The AI should surrender when all of their cities are at 1 HP.

[ElliotS]

As title. If an AI's cities are all at 1 HP (Or whatever healing after a turn of being at 1 HP is.) the warscore should be set to 100 and it should immediately capitulate.

Most of the time this will be it's last city, but sometimes you might get 2 or 3 to 1 HP.

It can be very annoying to forcefully vassalize a nation with few cities left, because you don't want to take them and break most of the infrastructure that you would be vassalzing them to leverage.

I think this is a straightforward and simple solution to a fairly common problem, any thoughts?

 

[Txurce]

As title. If an AI's cities are all at 1 HP (Or whatever healing after a turn of being at 1 HP is.) the warscore should be set to 100 and it should immediately capitulate.

Most of the time this will be it's last city, but sometimes you might get 2 or 3 to 1 HP.

It can be very annoying to forcefully vassalize a nation with few cities left, because you don't want to take them and break most of the infrastructure that you would be vassalzing them to leverage.

I think this is a straightforward and simple solution to a fairly common problem, any thoughts?

You're saying that, as a successful warmonger, you want to reap more spoils (and suffer less blowback) than you otherwise would, right? If so, I'm against that. I'd rather keep the present system, which hinders snowballing, in a very minor way.

 

[ElliotS]

You're saying that, as a successful warmonger, you want to reap more spoils (and suffer less blowback) than you otherwise would, right? If so, I'm against that. I'd rather keep the present system, which hinders snowballing, in a very minor way.

I think it's clearly closer to a bug than a feature. If an AI's cities are all at 1 health it should be willing to surrender. To not do so is death.

Moreover I've found (and the recently posted opinion guide) has shown you get far more blowback for vassals than finishing someone off. I think the play is to take the cities 98/100 times. Vassals are a boon later in the game when war weariness is super high and so is warscore. When it's not easy to push forward and take everything.

So for balance and basic logic I'd ask you to change your vote. If warmongering is too strong we can nerf it in ways that aren't the AI behaving irrationally.

 

[Txurce]

I think it's clearly closer to a bug than a feature. If an AI's cities are all at 1 health it should be willing to surrender. To not do so is death.

Moreover I've found (and the recently posted opinion guide) has shown you get far more blowback for vassals than finishing someone off. I think the play is to take the cities 98/100 times. Vassals are a boon later in the game when war weariness is super high and so is warscore. When it's not easy to push forward and take everything.

So for balance and basic logic I'd ask you to change your vote. If warmongering is too strong we can nerf it in ways that aren't the AI behaving irrationally.

I agree in the broad sense with this post, and am totally opening to changing my vote. You offered a very specific, self-serving reason in the OP, and in the second give an example where (I think) multiple cities are all at one health, which seems unlikely unless you're trying to game the system.

Opening up the conversation the way you just did in your second post makes a lot more sense to me: by discussing how vassalization as a whole should work. It might be better to lower the capitulation point. Personally, I know after 65 or so that it's a foregone conclusion the other civ is going to capitulate eventually. Now there is some question as to whether they'd do it before war weariness guts me. Lowering it to 70 or 75 may solve the problem by making vassalizing someone worth the diplomatic malus, because the cities aren't wrecked. They're also healthy enough to be able to recover and demand their freedom sooner than someone you've reduced to a heartbeat.

 

[Txurce]

I think it's clearly closer to a bug than a feature. If an AI's cities are all at 1 health it should be willing to surrender. To not do so is death.

Moreover I've found (and the recently posted opinion guide) has shown you get far more blowback for vassals than finishing someone off. I think the play is to take the cities 98/100 times. Vassals are a boon later in the game when war weariness is super high and so is warscore. When it's not easy to push forward and take everything.

So for balance and basic logic I'd ask you to change your vote. If warmongering is too strong we can nerf it in ways that aren't the AI behaving irrationally.

I agree in the broad sense with this post, and am totally opening to changing my vote. You offered a very specific, self-serving reason in the OP, and in the second give an example where (I think) multiple cities are all at one health, which seems unlikely unless you're trying to game the system.

Opening up the conversation the way you just did in your second post makes a lot more sense to me: by discussing how vassalization as a whole should work. It might be better to lower the capitulation point. Personally, I know after 65 or so that it's a foregone conclusion the other civ is going to capitulate eventually. Now there is some question as to whether they'd do it before war weariness guts me. Lowering it to 70 or 75 may solve the problem by making vassalizing someone worth the diplomatic malus, because the cities aren't wrecked. They're also healthy enough to be able to recover and demand their freedom sooner than someone you've reduced to a heartbeat.

