Report Questionable Behavior

[MartinHarper]

If I am busy conquering French lands, and the Incans capitulate them, then I get annoyed with the Incans. On the other hand, if the French are busy conquering me, and the Incans capitulate them and thus force them to make peace with me, then I like the Incans. Since capitulation forcibly creates peace treaties, the diplomatic consequence should be based in part on how much other civs want those peace treaties. Capitulation also forcible creates declarations of war, and these are already accounted for.

Conversely, if I am busy conquering French lands, and the Incans take them as protected vassals, then the French have effectively caused the Incans to declare war on me, and I dislike that. Therefore that should give the French a "-1 you brought in war allies against us" modifier.

 

[r_rolo1]

I strongly disagree. If I take, for example, Hitler as my vassal, and he has nuked tons of other countries, I think that the rest of the world should be horrified at my choice. I really don't mind the system that is in place now, I don't want to change it. I just want to expose it in the UI to players.

You took my sentence out of context. i was refering to the change proposal I had made ( -2 per hated vassal , + 2 per liked vassal ). With 8 vassals no one likes , that would give the absurd modifier of -16 , and that IMHO is too much. Also it had the potential to infinite up or down in diplo, a thing that is also undesirable ( and was from there that you took that sentence )

About how the system is... well, I beg to difer The current averaging system treats both vassal(s) and master in the same way , like if it was a partnership. By definition a master-vassal relationship is not a partnership, because one of the participants has the upper hand. IMHO treating as equals civs that are tied in a unequal relationship is simply stupid....

 

[Afforess]

You took my sentence out of context.

I'm really good at that.

About how the system is... well, I beg to difer The current averaging system treats both vassal(s) and master in the same way , like if it was a partnership. By definition a master-vassal relationship is not a partnership, because one of the participants has the upper hand. IMHO treating as equals civs that are tied in a unequal relationship is simply stupid....

Yes, but not in the way you think. Some Master-Vassal relationships are very biased towards the masters, and the masters get lots of benefits, free resources, etc... But some relationships help the vassal more than the master. You need to account for both.

If I make a stupid decision, and take a well-hated civ on as my vassal, I should have to pay the consequences diplomatically. The way the diplomatic effect works now, reflects that quite well. I don't mind that it averages. Although we may be technically two separate nations, for all intensive purposes, the Master tends to absorb the Vassal. They become one new nation, with a better leader (hopefully), more resources, and military power, and a new diplomatic score. Averaging seems the best way to do this. There is nothing wrong with this in my mind, and the way Firaxis made it works beautifully (for a change...), except for the UI. Jdog just needs to expose that to the players, and people will be a lot happier...

 

[Yakk]

And yet, taking capitulation pretends to be a viable option. Assuming the world has developed into poles, if you accept capitulation, all of your former allies will now treat you like scum.

And if you instead attack and take over an ally of theirs, and then accept capitulation, they like you more. It really, really, really doesn't make sense.

 

[MartinHarper]

for tech trades there's a decent chance the tech will find its way to the vassal.

I don't think so. If brokering is disabled, the master can't trade to the vassal. If brokering is enabled, the master may want to avoid the vassal trading the tech to someone else. Additionally, the master may want to keep the vassal under-teched, so that if it denies tribute it will be easier to conquer. To my mind, tech trades should only be based on team attitude for permanent alliances.

On a related note, there is some code in CvTeamAI::AI_techTrade that denies tech trading with "we would have nothing to gain" if the tech is 50% researched:

if (GC.getGameINLINE().isOption(GAMEOPTION_NO_TECH_BROKERING))
	{
		CvTeam& kTeam = GET_TEAM(eTeam);
		
		if (!kTeam.isHasTech(eTech))
		{
			if (!kTeam.isHuman())
			{
				if (2 * kTeam.getResearchProgress(eTech) > kTeam.getResearchCost(eTech))
				{
					return DENIAL_NO_GAIN;
				}
			}
		}
	}

The idea is that if the civ finishes researching the tech itself, it can trade it around. However, this logic should only apply if there is a civ somewhere that (a) doesn't have the tech and (b) we are not furious with. Fixing this might help backwards civs catch up when trade brokering is disabled.

