Game balance proposals (very big post ahead)

There are in theory a number of reasons why chain denouncement can happen. Here are a few:
Spoiler :

// Does this player feel it's time to denounce ePlayer?
bool CvDiplomacyAI::IsDenounceAcceptable(PlayerTypes ePlayer, bool bBias)

int CvDiplomacyAI::GetDenounceWeight(PlayerTypes ePlayer, bool bBias)
{
// Base Personality value; ranges from 0 to 10 (ish)
// Hostile: Bonus
iWeight += 6;
// Afraid: Penalty
iWeight += -10;
// Unforgivable: Big Bonus
iWeight += 10;
// Enemy: Bonus
iWeight += 5;
// Competitor: Small Bonus
iWeight += 2;
// Good Relations: Penalty
iWeight += -10;
else if(eOpinion == MAJOR_CIV_OPINION_FRIEND)
iWeight += -25;
else if(eOpinion == MAJOR_CIV_OPINION_ALLY)
iWeight += -50;

// We are at war
iWeight += 2;
// Look for other players we like or are strong, and add a bonus if they've denounced this guy, or are at war with him
// We must not be on bad relations with eThirdParty
// If we're hostile or planning war, we don't care about this guy
// We're close to this guy who's at war - want to gain favor
iWeight += 1;
// Are they strong?
iWeight += (GetPlayerMilitaryStrengthComparedToUs(eThirdParty)
// Are we friends with them?
iWeight += 4;
// Rand: 0-5
// Used when friends are asking us to denounce someone
if(bBias)
iWeight += 3;
 
Awesome, simple but elegant mod! :goodjob:

If you're looking for input, I'd agree with the others who said that balancing Religious beliefs should be the next priority (esp nerfing Tithe, Religious Texts, Itinerant Preachers).


@YertyL: Try the mod out before you rate it! The Library has had one :c5science:Science added; what OP listed are the net changes, not the new total effect. The Library has been buffed, in other words, not nerfed as you suggest.
 
There are in theory a number of reasons why chain denouncement can happen. Here are a few:

Yeah, you can get denounced even without warmonger penalties, one easy fix is to tweak Leaders Denounce Willingness, I could take a look at that.

Awesome, simple but elegant mod! :goodjob:

Thanks! That was the idea, even a lot of them are minimal changes, all of them are well thought.

If you're looking for input, I'd agree with the others who said that balancing Religious beliefs should be the next priority (esp nerfing Tithe, Religious Texts, Itinerant Preachers).

Uhm... I tweaked some other beliefs, but yeah, those one are the most used, now there's more room for other options anyways.
 
Uhm... I tweaked some other beliefs, but yeah, those one are the most used, now there's more room for other options anyways.
Definitely; I'm trying out the new Ceremonial Burial in my current game as Indonesia.


Also, I have no idea how feasible this is, but I just saw this in another thread and loved the Archaeologists by :c5faith:Faith idea:
Honor Finisher:
Gain :c5gold: Gold from killing units. May purchase Great Generals with :c5faith: Faith.

Suggested Revision:
Gain :c5gold: Gold from killing units. May purchase Great Generals and Great Admirals with :c5faith: Faith.

I'm overall pretty happy with the finisher. I've moved all Great Admiral bonuses to Honor however, as this lets Exploration focus a little better. (I'd let Exploration buy Archaeologists with Faith instead.)
 
Yeah, you can get denounced even without warmonger penalties, one easy fix is to tweak Leaders Denounce Willingness, I could take a look at that.

Would you be willing to modify the core DLL? The point is that denounce willingness doesn't really help with the chain denouncements except in a crude way. The actual weights are *hard coded* mostly in:

int CvDiplomacyAI::GetDenounceWeight(PlayerTypes ePlayer, bool bBias)

Leader Denounce Willingness is in the weighted range (0-10)

At the end of GetDenounceWeight we see two critical weights:

  • Are we friends with the third party that has denounced our target? weight (+4) hard coded
  • A random weight (0-5) hard coded
  • Yet another bBias hard coded weight of (+3) for friends asking us to denounce target

Add up just those three factors 4+5+3=12 and just the simple fact that there are other AI's that are your friends as well means that the hard coded weight of (12) will wash out the denounce willingness number (10) and lead to AI's being very willing to denounce their target.