 

[ElliotS]

I agree in the broad sense with this post, and am totally opening to changing my vote. You offered a very specific, self-serving reason in the OP, and in the second give an example where (I think) multiple cities are all at one health, which seems unlikely unless you're trying to game the system.

Opening up the conversation the way you just did in your second post makes a lot more sense to me: by discussing how vassalization as a whole should work. It might be better to lower the capitulation point. Personally, I know after 65 or so that it's a foregone conclusion the other civ is going to capitulate eventually. Now there is some question as to whether they'd do it before war weariness guts me. Lowering it to 70 or 75 may solve the problem by making vassalizing someone worth the diplomatic malus, because the cities aren't wrecked. They're also healthy enough to be able to recover and demand their freedom sooner than someone you've reduced to a heartbeat.

I was actually making this thread due to concerns from other threads, because G's mentioned it's easier to remember requested changes if made in the balance subform rather than version discussion. Hence why I had a short main post. Here they are:

I propose allowing soon-to-die AIs to capitulate to the victor without having to reach the magical war score number; I wanted to vassalize Venice but could not do so despite having brought the city itself to single-digit HP with my units surrounding it, all tiles pillaged, but without enough Venetian units to kill. Surely the beaten up single city civ should prefer life over death?

Also. I really miss seeing AIs using the demand function. They just never do it anymore. What's up with that?

Always? Shouldn't there at least be a small chance that they will rather fight till bitter end than submit?

Yeah, i had multiple games where i had to wait 30-40 turns not making peace deal and not attacking because i was waiting for Medieval Era to start. I knew that i will not be able to get that magic warscore 100 again because defeated AI will not have enough units for that

Which is why I dislike the medieval era for starting vassals since vassalage even existed in the classical era (See Ancient China for example where a number of "bloodline" states pledged loyalty to a stronger state that deserved the mandate of heaven and overthrow them when the mandate of heaven is lost to that stronger state. It's also really simple to change it to classical era as well, but I'm not one to ask for that.

P.S. if you want European examples, you could argue classical Greek city-states like Sparta and Athens(although to a lesser degree) used their allies like vassals as they taxed them and provided them protection in the same essence.

Hmm...you could always brutally murder him and all of his people like an uncivilized pre-Medieval Era brute?

In all seriousness, I agree, if their last city is about to fall, I think they should capitulate.

And the no-vassals-before-Medieval thing is relatable, I usually wait before going to war if I can to avoid that situation.

As for the comment about gaming the system: I don't think so. If you put in the extra effort to bring multiple cities to 1 HP rather than taking them, that would cause a player to call uncle, so the AI should too. It's not logical or fun for gameplay that declaring war on a civ with 1 city means you can't reasonably vassalize them, when that should be easy. Sometimes you want to trade them back their cities, which gives them a better chance to stage a comeback.

I'd be open to tweaks like a small % chance they won't surrender or the other ideas, but I want at least the simple tweak listed in the OP.

 

[Txurce]

I was actually making this thread due to concerns from other threads, because G's mentioned it's easier to remember requested changes if made in the balance subform rather than version discussion. Hence why I had a short main post. Here they are:










As for the comment about gaming the system: I don't think so. If you put in the extra effort to bring multiple cities to 1 HP rather than taking them, that would cause a player to call uncle, so the AI should too. It's not logical or fun for gameplay that declaring war on a civ with 1 city means you can't reasonably vassalize them, when that should be easy. Sometimes you want to trade them back their cities, which gives them a better chance to stage a comeback.

I'd be open to tweaks like a small % chance they won't surrender or the other ideas, but I want at least the simple tweak listed in the OP.

None of the posts you quoted mention salvaging buildings, the way you did in the OP. Most of them are about shifting vassalization from Medieval to Classical. They were actually the reason I said the discussion should be broader.

In my own games, I've often wished that vassalization were available pre-medieval. But I also appreciate how easy it is to reduce a neighbor to vassal status in the very early game -- pretty much all you have to do is make it Job 1 along with minimal expansion. I don't like making the game easier on myself like that and, after consideration, don't see any reason to make your game easier, either. The "effort" you put into reducing a civ to coma state shouldn't result in additional bounty for you.

