Capitulation threshold?

[Woody1]

Ah, okay, the bouncing your units outside of his borders does sound like a problem.

The AI capitulating to another civ sounds like a smart thing for it to do, but it shouldn't delay your attacks on the vassal.

IMO, what should happen is that you should automatically go to war with the new master of that civ, and your units should stay where they are. (Except for any units in the master's territory, which should get bounced because you're now at war with him too.)

Also, the "you declared war on me" diplomatic penalty should be a hit on the master, not you. (I realize that humans don't get these penalty, but "you" could be another AI, since this doesn't have to happen to a human player.)

So there does seem to be a problem with how the capitulations are done. That has nothing to do with the threshold, but I guess the problem is seen more often now.

Thanks for the info.

IMO, the fix should be to force a war with the master. If he knew a war with you was the result, he may be less willing to accept a capitulation.

 

[Fistleaf]

Why not create 2 versions of the unofficial patch: one with 3.17 capitulation threshold, and one with 3.13 capitulation threshold? Let everybody play around with both versions for a while and then vote on which version to keep.

I, for one, prefer a lower capitulation threshold. My experience with 3.13 is that civs refuse to surrender when they are clearly beaten.

And for the historical Qin Shihuang, several states capitulated to Qin even when they can still fight on. The last surviving state of Qi especially, surrendered without even fighting. Qi was a large state with substantial resources and can probably withstand Qin's onslaught for quite a while.

 

[Ace of Spades]

Hello Woody,

the behaviour you proposed occurs when the AI civ becomes a voluntary vassal of another civ.

I have to admit that this is difficult to handle. Personally, I do think that the current capitulation behaviour, where you are forced to peace, is usually the most sensible thing.

Example: You have a defensive pact with Civ A, and Civ B declares war on you. The defensive pact steps in, and A declares war on B. Now you and A are at war with B. Suppose you are winning, and conquering their cities. B wants to surrender to you, but you refuse. Therefore, B surrenders to A.

If the default result was that this causes you to declare war on A, this would be a bit strage... since he was your former war ally. Of course you could create an example were it does make sense to go to war with A, but that's something you can always do manually as a player, while sueing for peace manually is much more difficult, not to mention the diplomatic penalties.

So I guess the game designers just thought that having your units bounced once is acceptable, and - considering the alternatives - I am inclined to agree.

Best Regards,
Ace

 

[Woody1]

Well, it would be nice to get a choice when that happens.

(1) "Declare peace with the new vassal, because his master is your faithful friend."

(2) "Declare war on the disrespectful new master of your sworn enemy."


It just seems odd that you're forced into (1). Let's say the US was fighting Iraq, and then Iraq decided to become a vassal to Russia. Does it then make sense that the US would automatically have to pull out all its forces from Iraq, before declaring war on Russia & Iraq? And would Russia accept Iraq as a puppet state if it was involved in a war with another country?

 

[Willem]

It just seems odd that you're forced into (1). Let's say the US was fighting Iraq, and then Iraq decided to become a vassal to Russia. Does it then make sense that the US would automatically have to pull out all its forces from Iraq, before declaring war on Russia & Iraq? And would Russia accept Iraq as a puppet state if it was involved in a war with another country?

But the situation you're describing here is what happens if Iraq were to voluntarily become Russia's vassal. If both countries are at war with Iraq, then as soon as Iraq capitulates, the war is essentially over, the common enemy of both nations has been eliminated. It's the same sort of thing that happend in WW2. The US and Russia didn't automatically become enemies just because they both controlled parts of Germany. Same thing with England and the US against Japan. It just wouldn't make sense for what you're suggesting to occur.

 

[kazapp]

Well, it would be nice to get a choice when that happens.

(1) "Declare peace with the new vassal, because his master is your faithful friend."

(2) "Declare war on the disrespectful new master of your sworn enemy."

It just seems odd that you're forced into (1).

I believe the real reason is to be able to keep the code clean.

I mean, you can always accomplish (2) by accepting (1) and then DoWing the new master...

Let's say the US was fighting Iraq, and then Iraq decided to become a vassal to Russia. Does it then make sense that the US would automatically have to pull out all its forces from Iraq, before declaring war on Russia & Iraq?

I think a far more probable scenario would be that the Americans would gnash their teeth, clench their fists, complain to the UN, but withdraw their troops peacefully, not wanting to risk a World War III with Russia over anything as insignificant as Iraq...

And would Russia accept Iraq as a puppet state if it was involved in a war with another country?

Which country? In your example, Russia was at peace when it accepted Iraq as vassal.


Anyway, the Iraq example is I think flawed for a completely different, and much more important, reason: the civs you see in any game of Civilization are all supposed to be major powers. Local countries, like Iraq, Canada or Denmark would not be represented as individual civs. In early eras, they might be represented by settled barbarians perhaps.

Read up on the partitions of Poland or the fall of the Ottoman Empire, and you'll realize that there's a separate set of rules of conduct that applies to Major powers. The Ottoman Empire might have been weak when it was on its last legs, but even the risk of it allying with any other major power was considered a severe setback for any Power involved in a war against it.