Digging into the DLL might be necessary at some point.

EDIT: in the short term while we don't modify the DLL, to fix chain denouncements, add more civs than normal so that friendships are unlikely to form.

If they are unlikely, then the third party denouncement weighting will not be washed out by the joint friendship weighting. Basically what is happening is that you the warmonger are isolated and the friendships between others strongly biases for chain denouncements on you. But if they were more in conflict with each other, they would leave you alone more and not chain denounce as much. So indirectly you could fix chain denouncements by making each leader more aggressive generally or make friendships less likely to occur or just jump in and modify the DLL where it is hardcoded.

At least that is what the code is telling me it seems.
 
I'm avoiding recompiling the dll on purpose for this mod. Not that I can't, just I want to make the patch lightwheight and easily compatible with next game patches. One thing I couldn't change due to this is the GP faith purchases as the branch relation is hardcoded: I would give the ingeneer purchase to liberty for example. Also I don't like how any building yield change due to policies is not changed on the tooltip, with the exception of happiness. Also I wanted to change how the warmonger penalty is calculated (based on population % loss of the target, less harsh in earlied eras where your civilization is less civilized, less hate from cs conquering), but only the estimated number of cities is the only exposed value of the formula.

Things I plan to check in the future:

- Cover on siege is not shown on combat preview, even I'm not sure it's working. This could be related with the no terrain bonuses, I'm not sure.

- Policy cost growth: There's a point in the game where you can't generate culture enough for a decent policy progression, until you start to generate GPs from guilds. There's a value on the defines that points when the policies cost is calculated exponentially instead of multiplicatively, and it now points at the third policy. Maybe delaying this value to the fourth would make policy acquisition more constant.

- Also I think exploration is a bit bland compared to other contemporary SP trees, I was thinking of giving a free caravel on a second policy, some science boost (+1 gold and +1 science per lighthouse, harbor, seaport), bonus naval unit production... I have to still give it a good thinking.
 
- Cover on siege is not shown on combat preview, even I'm not sure it's working. This could be related with the no terrain bonuses, I'm not sure.

That's odd. Battering Rams and Siege Towers have both no terrain and Cover, don't they?

- Policy cost growth: There's a point in the game where you can't generate culture enough for a decent policy progression, until you start to generate GPs from guilds. There's a value on the defines that points when the policies cost is calculated exponentially instead of multiplicatively, and it now points at the third policy. Maybe delaying this value to the fourth would make policy acquisition more constant

Definitively worth a try.

- Also I think exploration is a bit bland compared to other contemporary SP trees, I was thinking of giving a free caravel on a second policy, some science boost (+1 gold and +1 science per lighthouse, harbor, seaport), bonus naval unit production... I have to still give it a good thinking.

Science boost sounds good. Someone once suggested that perhaps Exploration could give a culture bonus to Landmarks, thus giving civs which go wide for the sake of landmarks a tourism bonus once you get Hotels and Airports.
 
That's odd. Battering Rams and Siege Towers have both no terrain and Cover, don't they?.

Yeah, and all siege have cover I available in the common promotion course, It's weird, maybe just is not showing it in the tooltip? someone experienced something similar?

Science boost sounds good. Someone once suggested that perhaps Exploration could give a culture bonus to Landmarks, thus giving civs which go wide for the sake of landmarks a tourism bonus once you get Hotels and Airports.

Well, not bad idea, I take note for it.
 
Really like the light-weight nature of the mod, though it might be tempting to go heavy duty (DLL changes etc.), keeping it relatively simple is a plus AFAIC.

Just a couple of thoughts:

1. You note that Windmill has increased production to 3, given that 'Coffee House' replaces that building for Austria, shouldn't that also get the modded extra production?

