Nerf AI conquering City States?

[aimlessgun]

The liberation game can be something to do later on. It's pretty clear that there need to be more diplomatic options for city states though. Your options for keeping city states later in the game should not be:
A: dedicate 6 units per city state to surrounding them at all times
B: don't make alliances or DoFs and be ready to declare war even on your closest trading partners.

There desperately needs to be some kind of diplomatic "C: stay the hell away from my city state" option. If they are allied/dof with you and ignore your warning you can DoW them, and with no penalty.

 

[steveg700]

The liberation game can be something to do later on. It's pretty clear that there need to be more diplomatic options for city states though. Your options for keeping city states later in the game should not be:
A: dedicate 6 units per city state to surrounding them at all times
B: don't make alliances or DoFs and be ready to declare war even on your closest trading partners.

There desperately needs to be some kind of diplomatic "C: stay the hell away from my city state" option. If they are allied/dof with you and ignore your warning you can DoW them, and with no penalty.

Given that a declaration of friendship blocks declarations of war, seems reasonable at the very least that this could extend to suzerainties.

 

[Kruos]

The liberation game can be something to do later on. It's pretty clear that there need to be more diplomatic options for city states though. Your options for keeping city states later in the game should not be:
A: dedicate 6 units per city state to surrounding them at all times
B: don't make alliances or DoFs and be ready to declare war even on your closest trading partners.

There desperately needs to be some kind of diplomatic "C: stay the hell away from my city state" option. If they are allied/dof with you and ignore your warning you can DoW them, and with no penalty.

I would propose to link the diplomatic state of city state together with its suzerain : this way a war declaration to a CS would also trigger a war with its suzerain. With such principle, it would no longer be possible to declare wars to CS which have a suzerain allied with us. Seems logical, no?

 

[Ferocitus]

City-states aren't simply a player resource. As I've illustrated, this behavior can be very poor (or at least certainly puzzling) strategy for the AI. Are we just so jaded by the AI's limitations at utiliziing what's at hand effectively that we should see its willingness to immolate assets for both itself and the player as the next-best way to provide a challenge?

Early in the game, in the absence of a lot of information, getting an AI to
choose to nurture a CS for the long-term benefits rather than conquering it
immediately is not trivial.

It might seem obvious to human players to keep a CS that will help later on, but
what if x turns later you (or the AIs) find that Barbarossa is not far away? In
that case, choosing to keep the CS free in the hope of bonuses in the long-term
might be wasted time, effort and resources. Would you be just as critical of the
AI if it made an imprudent decision to not take the city quickly when it could?

With no warmonger penalties early in the game an easy city is not such a bad
(short-term!) option. Even razing it to keep it out of opponents hands might be
a good decision, That's one reason I prefer mod solutions, such as giving
CS walls and/or units. Players then have a choice in the type of game they want
right from the start.

 

[dexters]

AI will also take out city states someone else is suzerain of that they arent. If anything, losing all the invested envoys is a big setback.

Generally a city state that's in the path of their expansion often, but not always will have a target on their back. But Civ6 and (5) varies from game to game, and it's very hard to generalize since certain civs will be more agressive toward city states than others.

 

[Siesta Guru]

While I can sort of sympathize with the argument that everything that benefits the AI is something it should pick, the argument feels a bit flimsy when you realize that some behaviors would be highly advantageous to the AI but also aren't done for obvious reasons. The main one being that from a bots perspective, the best way to maximize the odds of winning is to immediately, in every single case, declare war on the human player the moment that player becomes visible. By having every player piledive the human asap, the odds of the human winning significantly decrease, giving bots a much better chance.
Another good tactic would be that if a human is attacking the AI, it could trade whatever city is being attacked away for a turn to a cooperating AI to knock all units out of the cities territory.

These kind of tactics are toxic, and have no place in a game that isn't just about winning. It makes no sense to have something like this in the game while there are so many glaring mistakes the AI makes that are all impact the AIs ability to compete far more.

It's definitely something that needs fixing. I do agree that the first step should be to make city states harder to conquer, or to give players a proper punishment for doing so, but I don't really think that will be enough. Even with my AI+ fix of having city states produce significantly more units, civs still attack them continuously, very frequently in a way that actually hurts the attacker far more than it could ever hope to win. The AI is just way overeager to attack city states in particular because there doesn't seem to be any mechanism in the code that tells it to value attacking city states differently until it becomes a suzerain. There is no good reason why a civ should value attacking a city state the same way as it values attacking a city of a player it's already at war with.

