Getting Started

[Ahriman]

We're drifting off into details again... Tomice gets the general idea of what I'm going for.

I get the general idea of what you're going for, I just think it would be incredibly difficult to balance, and I don't think your method (using AI trade values) is the right method for trying to balance it.
Fundamentally, its hard enough to balance strategic resource units vs non-strategic resource units. Its massively harder to do the three-way balance of strategic resource units vs non-strategic resource units vs strategic-resource buildings.

If I elect to use one of my horses for a building instead of a horseman then:
I gain a superior building and a normal unit, and I lose a superior unit and a normal building.

Supposing that the horseman is replaced by the archer, and the superior circus is replaced by a temple, for balance we would need:
V(horseman + temple) = V(archer + circus)
where V(.) is a valuation function of some kind.

That is a harder problem to solve than just balancing the strength of the horseman.

I'm fine with tying buildings into local resources, that doesn't bother me. Its a decent way of making different city sites more interesting, and helping terrain guide the city specialization.
But trying to insert military vs economy tradeoffs for strategic resources through the whole game is a big change that is very tough to balance, and I don't think its really needed.
I think you can achieve your basic design goal without it.

I've been trying to find a way to balance the Watermill for 4 months straight

The current version feels fairly balanced, I don't see a problem with it.

Maybe I'm just a curmudgeon, but I really hated corporations in BTS. The implementation was incredibly confusing (the formula for maintenance costs was very non-transparent to the player).
Civ5 solved this problem in theory with strategic resources; now there is a value for your second and third iron resource, they let you build more stuff. The reward for multiple cows and wheat is the tile yield bonus, and the reward for multiple luxuries is the ability to trade them to other civs more luxuries you don't have.

Seems to me this whole dynamic is missing. Each strategic unit should have a non strategic counter to it.

I think individual RPS hard-counters is the wrong way to go in a 1upt system.
The AI has a tough enough time already, if you introduce a hardcounter system (rather than a softcounter system) then you will just exacerbate the human's ability to outperform the AI, which I think is undesirable.

 

[Txurce]

Thal,

With regard to your game dilemma as I understand it, why wouldn't you always want the gold that comes from settling and developing both kinds of resources? If you're so far ahead in your game that the additional gold means nothing to you... start playing on Deity, buddy!

Responding to your overall thrust, I would have no problem with a slight buff to the benefits of strategics to the builder, as opposed to buffing the amount of strategics. Offering that option ought to make a war-tilted game more interesting for builders.

Polycrates,

As Ahriman may have already said, I don't think you can compare the value of strategics to a warmonger vs a builder, because you almost can't put a price tag on the value of units in a war.

The big issue for me was that resources should be relatively scare - just enough to give you something to work with, varied enough to make each game different, and not enough to eliminate choice. This is what I think Thal is doing, now that he is steering away from increasing yields and buffing effects instead.

Ahriman,

I don't think an adjustment should be made to what the AI pays for luxuries, because the AI are often artificially happy due to the game design. In most cases they would never buy resources, and that would create a design-generated imbalance for the human where there would be no opportunity to trade anything except strategics. In terms of existing imperfections, I think this one is minor compared to the poorly coded AI happiness buff, and if anything would focus on that before I would on making trade more "fair."

 

[Txurce]

I've been playing with the new mod, and have found that giving gold to CS produces extremely unsatisfactory gains. (I have two Patronage policies and have given about 1000g in total to Genoa, but am only halfway on the Friend scale.) The AI isn't having the same problem - 2 of the 4 CS on my continent are allied with the AI at 475AD.

Were the influence numbers changed? (I assume I'm in the right thread.)

 

[Sneaks]

I also mostly agree with Ahriman on this one. Balance the existing units as is, but no need to add a builder purpose to the resources. It simply overcomplicates the early game for the AI.

What I think is most critically missing from this mod, and something I think alpaca got extremely right with his, was making all special resources excellent for yields. Seeing a sheep and 2 wheats should make you as excited as horses and iron. If you make strategic/bonus resources especially good, the builder already gets his reward simply by working the plot. If you then choose to take the unit advantage of owning that plot and selling it off, so be it.

I also really really suggest not relying on AI trade value just yet. The AI is still horrendous at determining how to use trade to its advantage.

 

[Txurce]

What I think is most critically missing from this mod, and something I think alpaca got extremely right with his, was making all special resources excellent for yields. Seeing a sheep and 2 wheats should make you as excited as horses and iron.