2. Just a typo in your OP, under the Religious changes you've state:
World church: Boosted to 1 culture per 5 followers instead of 5.

You're missing a '4' there by the looks of it.
 
1. You note that Windmill has increased production to 3, given that 'Coffee House' replaces that building for Austria, shouldn't that also get the modded extra production?
Thanks for the feedback: much appreciated!! Yeah I keep changes consistent in unique replacements, I missed this one, I will check this out!

2. Just a typo in your OP, under the Religious changes you've state: World church: Boosted to 1 culture per 5 followers instead of 5.

You're missing a '4' there by the looks of it.
Yeah, that's a typo, its 4. I'll change it.

Thanks!
 
Just the kind of mod I was looking for! Thanks a lot, I'm gonna go test it out to see if it plays as well as it sounds on paper.
 
OK, having played a few games with the new honor, there is one thing I would like to adress:
-I still feel like the happiness honor poilicy is a bit too powerful, or at least non-specific. How about changing it to: "Forges and Stables give +1 happiness, +1 culture; courthouses give +2 happiness"? This would reward conquest, but make the policy less powerful for self-built empires.

In general, the IMO biggest imbalance left unadressed at the moment is the complete domination of ranged units (and city attacks). For me, the relation between great war infantry and artillery is what it should feel like: Arty reliably inflicts damage, but is heavily damaged when caught by a meleed unit, dead when caught by two. G.W. infantry can tank quite a few shots, and dish out quite a bit of damage, but takes shots before being able to attack. You need both.

Cbows, Xbows and city attacks completely dominate, however. I would suggest nerfing both ranged and melee strength of archery units, as well as city ranged attacks, to roughly 3/4 or 2/3 of their current value, and see how it plays out. I would guess that we will still see archery units after this change, but maybe not only them anymore. Would it not be cool if knights and swordsmen were actually useful again? :)

EDIT: One more thing that has been bugging me: What do you think about moving infantry to combined arms? Plastics is already arguably the most powerful modern tech because of research labs; It does not need a melee unit that completely dominates its era on top of that. On the other hand, the lower half of the modern tech tree still feels very unattractive.

EDIT2: A similar thing: What do you think about moving the "lumber mill +1 production" from scientific theory to steam power? Again, Sci.Th. feels too powerful, Steam power far too weak. It would even make sense thematically - lumber for your steam engines.

EDIT3: I had a few more ideas and started a new thread: http://forums.civfanatics.com/showthread.php?p=12946312#post12946312
 
OK, having played a few games with the new honor, there is one thing I would like to adress:
-I still feel like the happiness honor poilicy is a bit too powerful, or at least non-specific. How about changing it to: "Forges and Stables give +1 happiness, +1 culture; courthouses give +2 happiness"? This would reward conquest, but make the policy less powerful for self-built empires.

My objective was to get an average of two happiness and culture per city, I think all ancient policies should help with culture and happiness, the results are similar compared to G&K, but unlike that honor version, you're done in classical, while in G&K you could get up to 5 local happiness per city.

I didn't want to focus honor on city conquest, I wanted to focus on helping building and supporting an army, then you decide how you will make a good use of it.

In general, the IMO biggest imbalance left unadressed at the moment is the complete domination of ranged units (and city attacks).

Yeah, I know ranged still dominate until artillery: with the siege unit cover free promo it gets a bit better, but is only marginal and doesn't address it. I didn't want to enter much into combat balance, but I have two ideas in my head from some time ago:

- Lowering total city power to about 2/3 of actual power, but rising city health to 150%. City walls would play a bigger role in city defense/health.

- Chaging the ranged damage formula, to add a power decrease based on target HP: the explanation is simple: units represent squadrons, you get more power in melee vs damaged because you outnumber them and battle is easier, but ranged on the other side, attacking a damaged squadron means there are less objectives, so is likely you kill less individuals with a single burst.