 

[PieceOfMind]

And yet, taking capitulation pretends to be a viable option. Assuming the world has developed into poles, if you accept capitulation, all of your former allies will now treat you like scum.

And if you instead attack and take over an ally of theirs, and then accept capitulation, they like you more. It really, really, really doesn't make sense.

I think it makes sense just fine.

In your second example, where you take an ally of theirs, the bad diplo for doing such a thing should come from the "you declared on our friend". Once that friend is a vassal of yours you are under an agreement to protect that civ and so you are sort of more on their side. In this case it makes sense that your rival has better diplo with you.

If you are throwing out agreements all over the place to protect the most hated people in the world, the people who hate them will not like you for protecting those hated people.

The averaging of diplo across a whole team makes sense. The main problem that people have with vassal diplomacy is that the interface is not transparent about it. People like TMIT are very vocal about it, believing that everything in the interface has to be the absolute truth. While I don't think 100% transparency is necessarily the aim, it's certainly better to err on that side than lead players to believe that the AI and you are best mates when in fact the AI secretly hates you for vassalising the 3 most hated AIs in the world.

In the end, you need to remember the most important bit: For most intents and purposes masters and vassals are part of the one team. That one thing explains away all of the things that "don't make sense" if you think about it.

 

[Yakk]

But it is a nearly destroyed, enslaved, rump-kingdom. From an empire that was your steadfast ally for millenia.

And he went off, destroyed and conquered a 3rd party empire, then enslaved it as a puppet state with no remaining foreign policy.

This 3rd party empire -- if it was your friend, you are now happier than if it was your enemy.

I suspect there is a fundamental disconnect here. Capitulation, to me, means "we are defeated" -- Vichy France kind of thing. The allies of Germany didn't suddenly hate Germany for keeping around Vichy France; the allies of France didn't suddenly love Germany for keeping around Vichy France.

When you average without any weight, that is how large the impact is. A 1 city state that you are neutral about being the capitulated vassal of a 100 city empire you are +10 friendly with, results in a -5 relationship hit on the 100 city empire.

I don't think that makes sense, I think it is poor AI. You think it makes sense. /shrug.

 

[Dom Pedro II]

Because they are quicker than clean conquest (and less war weariness), avoid the vassal to another civ problem, and avoid to kill you with maintenance cos of tundra city. And they can be put to good use without too much problem.

And it would be better to have to slowly reduce the culture in each of the conquered cities rather than just raze their last crappy city to get a clean culture slate in the occupied cities? You already get hit with a diplomatic penalty simply for having a vassal. But look at it another way... Why should I get a big free diplomatic boost by vassalizing a popular civ?

Anyway, I think it's a terrible idea to make vassal painless, because IRL vassals weren't painless.

But IRL, occupying countries isn't painless either... but in Civ, it's fairly painless by comparison to the kind of trouble real world empires run into.

The averaging is bad for three reasons:

1) It can affect AI behavior too drastically. The AI has one particular view of a civ, and then for absolutely no good reason, it suddenly changes policy.
2) It's unrealistic from a roleplay standpoiint. Pretty much already explained by a few others here.
3) It makes vassal states too complicated by requiring players to weigh the potential vassal's standing with all remaining civs in the world and compare it against their own standing.

 

[Afforess]

But look at it another way... Why should I get a big free diplomatic boost by vassalizing a popular civ?

Because your essentially protecting the world's friend. If I think your protecting my friend, I would be happier towards you too.

The averaging is bad for three reasons:

1) It can affect AI behavior too drastically. The AI has one particular view of a civ, and then for absolutely no good reason, it suddenly changes policy.

Taking a vassal is not a decision for the faint of heart. If I vassalized RL Nazi Europe, instead of letting Berlin fall, I bet I would get a lot of pissed of countries invading me.

2) It's unrealistic from a roleplay standpoiint. Pretty much already explained by a few others here.

And I disagreed with the others too.

3) It makes vassal states too complicated by requiring players to weigh the potential vassal's standing with all remaining civs in the world and compare it against their own standing.