For your example to be really appropriate, you'd have to extend Iraq to something more, perhaps a "Pan-Arabic Bloc" - and even that might not be enough. In 1900 civilizations on a map of Europe would have been England, France, Germany, Austria, Italy, Russia and Turkey.

In 2000 civilizations on a Terra map would perhaps include the US, Russia, China, India, Brazil, and the EU. Not Iraq. Of course, if you run the map script for 50 civs, then this breaks down somewhat.

And if the US were at war with China or Brazil, and they were losing, and turned to the protection of India or the EU (!), sure as hell the US would have broken off the war immediately. Besides, after thinking about the military situation as the US Prez, you can always resume hostilities (DoWing India or the EU in this case, but going after the original target) after ten turns...

 

[Wolfshanze]

Why not create 2 versions of the unofficial patch: one with 3.17 capitulation threshold, and one with 3.13 capitulation threshold? Let everybody play around with both versions for a while and then vote on which version to keep.

I, for one, prefer a lower capitulation threshold. My experience with 3.13 is that civs refuse to surrender when they are clearly beaten.

Personally, I think the answer would be something inbetween 3.13 and 317 on capitulation threshold. Harder then 3.17, but easier then 3.13... I think you can find a happy medium here.

 

[DanF5771]

Has anybody investigated the code and found some hard numbers that could be compared?

 

[fizsh]

Slightly off topic, but in the same area. I just completed a game. The games started with 3.13 Bhuric's patch (1.11 I believe), got updated to Bts patch 3.17 then Solver's 2nd patch (not the current one). I would say at least 4 times during the game Japan renounced its protection from Mongolia and then accepted protection from Mongolia on the same turn, or rather the same in between turns for me. All this happened after the 3.17 patch. It doesn't bother me, I just treat it like it never renouncec in the first place, but it seemed odd. I am not a good player (play on Noble, win half the time) and am fairly unobservant when it comes to the AI agreements. But this one caught my eye. Mostly because I wanted to attack Japan and the first time I saw the renouncement (because of the sound that happens), I thought yeah, attack. But then noticed the acceptance of protection when I looked to closer.

 

[Woody1]

Yeah, the AI does seem to make and break voluntary vassalage very easily now. Not sure if that's good playing by the AI, but it doesn't appear to negatively affect the AI.

 

[DanF5771]

Could it be possible that the Vassal decides to break free in his turn and then the Master bribes him back into his Vassalage on his turn during the same IBT?

 

[fizsh]

Could it be possible that the Vassal decides to break free in his turn and then the Master bribes him back into his Vassalage on his turn during the same IBT?

That is the only thing I can think of. I am kind of wondering if anyone else has seen it, or at least, so often. As I said, I am not very observant, so I did not notice if Japan got maybe more gold or techs from it. They were both behind me tech wise, so they had nothing for me.

Another thing about that game, in line with capitulation. The Mayans did capitulate to the Mongols. The Mayans had a tech lead, in fact THE tech lead, but the Mongols led in power, big time. I took advantage of the Mongol/Mayan war to pick up 5 island cities from the Mayans, but Pacal wouldn't capitulate to me. The Mongols took a couple cities (2 or 3) from the Mayans, but on their continent, so that capitulation, to me, made sense. I may of taken more cities, but I was not a threat to the mainland. But, that is only one game. Takes me about a month to play a game

 

[lilnev]

Well I have yet to see an AI willing to capitulate to me. I beat Hattie down to one city, and she wouldn't surrender -- but she did vassalize to someone else, bringing another civ into the war against me. This has happened several times (not usually to the last city) -- either I have a war ally, in which case the AI capitulates to my ally, or I don't, in which case the AI will find someone else to capitulate to. And yes, I check every turn.

I feel like the change is that the AIs are now over-reactive to the power graph. I play Immortal, Agressive AI, and the stacks are huge. It's simply not possible to have more than about half the power of the typical AIs. I can still win wars -- but they're not very profitable when I come away with a handful of culture-swamped cities, and have to fight off an invasion from one of the larger cats on the block as well.

I've seen some snow-balling as well. It's 1650 in my current game, and Qin has four vassals. Pacal has one, and only Kublai and I are still on our own (large pangaea). I expect Qin to attack Kublai soon, making him the fifth vassal....

It may be that the threshold should vary with level and/or Agressive AI, and particularly the importance given to the Power graph.

 

[Fistleaf]

In my latest Monarch game, both the Ethiopians and the Mayans capitulated to me after I captured 3-4 big cities each(about 50% of their empire) and their Capitals. Not sure whether they would capitulate if I left their Capitals alone.
So I think the current capitulation threshold is good enough. Previously, civs would only capitulate when they have 1-2 cities. Now they are still willing to fight a hopeless war (I have tanks against their infantry, having pillaged their oil 1-2 turns before they discovered the tech for tanks) but will capitulate with a few main cities taken.

 

[lilnev]

Just an update, I lost when Qin took his sixth vassal -- everyone but me, since Zara had been eliminated -- pushing over the domination threshold.

 

[StMikael]

Well, this is all good and well, but WHERE and HOW do you adjust the capitulation threshold?