New Version - November 5th (11/5)

[lordrune]

I can confirm that this is also happening to me.

Honestly, if the AI only used this trading system for cities along their border (i.e. the six tile limit meant to be in place) I'd probably love this feature, but as it is it's a bit odd.

EDIT: That is, a city being traded for should be within six tiles of the civilization which is receiving it.

I'm only new to this patch, but I've seen some really weird stuff being offered, like the AI offering me a city within the heart of their empire (or on the other side) for one I've settled that's a long way away from them. It would have created two dis-contiguous empires. I thought it would have been considered a known bug, and I'm a little surprised it's considered to be working as intended (had I known the latter, I might have taken more note of the exact circumstances, screenshot etc).

And I've seen the AI accept similarly strange deals themselves. There are cases where city trading might make sense, but too often it's like Russia going to America: "Hey, I don't really have a use for St. Petersburg, but I'd like Portland, wanna trade?" and America going "Suuuure."

Not gonna happen, sorry. Territory exchange and sale is very much a historical thing, especially after war.

Well, depending on what sort of things I see as I play more games, it may be something that will turn me off the mod. I understand it's a cool feature and that it was an achievement to getting it working, but I do hope the team keeps an open mind about whether it really helps the feel and the flow of the game, or whether it adds an element that can feel slightly nonsensical, and don't close your mind to the possibility of altering it further. I'll reserve my final judgement until I've played it a bit more though.

 

[albie_123]

Not gonna happen, sorry. Territory exchange and sale is very much a historical thing, especially after war.

Okay. So, at least we've now established that this is a 'historical' feature, and not one rooted in the ideal of balance or 'fixes'.

 

[Gazebo]

Okay. So, at least we've now established that this is a 'historical' feature, and not one rooted in the ideal of balance or 'fixes'.

I'm not sure what you are getting at. It's a fix - the AI could always trade cities in vanilla BNW, it was just really bad it (so bad that the chance of the code working correctly was super rare) and deals were always dumb. Those AI cities you got in peace deals? That was just the AI adding up x% of their owned cities, with no consideration of value relative to the peace deal.

We've made it so that they know what they're doing and, just like a human player, they are allowed to do so with other civs. That's the goal of the CP - to make the AI capable of all game actions that a human player can do.

G

 

[lordrune]

TBH when the AI came up and said "Hey, I want that city you settled, here's one of mine which is on the other side of the continent to you" I was like Whhhaaaaaatttt? game broken?

Now I know it is actually planned that the AI does this, I'll pay more attention and see if I can figure out how the trades that the AI does some sort of sense. It still seems to me that the AI is overly eager to give away core parts of their empire in peacetime.

 

[albie_123]

I think we may be misunderstanding one another - it's not that the AI now being able to trade cities is inherently an issue, it's their overwhelming enthusiasm to do so. While trading for cities is of course something human players can do, and I'm glad that the AI now understands it better, it's not something human players ever really did outside of the occasional demand in a peace treaty, because it's almost never beneficial to trade cities of similar value rather than building up the cities you currently have.

The only time it's good to request a city in trade is when you don't have to give away a (non-useless) city in return, which almost always means military superiority and either a potential or currently ongoing war, and the only time it's good to sell a city in trade is if there's absolutely no chance of it ever being productive - in which case, the AI probably shouldn't want it without large concessions, and certainly wouldn't give a city for it. As such, city trade should only be taking place when:

a. A strong Civ is threatening a weak Civ to give up land,
b. A weak Civ is attempting to placate a strong Civ by giving up land,
c. A Civ has a useless chunk of land that it cannot make use of, and uses it as a bargaining chip in a trade.

What shouldn't happen (in my opinion, of course) is two Civs making a somewhat equal trade of two cities. It doesn't make sense from a gameplay perspective and it makes even less sense from a historical perspective.

Basically: I love that the AI can now actually value cities properly, and I'm really happy that they'll finally know what they're worth (will make for some satisfying peace deals). That is, as far as I'm concerned, a fix to their behavior. It's the change to their behavior - their sudden eagerness to trade cities, especially with one another - that is really offputting.

 

[Edaka]

You've got it correct, though. Standard = a map using one of the predefined <Worlds> sizes (i.e. not a custom size map).



Not gonna happen, sorry. Territory exchange and sale is very much a historical thing, especially after war.



What version? There's a hard-code on distance for the AI, but all things can break.

G

The latest - 11/5. It was in this version when AI got really active about trading cities, too, at least in my games.

 

[Deadstarre]

City trading kills immersion and I hope it is removed from the game (or made optional)

Not gonna happen, sorry. Territory exchange and sale is very much a historical thing, especially after war.
G

Spying is of course a historical thing and actually a very significant part of this game, but it is currently made optional.

Without fully understanding whats involved, can i ask, would it be a lot of work to make city trades optional?

 

[sleepyx732]

People are attempting to use a broken part of a feature to clamor for the whole feature to be shut down.

