this, but they seem to not use the casus belli to go to war and reform the realm when it falls apart. So it's probably something in that AI behaviour that broke.

I would rather do a cross reference like I did with my suggestion about "let the AI finally join in holy wars", were I first asked in the normal forum if "anyone ever seen the AI actually help in holy wars (aside from alliances through marriages)" and then made a suggestion thread - though there are already some threads which point out the pity state of the AI, I even wrote something in one or two of them, so I see no need atm for that.

 

[Matihood1]

I don't even care if you enable the AI to outright cheat - as long as the AI will finally be able to create and most importantly "hold those created realms together"!

I get your point but screw this logic with a 10-meter long stick. The whole "we can't make the AI good so we're just going to make it cheat" crap is the worst possible practice a developer of a single player game can do.

 

[AlipheeseXV]

I get your point but screw this logic with a 10-meter long stick. The whole "we can't make the AI good so we're just going to make it cheat" crap is the worst possible practice a developer of a single player game can do.

I'm very open for other suggestions - may even open a "collecting ideas thread" about this in the normal forum section later.

 

[MatthewP]

In EU4 there are higher difficulty settings which give the AI bonuses to make it stronger and more stable. This is great; I now only play on very hard and it’s extended the life of the game for me enormously. I have no idea why these don’t exist in CK3. Maybe they just wanted to iterate on the game a bit to figure out what the right bonuses were?

I realize this is more of a bandaid than a fix, but it’s a pretty good bandaid that can help with a wide variety of AI problems that will probably never get a real fix.

 

[Matihood1]

I'm very open for other suggestions - may even open a "collecting ideas thread" about this in the normal forum section later.

I mean, there isn't much that can be suggested imo. The devs need to polish the AI and that's that.

 

[InsidiousMage]

The devs need to polish the AI and that's that.

Aside from the fact that it isn't that simple to actually do. All of the complaints about the CK3 AI are all basically the same for the AI for Civilization 6. The fact that two successful franchises from successful companies make it's pretty clear that there are deeper issues with designing AI. These are complex games and making an AI that can play them as well as a human are probably a couple of decades away. Civ6 gives the AI percentage based modifier boosts based on the game difficulty level and it does help the AI keep up the player to a degree and even makes them dangerous to the player in the early game.

 

[AlipheeseXV]

I mean, there isn't much that can be suggested imo. The devs need to polish the AI and that's that.

Well:

Let AI vassals/independent rulers finally join normal holy wars

Okay, we all know the following situation: The player wants to declare a holy war against an hostile or evil faith neighbour. The "Declare War" window shows which county/duchy/once-in-a-lifetime-kingdom title you will fight for and the cost to...

forum.paradoxplaza.com

That's something that would make conquests much harder, therefore even the AI wouldn't "snowball".
Also, it would enable the AI to finally aid their liege!
So factions wouldn't be by default only antagonistic - AI chars with high opinion of their liege would finally be of use rather then just having them allied to prevent factions at all.

 

[Deshiba]

higher difficulty settings which give the AI bonuses to make it stronger and more stable.

That's indeed a cop-out to making the AI better. And it also doesn't address the issue of AI not forming and/or keeping titles.
If an AI were to actually form a France, Spain or Scandinavia empire, that would keep it from splitting into smaller kingdoms. However this rarely happens which makes 867 not much of a challenge.

That's something that would make conquests much harder

Allowing non-allied AI to join holy wars would make the game 1 big holy war depending on how you restrict it. If you look at the map, what if ALL Christian faiths were to war against ALL muslim faiths in 867 because 1 ruler declared war against another ruler? Forever war.

And if you do not include all Muslims, what if Muwalladi declared holy war on just a single Catholic ruler in Spain? Never gonna have Muslim Spain again.
There's so many ways it could break the game that you'd need to be really careful of what parameters you'd use.

---

The suggestions in here so far don't actually make the AI better at managing their own realm and/or forming and keeping titles alive.

