Diplomacy

[Perkus]

At the glacial late they've been patching, I wouldn't hold my breath for the C++ SDK to be released any time soon... Seems to me there is a skeleton crew left working on this thing.

 

[SlightlyMad]

Yeah, take2 bizarrely laid off most of the staff just prior to release. Honestly, they'd be better off ignoring patching and putting the SDK out to leave it to us.

 

[Thalassicus]

It feels like the equivalent of Caprica's cancellation. Fire the staff, push out what's done so far to the market, and be done with it because you're not expecting as much money as you want.

Still, I'd rather not get into that topic in these threads.

 

[SlightlyMad]

I'm not sure if this belongs here, or in fewer poor policies, but I had a thought about CS balance. I think the main reason they're broken is that getting free resources is already a great bargain at their price, especially in the early game when resources are more scarce. Additionally, the policy that gives you a portion of their science is still flat-out broken, and should really just be removed entirely.

What if that policy were replaced with one that gives you access to their resources, and without that policy you'd have no free resources from them at all? Is that something that's possible to mod in with the current tools?

 

[Thalassicus]

It's not possible with the current tools, though is certainly an interesting idea. At the very least, I could reduce the science bonus.

 

[Ahriman]

Did the science bonus change in the patch somehow? It seemed to be a fairly weak before patch.
Perhaps there was an implied buff by changing AI building priorities, so that those cities build more science structures?

I'm not convinced that the science policy is a problem, I think the problem is that its too easy to get city state alliances. Increase the cost of CS alliances (or reduce the influence yields from gold-buying) and I think you'll be fine.

 

[Thalassicus]

That was already done, fourth item in the list.

 

[SlightlyMad]

It's not possible with the current tools, though is certainly an interesting idea. At the very least, I could reduce the science bonus.

Darn, I was hoping that it would be possible since there's already some policy effects that alter their output and happiness values. I'd be in favor of reducing the science bonus, though. In games where I have any sort of focus on CS acquisition, the science bonuses tend to make up 30-50% of my total science output.

 

[Ahriman]

That was already done, fourth item in the list.

I know. I don't have a good feeling yet whether 80% is the right number, but it feels reasonable. My point was, I think this is the right lever to use.

City states are more important if having alliances with them is powerful. You want it to be worth going to war to protect your city state ally.

So I think its generally better to make it harder to get city state alliances than it is to nerf the value of those alliances.

I should have clarified; I meant the problem in *vanilla* was that its too easy to get alliances.
I have to playtest your mod more with the change.

* * *
Also, I have a really dumb/basic question for understanding diplomacy; do the aggressive/passive "you will regret this" or "we hope this won't damage our friendship" options have any impact on the AI at all? Or are they pure fluff?

 

[Tomice]

do the aggressive/passive "you will regret this" or "we hope this won't damage our friendship" options have any impact on the AI at all? Or are they pure fluff?

Good question from the "I always wanted to ask, but never thought of in the right moment" category thx for mentioning!

 

[Seek]

There was a thread in the strat forum a while back talking about this. Iirc, the consensus was that lying to the AI - giving a polite response and then ignoring the request - was the most harmful to relations. Don't know if anything's changed since the patch, though.

 

[Ahriman]

I assumed that was how it worked for, say, the AI warning you not to settle near it.
But how would it work for the AI informing you its attacking your city-state?
You get worse relations if you threaten them and don't declare war than if you declared war?

 

[Seek]

I assumed that was how it worked for, say, the AI warning you not to settle near it.
But how would it work for the AI informing you its attacking your city-state?
You get worse relations if you threaten them and don't declare war than if you declared war?

Lol, I doubt it, but you'd probably get a smaller "warmonger" hit (among all AI) if you threatened them and *did* declare. And a greater hit if you said, "oh, it's ok" then declared. I don't think any AI will think less of you if you don't declare war for any reason, unless of course if you had agreed to earlier with another civ. I'll try to find the thread.

EDIT: Looked for a while, but didn't find the thread. Sorry.

 

[Thalassicus]

I've been thinking over citystate quests for a long while. It seems a good way to do it might be a series of tiers. This would basically 1) ensure they do things that match the personality they supposedly have, and 2) request things logical for the timeperiod of the game and friendship status.