I don't know about food resources being as valuable as strategics, but more to the point, these mods addressed the food issue via buildings like the smokehouse. Wheat was skipped over in this regard, but is now being considered as a buff to a mill, I think.

 

[Sneaks]

Food resources having yields equal to strategics does not make them more valuable as you are not getting the physical resource to produce/trade.

I really would rather most of the yields not be tied to buildings but instead technologies, primarily because it allows for the advantage of location to boost a city when it matters most: during the growth of its first 5 citizens. As has been described by others, seeing 3 food resources in an area in Civ IV made you really excited. In Civ V, the bonus resources are honestly blah. There is no real advantage to settling here over there because the base tile yields are so similar, resource or not. If you tie these resource yield increases to buildings, the AI WILL NOT properly valuate settling locations.

 

[Thalassicus]

I agree balancing requirements on strategic buildings would be challenging. One reason it appealed is the very nature of the added complexity, much like the ideas for adjacency bonuses for terrain improvements (I like complex thinking games ). I also see the point it's probably big enough to be left to a different mod, though. I've mostly dropped the idea at this point.

I'd still like to identify ways we can make excess resources more meaningful, which is the goal.

My problem with the Watermill I'll detail in the City Development thread here: link.


@Txurce
The thing is, in vanilla the resource tile yields are usually identical or close to regular, improved tiles. Same with bonus resources. The end result is there's nothing really valuing one patch of land over another if our resource supply is full, so we can just drop settlers wherever at that point, without much thought involved.


@Sneaks
Valkrionn did something similar through making bonus resources tradeable and economic, used by new buildings. Alpaca modified the base yields. I'm exploring a middleground approach, detailed in the Terrain Improvements thread here.

Technologies can't be tied to resource yields, only improvements. I considered the improvement-buffing through techs method too, and took that route for the mine, so I agree it's a viable option. In an early version of terrain improvements mod I had a camp boost at gunpowder, plantation boost at economy (I think), and so on.

I'm very sleepy so don't quote me on this, but the general reasons I went with building improvements are: rewards investment more (three ways instead of two), and buffs 'one big city' placement decisions over 'many small cities'. Polycrates describes this rather well here.

That's why I really like what you've done with the smokehouse; it gives you a bit of a bonus at the outset for a good site, but really rewards you when you start actually investing in it. Likewise buildings like the Monastery, Mint and Forge - and for me it just really feels like a rewarding mechanic when you get these special bonuses from buildings. They promote picking a "perfect" site to maximize the number of those resources you can get in a particular city's radius, since the one copy of the building pays off more. And I think it helps empower special resources while still promoting well-developed cities.

Basically, with the bonus on the resource or improvement, it favors ICS a bit more: place 2 cities to get 2 resources right away instead of one city between the two (and somewhat further than either). Bonuses tied to the building help make city placement more interesting by getting "ideal" spots that can overlap multiple sources with 1 building to improve efficiency. It also follows precedent from the Monastery, Mint, and Seaport.

 

[Sneaks]

Thal, here is my biggest worry right now:

This modpack is getting massive. It is full of spectacular ideas all around. However, thanks to the crappy tech tree in this game, space on technologies is disappearing at record pace. Construction is already beyond this point with your mod. If you play as Inca:

-Construct Terrace Farm
-Bridges over Rivers
-Colosseums
-Agra Fort
-Great Wall
-Circus Maximus

Already missing something there.

I think it might be time to take some of the grander ideas and move towards a separate total overhaul-type mod where you can freely add more techs, and thus vastly expand.

 

[Thalassicus]

Thank you for pointing that out! I hadn't got that DLC so I didn't know it added the terrace farm to that particular tech. I'll move the Agra Fort to Masonry, which follows precedent from some other national wonders (building and NW on the same tech). If it doesn't work out there I might put it on Metal Casting.

Modularity is something that's received a lot of thanks over the months so I do want stick with it. Standard industry practice follows a modular pattern (the CiV modding tools for example), and I'm used to it at work so it's more comfortable for me than alternatives. The reason we had to use an all-in-one approach for civilization mods in the past was due to difficulties with compatibility, which are much reduced in CiV (though the developers were still lazy about the lua implementation... compare it to wow's nearly seamless inter-mod integration).

 

[Txurce]

Thal,

In addition to the seeming CS influence bug I mentioned earlier, I am now getting a free unit in the capital every time I sell a city. I am playing with your CB mod, Free Tech, Attila, Cope's, CivWillard, RR Boost, and WWGD. I recall having had the same bug a couple of months ago.