My guess is that Stalker0 will have a "Going for Gold" thread on vassalization. Given that, I think this conversation will be superseded there.

 

[CrazyG]

I see ElliotS's point. I've basically won a war, but I don't want to rack up a bunch of warmonger score. I don't want your cities, but the only way to end the war is to take them, and I can't give them back.

This happens in history too, where a nation took some land, then gave it back after the peace deal. But its really difficult to sell cities back in this game.

 

[Enginseer]

That's why one of my test mod has a reinforcement mod where a city may pop a random city event of reinforcing troops if the HP falls below 20% which was inspired by the Siege of Vienna and other historical relief forces.

It's difficult to sell back cities because it is essentially player griefing when abused. (Conquer a city, destroy its infrastructure, and give it back as a deadweight city). And this has definitely happened in history which means the AI will only accept their lost strategically important cities back.

 

[ElliotS]

I see ElliotS's point. I've basically won a war, but I don't want to rack up a bunch of warmonger score. I don't want your cities, but the only way to end the war is to take them, and I can't give them back.

This happens in history too, where a nation took some land, then gave it back after the peace deal. But its really difficult to sell cities back in this game.

Yeah this. I feel the position that vassals should serve isn't always served, which isn't good the the vassal AI or fun gameplay in general. You're way more likely to come back if the never need to take your cities and give them back, because you don't lose all your buildings. It's just smarter to say mercy and rebel later.

It's difficult to sell back cities because it is essentially player griefing when abused. (Conquer a city, destroy its infrastructure, and give it back as a deadweight city). And this has definitely happened in history which means the AI will only accept their lost strategically important cities back.

That makes sense, which is why instead of making them unconditionally accept back cities they should accept vassalage before needing to have them all taken.

 

[ElliotS]

None of the posts you quoted mention salvaging buildings, the way you did in the OP. Most of them are about shifting vassalization from Medieval to Classical. They were actually the reason I said the discussion should be broader.

Mentioning buildings specifically? No. That's implied though, because otherwise you can just take them as puppets. You really are better off with a puppet than a super weak vassal both for diplomacy and actual benefit.

They can annoy you in a ton of ways if they hate you, and your enemies are very likely to hate you because you get a regular malus AND then they increase all hostile and war feelings for you by 10% for each vassal you have.

 

[Txurce]

I see ElliotS's point. I've basically won a war, but I don't want to rack up a bunch of warmonger score. I don't want your cities, but the only way to end the war is to take them, and I can't give them back.

This happens in history too, where a nation took some land, then gave it back after the peace deal. But its really difficult to sell cities back in this game.

You can give a city back as soon as it fully heals. That's longer than I want sometimes, but not prohibitive.

But that's not your main point. You're saying that it should be possible to end a war sooner than it is now. I agree, and suggested a lowering of the threshold -- one not based on city health, but war score. Am I missing something here?

Mentioning buildings specifically? No. That's implied though, because otherwise you can just take them as puppets. You really are better off with a puppet than a super weak vassal both for diplomacy and actual benefit.

They can annoy you in a ton of ways if they hate you, and your enemies are very likely to hate you because you get a regular malus AND then they increase all hostile and war feelings for you by 10% for each vassal you have.

This makes sense to me. I wouldn't reduce the vassal hostility, so we're looking for a way to exit out of a war without vassalage. There are certainly historical exceptions where a people won't give up, but the majority do when their army is beaten (never mind how well their cities are holding up). I mentioned above basing it on war score rather than city health. It may also be worth considering basing it on army size vs city number/size -- if your army strength drops below a threshold, you automatically surrender.

I like that much more than the process of lining up every city at 1hp. What do you think about this alternative?

 

[Recursive]

AND then they increase all hostile and war feelings for you by 10% for each vassal you have.

Increases the WAR and GUARDED approaches, actually. HOSTILE isn't affected (unusually, since most changes to WAR also affect HOSTILE).

There's also some multipliers after that, so it's not quite 10%, but it's still notable.

Controlling many enemy capitals, however, exponentially increases hate for you (and generally to get a civ to surrender capturing their capital is the best route).

 

[CrazyG]

But that's not your main point. You're saying that it should be possible to end a war sooner than it is now. I agree, and suggested a lowering of the threshold -- one not based on city health, but war score. Am I missing something here?