For now, I wouldn't recommend any radical changes. Just fixing what is currently not working as intended, and see what people say then. I can understand the strategic importance of having a city elsewhere not connected to any of your cities, but I don't believe the AI is equipped to deal with this properly.

 

[shaglio]

I don&#8217;t know how city trading is currently coded, but is it possible to make the chance of a city trade be weighted by how close the respective cities are to the other player/AI&#8217;s capital compared to their own capital? In other words, if City A is closer to Player B&#8217;s capital and City B is closer to Player A&#8217;s capital, there is a greater chance of the trade being proposed.

 

[Funak]

I don’t know how city trading is currently coded, but is it possible to make the chance of a city trade be weighted by how close the respective cities are to the other player/AI’s capital compared to their own capital? In other words, if City A is closer to Player B’s capital and City B is closer to Player A’s capital, there is a greater chance of the trade being proposed.

The AI used to only be able to trade for cities within X tiles of their border, but I believe it got removed.

 

[shaglio]

The AI used to only be able to trade for cities within X tiles of their border, but I believe it got removed.

That seems more of a yes/no possibility. I think the distance to capital (or border) should increase the chance of the trade occurring, not toggle it on or off.

 

[ilteroi]

hello everybody.

i'd like to offer some comments on the city trading issue, as i'm the one who wrote the latest version of the code. most of the suggestions above are in fact already implemented. except world wonders.

currenty considered factors are:
- economic value of the city: yields + resources
- overlap with other cities owned by the buyer (is good! don't want a potential enemy in the middle of your empire)
- distance from other cities owned by the buyer (no maximum distance though)
- discount for unhappiness
- diplomatic relations between buyer and seller
- connection to trade network
- founding player, current ownership
- "original capital" status (as it is a victory condition)

example valuations (you can take a look yourself using e.g. the tracespy tool):

a city founded by a third party, ratio about 1:2

City value for Baghdad from Isabella to Ramesses II is 94347
City value for Baghdad from Ramesses II to Isabella is 38637

a city founded by Ramesses, ratio about 1:5

City value for Memphis from Ramesses II to Isabella is 54609
City value for Memphis from Isabella to Ramesses II is 285502

this is in gold, but it's trivial to scale the total values up or down.

overall, i'm quite content with the way it's working, but i do recognize that some tweaking might be required. i'll add the world wonder check and increase the premium for giving up control. i might also increase the city values across the board to make it more difficult for the AI to even out a deal with resources etc.

let's see how that goes.

 

[Funak]

hello everybody.

i'd like to offer some comments on the city trading issue, as i'm the one who wrote the latest version of the code. most of the suggestions above are in fact already implemented. except world wonders.

currenty considered factors are:
- economic value of the city: yields + resources
- overlap with other cities owned by the buyer (is good! don't want a potential enemy in the middle of your empire)
- distance from other cities owned by the buyer (no maximum distance though)
- discount for unhappiness
- diplomatic relations between buyer and seller
- connection to trade network
- founding player, current ownership
- "original capital" status (as it is a victory condition)

example valuations (you can take a look yourself using e.g. the tracespy tool):

a city founded by a third party, ratio about 1:2

City value for Baghdad from Isabella to Ramesses II is 94347
City value for Baghdad from Ramesses II to Isabella is 38637

a city founded by Ramesses, ratio about 1:5

City value for Memphis from Ramesses II to Isabella is 54609
City value for Memphis from Isabella to Ramesses II is 285502

this is in gold, but it's trivial to scale the total values up or down.

overall, i'm quite content with the way it's working, but i do recognize that some tweaking might be required. i'll add the world wonder check and increase the premium for giving up control. i might also increase the city values across the board to make it more difficult for the AI to even out a deal with resources etc.

let's see how that goes.

It all looks awesome, and it works perfectly, thanks for your hard work.

 

[Gidoza]

I just wanted to comment on the change regarding not being able to influence CS with diplomatic units if that CS is allied to someone who is at war. It would be necessary for the converse to be true as well: if a civ is at war, it can't influence City-States to become its ally. Otherwise, you'd have an exploit whereby your civ at war could remain so intentionally, steal lots of CS from others, at which point they would be locked under your influence because you are at war. And this makes sense - because why would a CS want to willingly ally itself to someone currently involved in an armed conflict? It gives a definite sense of being "used."

My suggestion would be that one could, in either case, still use Diplomacy units anyways to gain influence, but alliances are unshiftable until that particular civ's war comes to an end, or their influence falls low enough to eliminate the alliance (by coup or rigged election, etc...), and the same for that civ's influencing other CS. Gives an additional incentive for peace or to crush an aggressor that is being controlling.