 

[MatthewP]

That's indeed a cop-out to making the AI better. And it also doesn't address the issue of AI not forming and/or keeping titles.
If an AI were to actually form a France, Spain or Scandinavia empire, that would keep it from splitting into smaller kingdoms. However this rarely happens which makes 867 not much of a challenge.

Im not sure why you say that. You could easily give the AI vassal opinion bonuses, that alone would make them more stable. And being more stable would make them more likely to form higher titles. Note this is just one example, certainly not the only way to make them stronger and more stable with simple bonuses. Of course with partition it’s always going to be tricky getting empires, but that’s sort of by design.

Allowing non-allied AI to join holy wars would make the game 1 big holy war depending on how you restrict it. If you look at the map, what if ALL Christian faiths were to war against ALL muslim faiths in 867 because 1 ruler declared war against another ruler? Forever war.
View attachment 745236
And if you do not include all Muslims, what if Muwalladi declared holy war on just a single Catholic ruler in Spain? Never gonna have Muslim Spain again.
There's so many ways it could break the game that you'd need to be really careful of what parameters you'd use.

---

The suggestions in here so far don't actually make the AI better at managing their own realm and/or forming and keeping titles alive.

This is how it worked in ck2 and it was not an issue. It was restricted by geography and possibly other things. Only the defender could get help from non-allies. The scenario you suggest of every holy war being like a crusade did not happen. It did make conquest by holy war less trivial. Similar rules could easily be used in ck3.

 

[Deshiba]

Im not sure why you say that. You could easily give the AI vassal opinion bonuses, that alone would make them more stable. And being more stable would make them more likely to form higher titles. Note this is just one example, certainly not the only way to make them stronger and more stable with simple bonuses. Of course with partition it’s always going to be tricky getting empires, but that’s sort of by design.

I'm saying that scaling difficulty with opinion bonuses is a cop-out because it's true. Giving them scaling bonuses depending on the difficulty setting doesn't make the AI inherently better at managing a realm. In fact this only becomes the case at higher difficulty levels when arguably you want the AI to be able to form and keep titles at any difficulty. Answer yourself the following question if you doubt my statement.

Why would the empire of France, Spain or Scandinavia only form on hard mode?​

This is how it worked in ck2 and it was not an issue. It was restricted by geography and possibly other things. Only the defender could get help from non-allies. The scenario you suggest of every holy war being like a crusade did not happen. It did make conquest by holy war less trivial. Similar rules could easily be used in ck3.

But you do see my concern about it, right?
Even when only the defender gets to call in people from a religion AND it's restricted to geographical regions there's still an issue in regions like Spain. In order for there to be a victor between Christianity and Muslims all their regions have to be united to be strong enough to invade the other religion. With the tendency for AI to be unable to keep titles together this will literally never happen, so there will be no Spanish empire by AI ever. The only thing I see this doing is make it even more player exclusive.

Again, this doesn't fix the OP issue of titles not being formed or held, it just makes it less likely for AI to Form larger titles like empires.

 

[Ezumiyr]

The reason why realms keep getting split isn't because "the AI is weak". There's some confusion in this thread between AI realms and how the AI of CK works.
If you just buff every AI realm, nothing will change, because vassals will also be stronger, and the issue is the faction system.
You could change the AI of CK so it's less prone to join factions. It would indeed solve the issue... somewhat.
But the real cause of the issue isn't the AI (and even less how weak AI empires are), it's balance and features. Factions are too monodimensional, they only weaken your realm, and won't accept concessions. It's all or nothing.

Then there's also the vision of the game. OP thinks that empires should be formed by the AI, and not only the HRE, but also Scandinavia... The thing is that empires are rare historically, and they often crumbled. Should the AI really be able to forme huge nonsensical empires just because it suits your playstyle? I completely disagree. Or at least, it should be an optional game rule - by default, empires should be really rare... but kingdoms should also be more stable.

 

[Deshiba]

The reason why realms keep getting split isn't because "the AI is weak". There's some confusion in this thread between AI realms and how the AI of CK works.
If you just buff every AI realm, nothing will change, because vassals will also be stronger, and the issue is the faction system.
You could change the AI of CK so it's less prone to join factions. It would indeed solve the issue... somewhat.
But the real cause of the issue isn't the AI (and even less how weak AI empires are), it's balance and features. Factions are too monodimensional, they only weaken your realm, and won't accept concessions. It's all or nothing.