 

[Thalassicus]

I'd completely forgotten it's possible to sell cities.

Shows how much I used it...

From your description it sounds like the 'give city' code likely calls the SerialEventCityCaptured event (last line of BC - General.lua), which means it's running the partisan mechanic. I hadn't thought to check if this event occurred for cities that were not technically captured. I think if I check "is the city in revolt?" that should hopefully identify if it was captured or given away... unless the game runs this event before adding revolt status, which will make it much more of a pain to solve.

Either way, thank you for pointing it out.

I've been experimenting with influence, the variables to modify it are very ambiguous and hard to use, so I haven't managed to get it right yet. I've been trying to achieve -25% influence from gold. Influence from quests was increased a lot a few release versions back, so they're very much worth completing now. For example, conquering a CS for a quest now gives the CS who asked it a much larger boost, to counteract the fact influence degrades faster with each CS you kill. Conquering any CS also gives an instant boost equal to 30 turns of friendship (I could buff or reduce this if it feels too much/little).

 

[Txurce]

I've been experimenting with influence, the variables to modify it are very ambiguous and hard to use, so I haven't managed to get it right yet. I've been trying to achieve -25% influence from gold. Influence from quests was increased a lot a few release versions back, so they're very much worth completing now. For example, conquering a CS for a quest now gives the CS who asked it a much larger boost, to counteract the fact influence degrades faster with each CS you kill. Conquering any CS also gives an instant boost equal to 30 turns of friendship (I could buff or reduce this if it feels too much/little).

In my opinion the level that it's at now makes allying with a CS prohibitive at Immortal until later in the game, and then only if you're successful - even the AI is doing it less. "Getting a maritime alliance asap" is not on the menu!

Let's see what others think as they experiment with the dev mod.

 

[Txurce]

Thal,

I just spent 1000g on Genoa, then noticed it gives me 5 influence. That's better than 0 at 500g, but still not really worth my while.

 

[Sneaks]

FYI:

I think you have Aqueduct and/or Baths at Construction too. So that makes 7-8 items on the Tech.

 

[V. Soma]

Let me have some thoughts on resources.

If I understand well in general, possessing iron favors the conquest player.
Now, if this player has iron and all other good thngs at hand in the cities,
that is, has iron PLUS lots of food and luxury resources,
then that palyer is in a winning situation.
He has both building and war success...

I believe it would be an interesting situation in the first phase of the game
if iron was to be located at places where less of other juicy things are present.
Thus the player would have to make interesting decisions:
should I go for luxury or iron? should I go for food or iron?

Therefore a concept idea:

Would it be OK if resources are to be tied to geography?
(Maybe it would be a plus for realism immersion, too)

My idea - if possible:

Make iron appear on hills, and only on those hills that are adjacent to a mountain tile.

This would perhaps strike a balance thing:
you can get iron, but take the hit of having a useless mountain near it...

Other resources also could have geographic specifications added...

 

[Tomice]

Thal,

I just spent 1000g on Genoa, then noticed it gives me 5 influence. That's better than 0 at 500g, but still not really worth my while.

This has to be a bug, no way Thal intended it this way.




@ V.Soma: This is already implemented, all ressources are placed with an intention. I've seen Thal talking about terrain-dependant placement for oil and horses, but from my experience also other ressorces are tied to a certain pool of terrain types.

 

[Thalassicus]

Thal,

I just spent 1000g on Genoa, then noticed it gives me 5 influence. That's better than 0 at 500g, but still not really worth my while.

Yeah, it's not intended to drop that low. Here's the issue... there's only two variables:

MINOR_CIV_GOLD_GIFT_GAME_MULTIPLIER
MINOR_CIV_GOLD_GIFT_GAME_DIVISOR

Logically you might think, gold * mult/div = influence, right? So changing it from the default 2/3 to 1/3 should halve influence gain.

It doesn't work this way, as you probably guessed. Changing it to 3/3 decreases influence gain, but not enough. So I tried setting it to 4/3 in the current dev version, and it decreased it too much! Trying values like 8/7 don't hit some good middle ground, either, and I've tried things like 2/1 or 2/4. These are variables in some longer, complicated formula hidden to us. Slowpoke tried figuring it out, though the answer he came up with doesn't work for me, so I asked him in his mod thread if he's figured out anything new. I've mostly just been guessing at random and trying stuff out to see what happens.