 

[azmundai]

whats funny is I use the mod that boosts the units in city states, and they still drop like flies, cuz you know, unit spam is pro tactic i guess ...getting kinda bored of always having to spam archers to kill ai civs to avoid playing 3 cities in a corner all game.

 

[Pietato]

I just loaded up a game of mine on Emperor, continents, large maps, standard settings and counted 10 city states on turn 328

Standard large maps start with 15 city states, so 5 got taken over.

Hard to judge from end-game totals, seeing as some would be liberated.

 

[Ferocitus]

While I can sort of sympathize with the argument that everything that benefits the AI is something it should pick

I wouldn't go quite that far. It's just very difficult to include sensible
long-term strategies into an AI. I hope you can find a way to mod
the game to suit players who don't like what Firaxis chose for
the base game. (IMO, the base game is just that - something that
can be moulded into many very different types of game.)

These kind of tactics are toxic, and have no place in a game that isn't just about winning.

Again, that's something that can be modded to suit players with different
expectations, and I hope better tools are released soon to help you and others
achieve that. However, there are a lot of people who don't mind playing as, or
against aggressive civs. Your proposed tweaks and modifications will please many
who want CS to defend themselves better, but it won't be as welcome by those who
like play as Fred Barbarossa with his agendas and bonuses against CS.

I agree with a lot of what you say, but IMO all you are doing is making
arguments for mods that create different games that suit players with different
expectations, and all power to you. The inadequacies and glaring errors made by
the AI are separate serious issues, but aggression towards CS isn't one of them.

 

[Athmos]

Agreed but a player can utilise them better. Removing an option for gameplay at a level where the AI has huge advantages does make things harder. That’s all I am saying.

A player can also make better use of 1UPT and City planing. And indeed, AI was better at playing civ4 in these regards (although it was even worse in naval wars).

Frankly, this is not about the game being hard or not, it is about an element of gameplay simply disappearing. I don't know what was changed in the AI to make it so much more aggressive toward CS, I guess the devs wanted emergencies to trigger or something, but they overdid it somehow.

Not only protectorate wars casus belli appears much too late (civic service), but it is only an a posteriori reaction, no way to deter AIs to attack CS in the first place.

I think city states should start with as many warriors as the AI, or maybe just ancient walls.
The player and AIs should be able to pledge to protect a CS for 30 turns, during which war declaration with the CS would be a War declaration against the protector as well - I don't see how an ability to pledge protection on a city state would harm the AI. They could pledge to protect CS from the player or each other too.


In my last game, all 6 city states (Lisbon Muscat Kumasi Antananarivo Brussels Hong Kong) on my starting continent were conquered in the first 35 turns, I could have at most gained one suzerainship in that time If I had pushed for the mysticism envoy. Not that gaining suzerainship would have brokered peace or changed anything anyway, but maybe it could have triggered an emergency ? Do we have some data about emergencies triggers ?

 

[historix69]

... The main one being that from a bots perspective, the best way to maximize the odds of winning is to immediately, in every single case, declare war on the human player the moment that player becomes visible. By having every player piledive the human asap, the odds of the human winning significantly decrease, giving bots a much better chance.
...

Should the AI know which player is human and which is AI? -> I don't think so. The AI should treat all other players as players, possible enemies and allies. If a player (human or AI) tries to win too much, the other players may team up to stop him, also depending on individual interests.


Regarding the conquest of CS :
I have less problems with AI conquering CS, but I don't like AI removing CS permanently from the game by razing a CS or by assimilating them (like Austria in Civ 5). There should be an option for players to forbid CS-Elimination by rule, similar like the rule for player capitals.

 

[Ferocitus]

Frankly, this is not about the game being hard or not, it is about an element of gameplay simply disappearing. I don't know what was changed in the AI to make it so much more aggressive toward CS, I guess the devs wanted emergencies to trigger or something, but they overdid it somehow.

Not only protectorate wars casus belli appears much too late (civic service), but it is only an a posteriori reaction, no way to deter AIs to attack CS in the first place.

That's their vision for the base game and how it should work.
Clearly, you and many others don't like it; I and many others don't mind it.
I loathe the maps they ship as standard, most people seem happy with them.