Again, taking a vassal isn't a decision for the faint of heart. If you think the good < bad, just kill them off completely. For instance, your at war with Player 2, and I hate him too, and am 3 turns away from getting my entire army ready for war. Then, you take his capital, and he offers to capitulate. You accept. Essentially, your protecting my enemy. I have this huge army that I was going to use against him, and now I'm ticked off at you for protecting him. Want to guess what happens next?

 

[Roland Johansen]

Shouldn't the vassal effect on diplomacy decrease over time. I'd say that after a while, the vassalised nation is becoming part of your nation and getting less independent relations that would affect yours.

 

[Afforess]

Shouldn't the vassal effect on diplomacy decrease over time.

It already does. All diplomacy* has decays already, they are just set ridiculously high. (high = low probability of forgetting). As time wears on, the AI will hate your vassal less, and thus, you less.

* Not really all. That's why I made a realistic diplomacy modcomp.

 

[MartinHarper]

Looking at the Vichy France example, I think the big difference is that the French liked America, because America declared war on Germany. However, in Civ IV, the French would give the Americans "-3 You declared war on us", even though it was a war of liberation. Therefore, I think that when war is declared on a master and its capitulated vassal, only the master should give a "-3 You declared war on us". The vassal would, at best, give "-1 You declared war on our friend".

This is also good gameplay: a capitulated vassal should welcome DoWs on their master, as it presents an opportunity to break free, either because the master gets killed off, or because the vassal loses 50% of its cities. However, colonies, protection vassals, defensive pacts, and permanent alliances should work the same way as now.

 

[Dom Pedro II]

Because your essentially protecting the world's friend. If I think your protecting my friend, I would be happier towards you too.

Yeah, I'm sure Manchuria's friends let out a big sigh of relief when Japan took them as a protectorate in 1932. Forcing a country to capitulate is not "protecting" them. It's subjugating them. Being happy towards a friendly civilization's master seems awfully pollyannaish, don't you think?

Taking a vassal is not a decision for the faint of heart. If I vassalized RL Nazi Europe, instead of letting Berlin fall, I bet I would get a lot of pissed of countries invading me.

You missed my point. The point of the first argument was that the AI whose attitude is being modified could be dramatically changed in a way that is detrimental for itself. If the AI is pursuing a helpful relationship with a player, but that player keeps one city alive of its worst enemy, that could cause them to become hostile, which would hurt the AI player. Or, conversely, if the AI player is prepping for war against one player, but that player vassalizes a friend, the AI player might actually scrap its war... again... for no particularly good reason. The vassalizing of a third party has very little effect on them except, perhaps, to work in the exact opposite way from what the averaging does.

If Player A vassalizes my enemy, and I'm friendly with Player A, it means my enemy can no longer attack me (unless Player A declares war on me). On the other hand, if I'm on bad terms with Player A and they vassalize my friend, that means my friend might be forced to go to war with me in the event of war between myself and Player A. I shouldn't like Player A more for turning a potential ally into a potential enemy.

But eventhough your example of Nazi Germany was not really responding to the point I was making, I'm going to address it anyway. To be more specific regarding your initial point, suppose the Western Allies of Britain/France/USA had allowed the Nazi leadership to keep whatever little piece of Germany they hadn't taken yet in exchange for completely surrendering their sovereignty to the Western Allies. The USSR would be furious to be sure. However, the USSR would have also been furious if the Western Allies had signed a peace treaty with Hitler and left the Soviets high and dry. But in the game, I can sign a peace treaty with a mutual enemy, and they AI couldn't care less aside from losing the "mutual war" attitude bonus.

Your extremely unusual Nazi Germany example is about the closest to justifying averaging of attitudes as one can get, but the reason has more to do with previous agreements on post-war Europe between the USSR and the Western Allies, and the extremely heinous conduct of the Nazi regime. But the game should not be coded to satisfy extreme examples.

Again, taking a vassal isn't a decision for the faint of heart. If you think the good < bad, just kill them off completely. For instance, your at war with Player 2, and I hate him too, and am 3 turns away from getting my entire army ready for war. Then, you take his capital, and he offers to capitulate. You accept. Essentially, your protecting my enemy. I have this huge army that I was going to use against him, and now I'm ticked off at you for protecting him. Want to guess what happens next?