Then there's also the vision of the game. OP thinks that empires should be formed by the AI, and not only the HRE, but also Scandinavia... The thing is that empires are rare historically, and they often crumbled. Should the AI really be able to forme huge nonsensical empires just because it suits your playstyle? I completely disagree. Or at least, it should be an optional game rule - by default, empires should be really rare... but kingdoms should also be more stable.

Currently France splits in 3 kingdoms, which then splits into individual dukedoms. It's real easy for the player to just wait out the big divide and then gobble it all up piece by piece. The same seems to happen in every major empire title on the 867 start date.

 

[AlipheeseXV]

The reason why realms keep getting split isn't because "the AI is weak". There's some confusion in this thread between AI realms and how the AI of CK works.
If you just buff every AI realm, nothing will change, because vassals will also be stronger, and the issue is the faction system.
You could change the AI of CK so it's less prone to join factions. It would indeed solve the issue... somewhat.
But the real cause of the issue isn't the AI (and even less how weak AI empires are), it's balance and features. Factions are too monodimensional, they only weaken your realm, and won't accept concessions. It's all or nothing.

Then there's also the vision of the game. OP thinks that empires should be formed by the AI, and not only the HRE, but also Scandinavia... The thing is that empires are rare historically, and they often crumbled. Should the AI really be able to forme huge nonsensical empires just because it suits your playstyle? I completely disagree. Or at least, it should be an optional game rule - by default, empires should be really rare... but kingdoms should also be more stable.

Then, well, we can just give the 867 start date to the byzantines...

It's not even about playstyle - I mentioned empires I've seen on a somewhat regularly basis build.
As far as I find your idea about "rare empires" interesting, it will be super boring to watch EVERY single time in the 867 start date the byzantines mopping up like everything.
With rare empires, there will be nothing as opposition to byzantion. Even "stable kingdoms" aren't really a thread to them if led by the AI. Only player intervention will stop the byzantines.

And for "stable kingdoms": The AI is far from managing even small kingdoms. I always play with custom kingdoms on. And, well, the AI seems at least more able to form them - but they normally vanish in the timespan of around 30 years.
That's it. Kingdom titles "lying around" aren't formed in 90% of the time.
So my initial point stands: For now the AI is all about "I want to be an independent count!"
Nothing else.

 

[MatthewP]

I'm saying that scaling difficulty with opinion bonuses is a cop-out because it's true. Giving them scaling bonuses depending on the difficulty setting doesn't make the AI inherently better at managing a realm. In fact this only becomes the case at higher difficulty levels when arguably you want the AI to be able to form and keep titles at any difficulty. Answer yourself the following question if you doubt my statement.

Why would the empire of France, Spain or Scandinavia only form on hard mode?​

But you do see my concern about it, right?
Even when only the defender gets to call in people from a religion AND it's restricted to geographical regions there's still an issue in regions like Spain. In order for there to be a victor between Christianity and Muslims all their regions have to be united to be strong enough to invade the other religion. With the tendency for AI to be unable to keep titles together this will literally never happen, so there will be no Spanish empire by AI ever. The only thing I see this doing is make it even more player exclusive.

Again, this doesn't fix the OP issue of titles not being formed or held, it just makes it less likely for AI to Form larger titles like empires.

Giving the AI an opinion Bonus would make it have better outcomes in managing a realm. Would it be “inherently” better? Maybe not, if you like. But who cares? Of course, you are right that Bonuses like this would only work when they exist. The truth is other posters are right when they say that having lots of empires form is pretty ahistorical. Which is why it’s perfect for a game/Difficulty setting that could make the game more fun for people who want it to play out that way.

On the holy wars, I agree that giving holy war defenders advantages doesn’t make it more likely for large empires to form. I still think it is a good idea, separately. Holy wars are too powerful.