 

[kaspergm]

hello everybody.

i'd like to offer some comments on the city trading issue, as i'm the one who wrote the latest version of the code. most of the suggestions above are in fact already implemented. except world wonders.

currenty considered factors are:
- economic value of the city: yields + resources
- overlap with other cities owned by the buyer (is good! don't want a potential enemy in the middle of your empire)
- distance from other cities owned by the buyer (no maximum distance though)
- discount for unhappiness
- diplomatic relations between buyer and seller
- connection to trade network
- founding player, current ownership
- "original capital" status (as it is a victory condition)

example valuations (you can take a look yourself using e.g. the tracespy tool):

a city founded by a third party, ratio about 1:2

City value for Baghdad from Isabella to Ramesses II is 94347
City value for Baghdad from Ramesses II to Isabella is 38637

a city founded by Ramesses, ratio about 1:5

City value for Memphis from Ramesses II to Isabella is 54609
City value for Memphis from Isabella to Ramesses II is 285502

this is in gold, but it's trivial to scale the total values up or down.

overall, i'm quite content with the way it's working, but i do recognize that some tweaking might be required. i'll add the world wonder check and increase the premium for giving up control. i might also increase the city values across the board to make it more difficult for the AI to even out a deal with resources etc.

let's see how that goes.

Thanx for the explanation, but what we need seems to be some sort of sanity check.

Problem is, trading cities is not like trading resources. If I have one extra unit of Spices and AI has one extra unit of Salt, it'll consider that these items are equal values to use (if we are on good terms at least) and propose a trade. That makes sense, because both of us benefit from the trade.

However, if the AI has a city with a certain value based on the resources and buildings it has, and one of my cities has the same value to the AI, and vice versa on my end, that doesn't mean that it's a good idea for any, much less both of us, to swap the cities. And the AI doesn't seem to understand this, cf. multiple reports of multiple cities being swapped back and forth between multiple players in the game. So either there is some sort of error in your code, or the code needs to be revised to:
- Put much more emphasis on the strategic location of the city. Distance from borders must play a crucial role as well as not breaking up your own empire by giving the enemy a city in the middle of your empire.
- Be much less prone to sell off it's own cities spontaneously (i.e. without any motivation spawned by war).



EDIT > Just to give a show what happened in my case:

First, Austria swaps their city Graz with Siam in return for the Siam city Muang Saluang. This swap at best makes marginal sense for Austria, because they already have access to Spices in their city Linz, but at least it doesn't disconnect their empire. For Siam, on the other hand, this trade makes zero sense in terms of location of their empire and resources (the city of Graz is here Roman because it was later traded again).

Spoiler :

Siam then swaps the city of Graz with Rome in return for the city of Antium which makes even less sense for both parts, because soon this city will be inaccessible from the rest of the Siam empire and it cuts the Roman empire literally into two.

Spoiler :

It was at this point Maria Theresa came up with the suggestion that I swap my city of Guanzhou in return for the city she aquired from Siam, something that clearly made little sense for me (unless I was dying to get those spices).

Spoiler :

 

[ilteroi]

Thanx for the explanation, but what we need seems to be some sort of sanity check.

Problem is, trading cities is not like trading resources. If I have one extra unit of Spices and AI has one extra unit of Salt, it'll consider that these items are equal values to use (if we are on good terms at least) and propose a trade. That makes sense, because both of us benefit from the trade.

However, if the AI has a city with a certain value based on the resources and buildings it has, and one of my cities has the same value to the AI, and vice versa on my end, that doesn't mean that it's a good idea for any, much less both of us, to swap the cities. And the AI doesn't seem to understand this, cf. multiple reports of multiple cities being swapped back and forth between multiple players in the game. So either there is some sort of error in your code, or the code needs to be revised to:
- Put much more emphasis on the strategic location of the city. Distance from borders must play a crucial role as well as not breaking up your own empire by giving the enemy a city in the middle of your empire.
- Be much less prone to sell off it's own cities spontaneously (i.e. without any motivation spawned by war).

from my perspective the "problem" is that the AI is just too good at finding matches

however, i just increased the asymmetry between buying value and selling value, so it should be harder to make a trade now.

 

[Gazebo]

from my perspective the "problem" is that the AI is just too good at finding matches

however, i just increased the asymmetry between buying value and selling value, so it should be harder to make a trade now.

If a computer is good at anything, it is matching games.

G

 

[kaspergm]

from my perspective the "problem" is that the AI is just too good at finding matches

however, i just increased the asymmetry between buying value and selling value, so it should be harder to make a trade now.

Sounds good, will the new values be released in a new version soon?

 

[Eagle Pursuit]

I'm not sure where to post this comment, but...

I first learned about this mod from Quill18's videos, and I had to try it. It's really great. It's like a whole new game.

Great work, modders.

 

[lordrune]

Really appreciate the explanation and the tweaks, Ilteroi. I think they should be all that's needed, hopefully we will still see some great trades going on but not so many of the 'WTH' (pardon the vernacular) ones.

And I'm with Eagle Pursuit - learned about this mod the same way What's really nice is having a ready-to-go integrated mod setup that uses EUI and Civ 4 diplomacy - along with CSD and the other balance features that, for the most part, seem to make the game stronger - and built with compatibility patching. I've played with some of these things before but compatibility is sometimes a bridge too far for me technically.

Yes, there's snowball in this mod, but Civ is all about snowball, if we didn't really like it we'd be playing a different genre!