Citystates would pick a quest from the first tier they match criteria for. If everything's already done they make no requests (don't match any criteria, or all options completed for their tier). In addition, status is re-evaluated periodically to see if the civ should go up a tier (but not down a tier).

The personality types:

• Hostile or Militaristic - Warlike.
• Peaceful
• Irrational
• Neutral - Isolationist.

-- Warlike --

1. Unfriendly and undiscovered civs or natural wonders
   Check next tier.
2. Barbarian camps nearby
   Request assistance against a camp.
3. City-states nearby
   Request a CS killed.
4. Friendly
   Road.

-- Peaceful --

1. Unfriendly and undiscovered civs or natural wonders
   Exploration quest (find X).
2. Barbarian camps nearby
   Request assistance against a camp.
3. City-states nearby
   Check next tier.
4. Friendly
   Development quest (road, wonder, great person, resource).

-- Irrational --
Random hostile or peaceful personality when analyzing tier level.

-- Isolationist --
Only can be bribed.



In the standard game it works like this:

1. Unfriendly
   Random chance between barbarian or kill-neighbor quests.
2. Friendly
   Random chance of all quests, with some more likely based on personality.

While the vanilla method is simple to code, the problems are 1) kill-quests pile up and 2) it seems odd for CSs to give quests for things diametrically opposed to their personality.

The perfect example of this are "Peaceful" citystates that request bloodthirsty massacres of their neighbors. Even though this has a low chance to occur in any given dice roll, since the kill quest and barbarian quest are the only options before friendly, and the kill quest is the only one that can theoretically never expire, and tend to accumulate. Just making kill-quests have a limited duration is a decent band-aid solution but will still result in "peaceful" and "mind your own business" citystates requesting strange things. The kill quests also come with a serious downside: for each citystate killed your influence with its neighbors drops faster, even if the neighbor requested the kill. Completing it should give a "me love you long time" influence boost to counteract this.

We don't have access to the code yet that handles this and the current system is very inflexible (and incomplete, with only a few modifiers available), but citystate quests are still something that's been bothering me for a very long time. Citystate quests and capture bonuses will probably be the first thing I work on whenever we get c++ access.

 

[Seek]

We don't have access to the code yet that handles this and the current system is very inflexible (and incomplete, with only a few modifiers available), but citystate quests are still something that's been bothering me for a very long time. Citystate quests and capture bonuses will probably be the first thing I work on whenever we get c++ access.

Yet another thing I'm looking forward to the dll access for! Agree these sorely need change.

A little error in the current Balance - Diplomacy build:
The influence from kill barbs quest has dramatically increased (to 30), but the tooltip on the notification still says you have gained 5 influence.

Also, is there any way you can fix the road quest? It's been broken since the patch.

 

[Thalassicus]

Interesting... so the tooltip is hardcoded instead of read from the XML. How am I not surprised?

 

[V. Soma]

Thal, I don't kow if this belong here or not but:

The following I consider almost a bug, but definitely wrong:

I have Open Borders with Civ A
I have war vs. civ B and I fight enemy units on the land on Civ A.

OB with civ A ends and hey - my (and civ B's) units are thrown out of the battlefield
to past the borders of civ A, and THEN I can re-new the OB.

This is way stupid, uhh!

Should be:

OB with civ A ends, then diplo window come up automatically and lets me to re-new the OB,
BEFORE anything happens with units on the land of civ A...

Thal, what do you think about it?
Can this thing be solved?

 

[Thalassicus]

Not something I can deal with sadly... that border-tossing was added in a patch in the internal code.

 

[Sneaks]

Border toss is a necessary evil to stop ob troop positioning then Dow. I am not sure if I could ugly hack this in the lua but I will look

 

[V. Soma]

Border toss is a necessary evil to stop ob troop positioning then Dow. I am not sure if I could ugly hack this in the lua but I will look

I am no coder but logic would dictate that the two situations are nicely different,
thus could be handled differently:

A:
You are in the land of a civ and you DoW on the same civ:
your troops should be tossed out of the land, no matter you have OB or not...

B:
You are on land of civ A with OB and you DoW or have ongoing war vs a different civ:
your troops should stay on land of civ A
AND
end of OB should not toss out troops in the same turn,
only if OB is not renewed in that turn...

*

I understand if coding in practice is difficult to make it,
but in theory, it should be possible, IMO...