Then why won't you be upset with me when I take all those tempting cities you were going to help yourself to?

 

[Dom Pedro II]

This is also good gameplay: a capitulated vassal should welcome DoWs on their master, as it presents an opportunity to break free, either because the master gets killed off, or because the vassal loses 50% of its cities. However, colonies, protection vassals, defensive pacts, and permanent alliances should work the same way as now.

Great point. Being a vassal is generally not beneficial to the vassal. They should welcome the chance to undermine their master's control.

 

[PieceOfMind]

Well I see there is some fundamental disagreement about the nature of vassals in this thread. It's really interesting to see the spread of views on this, and it made me ask a few questions...

I'm finding it hard to construct a very coherent argument here. The main thing I'm concerned about is that we don't turn capitulating vassals into something that is done without thought. The ability to take vassals and the consequences surrounding it are core to the gameplay and changing the code here could drastically change the direction of games.

To those who are proposing significantly changing the diplo about vassals, what would the negative consequences, in your mind, be for capitulating a small but very hated civ?

 

[Dom Pedro II]

To those who are proposing significantly changing the diplo about vassals, what would the negative consequences, in your mind, be for capitulating a small but very hated civ?

I have to answer this question with a question... if the game developers hadn't just decided to just slightly modify the code for averaging the universally accepted team attitude to vassals, would ANYBODY be defending it so fervently?

I haven't heard anybody clamoring for a diplomatic penalty if the AI asks another player to declare war, that player agrees, does absolutely nothing, and then signs a peace treaty with that most hated enemy letting them go on free and unrestricted to fight another day. But I bet you if the game developers had included it and somebody wanted to get rid of it, there'd be a line of people waiting to tell us how it would be terribly unbalancing to take it out of the game.

Vassal states already has built-in flaws. Personally, I find them sufficient to almost never use vassal states. Was there any real need to further gimp it? My guess is that this wasn't even an intentional penalty. I think the probably just didn't think it through that much.

 

[PieceOfMind]

I have to answer this question with a question... if the game developers hadn't just decided to just slightly modify the code for averaging the universally accepted team attitude to vassals, would ANYBODY be defending it so fervently?

I'm not sure what modification you're talking about so if one assumes I'm defending the code then my answer would be yes.

I haven't heard anybody clamoring for a diplomatic penalty if the AI asks another player to declare war, that player agrees, does absolutely nothing, and then signs a peace treaty with that most hated enemy letting them go on free and unrestricted to fight another day. But I bet you if the game developers had included it and somebody wanted to get rid of it, there'd be a line of people waiting to tell us how it would be terribly unbalancing to take it out of the game.

You raise a good point. There should perhaps be a penalty for that. That said, treaties of peace seem to me to be more like a neutral agreement. It is only when you are trading with an ally's enemy he gets exceptionally peeved. As it is now, if you are a neutral third party watching as your friend fights someone they hate, they will neither hate you for not being in the war nor like you for joining in the war (I'm talking about the mutual military struggle bonus). So you end up somewhere in the middle.

The actual event of signing a peace treaty shouldn't IMO cause a diplo change. It should be the war/peace status that causes a diplo modifier. This already happens - if you sign peace with your friend's enemy you no longer get the mutual military struggle.

What you are asking for is a reworking of the way you bribe enemies into war. The agreement should be somehow more binding like having you be unable to sign peace until he does.

Or even better, if you declare peace with his enemy you'll suffer a relations hit but only until he also signs peace. That way, the sour feelings only exist while he's still at war and you're doing nothing.

Vassal states already has built-in flaws. Personally, I find them sufficient to almost never use vassal states. Was there any real need to further gimp it? My guess is that this wasn't even an intentional penalty. I think the probably just didn't think it through that much.

Vassal states are pretty good already.