As Engineseer described, the issue is when I conquer your city, I destroy all the infrastructure before I can give it back, and I think you retain some warmonger score if you conquer then sell back the same city. A threshold change could work too I suppose.

 

[Gazebo]

In an ideal world you'd have a system like EU4 etc. where you 'occupy' a city until the war is over, and then exchange war score for cities...but that's never going to happen (and merely mentioning it will make you pine for what will never happen).

G

 

[ElliotS]

I like that much more than the process of lining up every city at 1hp. What do you think about this alternative?

I'm open to it, but it seems like G wants minimal changes on some parts of the mod so I wanted a more limited scope with less chance of affecting other stuff.

 

[HalfEmptyMug]

There are certainly historical exceptions where a people won't give up, but the majority do when their army is beaten (never mind how well their cities are holding up).

In more crowded games there are a few situations where fighting to the last breath might be more appealing, such as not wanting to join your side in a multi-front war where they might be conquered anyway, or being considered a backstabber by their former allies if they survive. Also, there is some merit in risking becoming a martyr, where they "know" completely conquering them will draw just enough attention from the rest of the world to make them begin fighting back, and they're already out of the race to win anyway. I remember many, many versions ago actually liking when the AI would declare war with their single city and was incredibly stubborn, forcing me to choose between dealing with them or maintaining global relations (though I get that's not for everyone).

I agree, and suggested a lowering of the threshold -- one not based on city health, but war score.

You're way more likely to come back if [they] never need to take your cities and give them back, because you don't lose all your buildings. It's just smarter to say mercy and rebel later.

I think the bigger problem is that the AI has a difficult time judging threat beyond what's happening in the moment, and adding an hp or warscore qualifier won't really solve that issue. If you've routed all of their military units and are besieging one of their core cities at a rate they can't hope to stop, they should be willing to negotiate assuming their other cores will probably be lost too — regardless of what hp it has. I'm not sure how difficult that would be to pull off though since that's pretty much been the fight with the tactical ai for some time now.

As it is, I'm pretty sure the ai does take into account current city threat and health, just not to the point where they can predict the eventual war score loss. Would it be possible to take into account the rate at which a city is losing its health per turn to predict its loss? And/or the per turn loss of units compared to their opponent? If that's not practically worth doing, then I would support a combination of ElliotS' and Txurce's solutions, along with more lenient liberation requirements (Since I've never been able to demand another civ release a vassal, nor anyone else mine).

TL;DR: I think the AI needs to get better at realizing when it's defeated beyond just warscore and health to vassalize, plus making it easier to be liberated so they can bounce back and not be gamed.

 

[ElliotS]

I think the AI needs to get better at realizing when it's defeated beyond just warscore and health to vassalize

This is the problem. As nice as it would be many changes would likely mean AIs that can win a war capitulate needlessly. That's why I'm asking for a very specific change with no real potential to backfire. An AI might pull out a win after getting routed at 50 war score, but not when ALL of it's cities are at 1 HP. That's undeniable tactical dominance. The only challenge I can see is making sure they go to 100% warscore with the right people, not letting anyone in the war at all swipe a free vassal from under the leader's nose.

 

[black213]

Thanks for making a thread for it @ElliotS
Voted yes, although it should be something more 'natural' rather than a certain figure. Like AI's evaluation of the situation; "Am I about to die? If yes, offer capitulation to victor (chance of fighting to the death depending on the boldness of the specific leader)"

 

[tu_79]

This is the problem. As nice as it would be many changes would likely mean AIs that can win a war capitulate needlessly. That's why I'm asking for a very specific change with no real potential to backfire. An AI might pull out a win after getting routed at 50 war score, but not when ALL of it's cities are at 1 HP. That's undeniable tactical dominance. The only challenge I can see is making sure they go to 100% warscore with the right people, not letting anyone in the war at all swipe a free vassal from under the leader's nose.

The problem with your suggestion is that it's not clear to state ingame that you have to reduce all of their walls to 1 hp in order to gain such status.
I see the same problem with razing cities increasing warscore. There's not a tooltip when hovering over the 'RAZE' button saying how much warscore the player is going to gain by killing the former people of the civ you are at war, isn't it?
So, how could you inform a player that, in order to force a vassallage to AI, you have to reduce all their cities to 1HP?

The clean solution is having its author to work on AI logic until it manages to take better decisions, but it takes time and work, and it has to merit it.

Related question, if you raze one of AI former cities, can you gain enough warscore as to force a vassallage?