I think city states should start with as many warriors as the AI, or maybe just ancient walls.
The player and AIs should be able to pledge to protect a CS for 30 turns, during which war declaration with the CS would be a War declaration against the protector as well - I don't see how an ability to pledge protection on a city state would harm the AI. They could pledge to protect CS from the player or each other too.

I don't like your proposals for the base game.
So, how are you going to keep the majority of us happy if it's not through the
use of mods?

 

[Athmos]

(...)
I don't like your proposals for the base game.
So, how are you going to keep the majority of us happy if it's not through the
use of mods?

Well, as I understand it Ferocitous, from what you said in another thread :

I only play marathon pace with 33 civs and 28 CS

So I'm sure how relevant your opinion actually is regarding the "base game".

 

[Ferocitus]

So I'm sure how relevant your opinion actually is regarding the "base game".

Why should I take your opinion seriously when you just want your changes
to be accepted and for others to get mods to make the game what
it is now with respect to CS?

The question still stands: how are you going to keep the majority content
without resorting to mods?

 

[Athmos]

Why should I take your opinion seriously when you just want your changes
to be accepted and for others to get mods to make the game what
it is now with respect to CS?

The question still stands: how are you going to keep the majority content
without resorting to mods?

I don't see where I'm making more assumption about the majority's opinion than you.
Actually, I don't think I ever said my opinion was the majority's, because I simply don't know if it is.

I'm arguing that the current way the AI is behaving (at least in Deity on all-standard continent maps since those are the settings I'm using) is making a game sub-system almost irrelevant, and I'm suggesting that the balance of the benefit of peaceful interaction vs conquest toward city states should be altered as the game is currently too stacked toward aggression.

I don't have a special communication channel with Firaxis you know, you don't have to feel threatened by my opinions or ideas.

 

[Ferocitus]

I'm arguing that the current way the AI is behaving (at least in Deity on all-standard continent maps since those are the settings I'm using) is making a game sub-system almost irrelevant, and I'm suggesting that the balance of the benefit of peaceful interaction vs conquest toward city states should be altered as the game is currently too stacked toward aggression.

Yes, just another opinion like mine and from others. You can use
mods to get closer to what you want without changing the game
as its shipped now.

I don't have a special communication channel with Firaxis you know, you don't have to feel threatened by my opinions or ideas.

Huh? Irrelevant at best.

 

[bbbt]

I do wish the AI had more variety in personality in thth (and many other) respect. I.e. when I meet Poundmaker and he's taken over all his neighboring city states, my first thoughts are basically 'Well this is exactly the sort of thing the chief was complaining about'.

 

[Ferocitus]

I do wish the AI had more variety in personality in thth (and many other) respect. I.e. when I meet Poundmaker and he's taken over all his neighboring city states, my first thoughts are basically 'Well this is exactly the sort of thing the chief was complaining about'.

A clever line for every occasion would be nice.

If you were playing as Teddy, he could kill you immediately while repeating
your pre-presidency quote:
"I don’t go so far as to think that the only good Indian is the dead Indian
but I believe nine out of every ten are, and I shouldn’t like to inquire too
closely into the case of the tenth."

 

[Athmos]

Huh? Irrelevant at best.

Well, I stated that as you are constantly framing the discussion in terms of how the game should be and how others should just use mods :

Yes, just another opinion like mine and from others. You can use
mods to get closer to what you want without changing the game
as its shipped now.

As far as I know, we are simply discussing aspects of the game here, not making decisions that Firaxis would be constrained by. Anyway, all the horses here have been beaten to death multiple time over, I'll move on to see if/how the current system can be gamed to use envoys to generate emergencies, I suspect that might be the most profitable use of CS in the current version.

 

[TheMeInTeam]

On Emperor some are always left. deity is supposed to be tough.

"Deity is supposed to be tough" is a non-argument in the context of this thread's discussion. It doesn't address the complaints made, and it doesn't build a supporting case for the behavior or design ease. Can't we move past that?

On deity, 2-3 city states down by turn ~20 (on standard) is pretty common. Does that significantly contribute to deity's difficulty? I'm not convinced. Does it centralize the game? Yes. Do some of these often go to the player? Also yes.

City states should scale like AI civs, so they can be conquered but aren't settler-steals for everyone.