Benefits:

• +1 in every city
• Can demand resources
• Can use territory for healing and other military benefits that come with having home territory (e.g. defense in cities, paradrops from forts etc.)
• Can have foreign trade routes under Mercantilism
• Military assistance in conflicts
• Free defenders whenever you gift a city to them (e.g. colonies).
• Vassals population and land contribute partly to your score towards Domination victory
• Vassal does not need to be killed for Conquest victory
• Can tell a vassal what to research (true for teammates and perm allies too of course)

Drawbacks:

• I could have had more cities instead.
• Vassal might still use espionage against me (I think this is improved in BBAI though)
• Sometimes vassal screws you by doing something like build a spaceship part in a small city

Please add to the list...

 

[winthrowe]

By definition a master-vassal relationship is not a partnership, because one of the participants has the upper hand. IMHO treating as equals civs that are tied in a unequal relationship is simply stupid....

this got me thinking, what if keeping a a civ 'under your thumb' was a good thing. something like:
assuming currently averageAttitude = (vassalToCiv+masterToCiv)/2

changing to something like:
AverageAttitude = (vassalToMaster*vassalToCiv/abs(vassalToMaster) + masterToCiv)/2

so if you've almost destroyed that evil civ and forced them to capitulate to you, others would enjoy your policing ("+x We like you keeping our former enemy in check"), but if monty has been terrorizing others and then offers his services, your good relationship with him would give a similar penalty as today. maybe off the wall and a horrible idea, or might spark someone else.

 

[Afforess]

Yeah, I'm sure Manchuria's friends let out a big sigh of relief when Japan took them as a protectorate in 1932. Forcing a country to capitulate is not "protecting" them. It's subjugating them. Being happy towards a friendly civilization's master seems awfully pollyannaish, don't you think?

No, I don't. I disagree completely. When it's a choice between dissolving one nation's national bonds and recognition on a world-wide stage, or merely subjugating them, I think I would rather see a friend live then die.

How about a more RL example. You have a pet, that you love. I'm a repo man, and you haven't paid your taxes. Would you rather I kill the pet, or sent it to the local animal shelter? That's the differeance.

You missed my point. The point of the first argument was that the AI whose attitude is being modified could be dramatically changed in a way that is detrimental for itself. If the AI is pursuing a helpful relationship with a player, but that player keeps one city alive of its worst enemy, that could cause them to become hostile, which would hurt the AI player. Or, conversely, if the AI player is prepping for war against one player, but that player vassalizes a friend, the AI player might actually scrap its war... again... for no particularly good reason. The vassalizing of a third party has very little effect on them except, perhaps, to work in the exact opposite way from what the averaging does.

Fine, then expose the potiental attitude shift the AI decision making process. That's what Better AI is all about, right?

If Player A vassalizes my enemy, and I'm friendly with Player A, it means my enemy can no longer attack me (unless Player A declares war on me).

Really, you prefer stronger allies? I'd rather have a lot of weak allies I know I could kill later on.

On the other hand, if I'm on bad terms with Player A and they vassalize my friend, that means my friend might be forced to go to war with me in the event of war between myself and Player A. I shouldn't like Player A more for turning a potential ally into a potential enemy.

You're looking at it the wrong way. It isn't whether the other civ is a potential ally or not, it has to do with whether they are still alive to fight another day or not. Your looking at it from the wrong standpoint.

But eventhough your example of Nazi Germany was not really responding to the point I was making, I'm going to address it anyway. To be more specific regarding your initial point, suppose the Western Allies of Britain/France/USA had allowed the Nazi leadership to keep whatever little piece of Germany they hadn't taken yet in exchange for completely surrendering their sovereignty to the Western Allies. The USSR would be furious to be sure. However, the USSR would have also been furious if the Western Allies had signed a peace treaty with Hitler and left the Soviets high and dry. But in the game, I can sign a peace treaty with a mutual enemy, and they AI couldn't care less aside from losing the "mutual war" attitude bonus.

If you sign with a mutual enemy, the AI will average it's hate toward the enemy vassal into your diplomatic stance. It does have an effect.

Your extremely unusual Nazi Germany example is about the closest to justifying averaging of attitudes as one can get, but the reason has more to do with previous agreements on post-war Europe between the USSR and the Western Allies, and the extremely heinous conduct of the Nazi regime. But the game should not be coded to satisfy extreme examples.