Those changes need lots of testing and the second involve dll edit, witch I tried to avoid in order to make a lightwheight bug-free mod.


I'll respond you there about this changes you propose. I didn't want to enter into tech modifications, other than a few wonder-related changes, but yeah, to me tech tree needs a big overhaul, moreover, science needs an overhaul, it's so dominant you end playing science most of the time.
 
My objective was to get an average of two happiness and culture per city, I think all ancient policies should help with culture and happiness, the results are similar compared to G&K, but unlike that honor version, you're done in classical, while in G&K you could get up to 5 local happiness per city.

I didn't want to focus honor on city conquest, I wanted to focus on helping building and supporting an army, then you decide how you will make a good use of it.

I can understand that, but I still feel like this single honor policy gives you roughly the bonuses of Liberty opener, Represantation and meritocracy in a single policy, at least short term. I feel that ATM if you compare liberty and honor total bonuses, there is too little motivation to go liberty for peaceful expansion.

I would prefer it if e.g. the honor opener gave culture for every destroyed unit, making it useful long-term, but to compensate this policy is nerfed down to roughly +1 +1 per city, more happiness long-term (which would be the case in my suggestion). If you do not want to force a player into conquest, you could still e.g. go for armory and military academy bonuses or something similar, but long-term.

Yeah, I know ranged still dominate until artillery: with the siege unit cover free promo it gets a bit better, but is only marginal and doesn't address it. I didn't want to enter much into combat balance, but I have two ideas in my head from some time ago:

- Lowering total city power to about 2/3 of actual power, but rising city health to 150%. City walls would play a bigger role in city defense/health.

- Chaging the ranged damage formula, to add a power decrease based on target HP: the explanation is simple: units represent squadrons, you get more power in melee vs damaged because you outnumber them and battle is easier, but ranged on the other side, attacking a damaged squadron means there are less objectives, so is likely you kill less individuals with a single burst.

Those changes need lots of testing and the second involve dll edit, witch I tried to avoid in order to make a lightwheight bug-free mod.

Yes, these changes sound like a good idea; though I still feel like a simple nerf of archer units' ranged and melee strength would make them less dominant without obsoleting them.


I'll respond you there about this changes you propose. I didn't want to enter into tech modifications, other than a few wonder-related changes, but yeah, to me tech tree needs a big overhaul, moreover, science needs an overhaul, it's so dominant you end playing science most of the time.

If I may ask: Why do you not want tech modifications? In my mind, the complete dominance of certain tech paths (looking at you, civil service-> education) is an equally large problem as the complete dominance of a social policy. Both force you into a certain playstyle.

In my mind, tech does not necessarily need a complete overhaul (though I would still lower library's percentual bonus from an effective 50% to 25%, e.g. 1 science per 4 citizens). Focusing on science techs should be the equivalent of teching straight to battle cruisers in SC2: Very powerful if you pull it off, but usually you are punished if you don't defend before that; plus, you give up map control, i.e. expansion space.

The problem is that such a large arsenal of "auto-defends" exist: Relatively passive AI, massive discouragement of wars and expansion, auto-leveling up of your cities, and defensive units that you do not need to invest beakers in, i.e. pikes and infantry. If these would be mostly eliminated, it would not be a problem if tech is powerful long-term, as long as it is risky short-term. That is the essence of most strategy games. IMO things like nerfing the auto-progression of ranged city strength, as well as balancing the tech tree in such a way that science techs really give only science, while you need to research other techs to defend yourself, would contribute to this.
 
...

Yeah, I know ranged still dominate until artillery: with the siege unit cover free promo it gets a bit better, but is only marginal and doesn't address it. I didn't want to enter much into combat balance, but I have two ideas in my head from some time ago:

- Lowering total city power to about 2/3 of actual power, but rising city health to 150%. City walls would play a bigger role in city defense/health.

- Chaging the ranged damage formula, to add a power decrease based on target HP: the explanation is simple: units represent squadrons, you get more power in melee vs damaged because you outnumber them and battle is easier, but ranged on the other side, attacking a damaged squadron means there are less objectives, so is likely you kill less individuals with a single burst.