I believe it would be an interesting situation in the first phase of the game
if iron was to be located at places where less of other juicy things are present.

Would it be OK if resources are to be tied to geography?

This is how it's set up, I'll try and explain. Basically there's 2 things that create strategic and bonus resources:

• PlaceStrategicAndBonusResources
• PlaceSmallQuantitiesOfStrategics

The first does big deposits, second small ones. (Luxuries are handled elsewhere, I'm just giving these as an example.) The big-deposit one goes through actions for each terrain type. For each N tiles of a terrain type, it adds one strategic, and does so based on a probability between a list of resources. For example, this is the one for hills, and I've highlighted the important numbers:

local resources_to_place = {
{self.iron_ID, iron_amt, 26, 0, 2},
{self.coal_ID, coal_amt, 35, 1, 3},
{self.aluminum_ID, alum_amt, 39, 2, 3} };
selfrocessResourceList(22 * strategic_major_multiplier, 1, self.hills_list, resources_to_place)

What this means is for every 22 hill tiles on the map, pick a random tile, and add a resource with a 26% chance for iron, 35% for coal, and 39% for aluminum. Does this make sense? There's one of these sections for various types of terrain. For example, here are the ones for desert and grassland:

local resources_to_place = {
{self.oil_ID, oil_amt, 65, 0, 1},
{self.iron_ID, iron_amt, 35, 1, 1} };
selfrocessResourceList(6 * strategic_major_multiplier, 1, self.desert_flat_no_feature, resources_to_place)

local resources_to_place = {
{self.horse_ID, horse_amt, 100, 2, 5} };
selfrocessResourceList(33 * strategic_major_multiplier, 1, self.dry_grass_flat_no_feature, resources_to_place)

There's about 4 times as many resources placed in desert as on hills, and 6 times as much on flat desert than flat grassland. Basically, low-yield terrain has a tendency to give more resources to compensate. In addition, areas with lots of rough terrain like forests and hills will generate a lot of iron (whose units are better on rough terrain) while regions of open terrain will create lots of horses (good on open terrain).

The PlaceSmallQuantitiesOfStrategics function is much more complicated, but still based off geography, and favors more resources on bad terrain to compensate for the lower yields. Bonus resources are handled further down in the same function as major strategics. Sea oil has its own separate function, where it places a number of deposits equal to (0.75 * land oil total / sea oil deposit size). Basically it's a mess, but works.

 

[V. Soma]

Thanks for the detailed resource placement explanation, Thal.

I still would like irons be close to mountains, though, but OK, I understand the concept of the game as of now,
and it makes sense this way already, too.

 

[Txurce]

Yeah, it's not intended to drop that low. Here's the issue... there's only two variables:

MINOR_CIV_GOLD_GIFT_GAME_MULTIPLIER
MINOR_CIV_GOLD_GIFT_GAME_DIVISOR

Logically you might think, gold * mult/div = influence, right? So changing it from the default 2/3 to 1/3 should halve influence gain.

It doesn't work this way, as you probably guessed. Changing it to 3/3 decreases influence gain, but not enough. So I tried setting it to 4/3 in the current dev version, and it decreased it too much! Trying values like 8/7 don't hit some good middle ground, either, and I've tried things like 2/1 or 2/4. These are variables in some longer, complicated formula hidden to us. Slowpoke tried figuring it out, though the answer he came up with doesn't work for me, so I asked him in his mod thread if he's figured out anything new. I've mostly just been guessing at random and trying stuff out to see what happens.

What was wrong with the prior equation you were using? I didn't realize that CS balance was still in play. Probably the biggest issue there now is the road request not having a payoff, thanks to the post-patch bug.

Regardless, could we go back to the earlier, working version... or just tell me which folder needs replacing - Diplomacy?

 

[Thalassicus]

I've updated again. Version x9.13dev contains primarily changes to terrain, detailed in the CD and TI readmes.

Influence balance is something I've been working on from time to time. I've never really been happy with the -10% or so effect 3/3 had. Keep in mind these dev versions are just my current stuff dropped into a zip file, so it's not always going to be working properly.

The two variables are in the diplomacy mod's BD - General.xml file, halfway down. If you could, please test them at various combinations like 3/3, 10/9, and 5/4, and tell me how that impacts your current game. My game's at turn 300 and influence is still alright, so it sounds like you're in a later era.