Why not? Since you are the one arguing for change here, and Firaxis obviously agrees with me, you need to give a convincing argument to win me over.

Great point. Being a vassal is generally not beneficial to the vassal. They should welcome the chance to undermine their master's control.

It can be. If I know Player A is much stronger than me, and gearing up to kill me, and Player B is strong enough to hold A back, I'd be Player B's vassal in a heartbeat. Temporarily, of course.

To those who are proposing significantly changing the diplo about vassals, what would the negative consequences, in your mind, be for capitulating a small but very hated civ?

I have to answer this question with a question... if the game developers hadn't just decided to just slightly modify the code for averaging the universally accepted team attitude to vassals, would ANYBODY be defending it so fervently?

Probably not. But that doesn't matter. Firaxis DID code it that way, and players have gotten used to that game mechanic. Better AI's goal here isn't to go change game mechanics whenever they feel like it. The goal is to improve the AI's behavior.

Vassal states already has built-in flaws. Personally, I find them sufficient to almost never use vassal states. Was there any real need to further gimp it? My guess is that this wasn't even an intentional penalty. I think the probably just didn't think it through that much.

Those aren't flaws, those are features.

The way vassalage works doesn't need to be changed here. If you want to change it, do it in your respective mods, but Better AI isn't the place for altering game mechanics. I think Jdog just needs to expose the diplo penalty hit to the AI decision making, to help decide if they want to take a vassal, and to expose it to the UI, for players to see.

 

[r_rolo1]

I have to agree with what Yakk said some pages ago that Firaxis had putted in the same vassal bag 3 very diferent status . colony ( or should I say ex-colony ? ), protectorate ( aka voluntary vassal ) and capitulated civ .That made me wonder what was in the mind of the coder for the concept of vassal ...

First of all , in most aspects the way that vassal behaviour is coded points to the concept of capitulated : someone that was beaten to crap , but that still is allowed to have some kind of control about their diplomacy ( think Carthage after the second Punic war or even most of the countries ocupied by Germany in WW II , like Denmark or the Netherlands ) and even the economical facets of the issue ( trade routes with the vassals not considered foreign , the fact that the master can direct research and military operations, the more recent inclusion of the removal of diplo-driven obstacles to tech trading .... ) also point in that direction ( it could also point to the other 2 options though ) . This does not bode well for protectorates and colonies ( that were, at best, badly implemented IMHO ... they mixed vice-royalties, dominions and independent coutries with cultural links with the motherland all in the same bag ) but it what we have now ...

Given that the mechanism we have from the original coder is clearly geared towards capitualded civs, IMHO it is hardly defendable that master and vassal should be treated as equals in terms of diplomacy.

First of all, vassals in Civ IV have no decision on war and peace issues ... that alone makes stupid that the AI war decisions are made with the presumtion that master and vassals ( note the use of the plural ) are equal partners in terms of diplo, like if the master called the vassals to a council and the diplo was decided by democratic vote . IMHO atleast this kind of decisions should be made from free civ to free civ ( the only entities that can make war and peace, remember ) with some weighting by the like/ dislike of vassals. In fact a similar aproach is followed by this mod in terms of the capitulating rules .... same for demands: I really can't see why it is allowed to demand things from vassals of other civs without passing by their master ... and for the diplo votes if the candidate(s) is(are) master(s) ( not sure of how to deal with vassals and diplo victories though I would really find strange if RL Kabul or Bagdad governement was elected to rule the world though )

In terms of trading things are a little more fuzzy, though ... tech trading ( in case of no brokering is diferent OFC ) is probably better suited for the current average model, because , with the current 3.19 rules , if you tech trade with a vassal, the master ( and probably the other vassals as well ) will receive the tech fast enough. Resource trading is probably more suited for treating each civ separately with some weighting ( you can't broker or give resources you buy )

Anyway, regardless of what model is implemented, the real diplo status should be seeable ( not that i think that anyone disagrees with this ) . Too much in the decisions you have to do during a SP game revolves on knowing the AI attitudes towards you and others to allow having AI forgetting to tell you that, in spite they are Friendly with you, they will treat you as being Annoyed with you because you have two vassals they are Furious with....