Those changes need lots of testing and the second involve dll edit, witch I tried to avoid in order to make a lightwheight bug-free mod.



I'll respond you there about this changes you propose. I didn't want to enter into tech modifications, other than a few wonder-related changes, but yeah, to me tech tree needs a big overhaul, moreover, science needs an overhaul, it's so dominant you end playing science most of the time.

That sounds like good thinking to me, in general anyway. If there are two areas that could still do with tweaking it's the range attack, but particularly city range defence which is too high, though nerfing by 2/3 could be risky. One of the things I do like about Civ 5 is that you don't need to build a big army to protect your cities early on. You can risk a 'branching out' knowing that a city is vulnerable but still strong enough to make a would be attacker think twice about having a shot at one that appears an easy take.

The other area you mention (science/tech domination) is more of an issue IMHO. I'm relatively new to Civ5 so my opinion could be off, but I'm a veteran of previous games and especially Civ4 where I created a mod that adjusted a few things. One was adding a few key techs (with corresponding bits and bobs), but mostly by slowing the tech development and science points available relative to all other elements. I did this simply by increasing by an incremental % the number of science required to advanced. It required a corresponding tweak of the years per turn too so that you weren't still in the renascence in 2050 :D.

It's probably more than you would want to do with this mod, but it worked a treat with Civ4 mod that I built. Simply rushing to a particular tech became a risky thing because it was more costly and could mean losing out on other aspects because it required more of a concerted effort (where available). You were forced to live longer with the tech decisions you did make.

I've not thought it through too much with the Civ5 BNW version, and it could well fail badly. But after a handful of games, with and without mods - I'm already itching for something that takes the 'value' out of science a bit ... or maybe, more accurately, makes it more costly to advance through the tech tree and so makes other aspects more attractive. Kind of like what BNW introduced into the late game, but stretched right through the game. Not sure of your thoughts on this, but that’s my 2 cents worth.
 
If I may ask: Why do you not want tech modifications? In my mind, the complete dominance of certain tech paths (looking at you, civil service-> education) is an equally large problem as the complete dominance of a social policy. Both force you into a certain playstyle.

In my mind, tech does not necessarily need a complete overhaul (though I would still lower library's percentual bonus from an effective 50% to 25%, e.g. 1 science per 4 citizens). Focusing on science techs should be the equivalent of teching straight to battle cruisers in SC2: Very powerful if you pull it off, but usually you are punished if you don't defend before that; plus, you give up map control, i.e. expansion space.

The problem is that such a large arsenal of "auto-defends" exist: Relatively passive AI, massive discouragement of wars and expansion, auto-leveling up of your cities, and defensive units that you do not need to invest beakers in, i.e. pikes and infantry. If these would be mostly eliminated, it would not be a problem if tech is powerful long-term, as long as it is risky short-term. That is the essence of most strategy games. IMO things like nerfing the auto-progression of ranged city strength, as well as balancing the tech tree in such a way that science techs really give only science, while you need to research other techs to defend yourself, would contribute to this.

No, you are right, tech need modifications, but the main problem is that if it science or growth (science again, uh) related, the tech automatically is great. Civil service and Education are SO important because you usually are playing a catch-up game, and those techs will give you the power needed to get on the tech parity soon enough to get those great wonders of the renaissance.

And yeah, little empires with the tradition bonuses, super-cities and overpowered ranged units feel powerful enough to play defensive with minimun units. I couldn't be more agree on the Starcraft comparison: you take a risk and thus get a better position on the long term, but with current mechanisms whitch hurt wide and conquer, you take little risk with huge reward.

I've not thought it through too much with the Civ5 BNW version, and it could well fail badly. But after a handful of games, with and without mods - I'm already itching for something that takes the 'value' out of science a bit ... or maybe, more accurately, makes it more costly to advance through the tech tree and so makes other aspects more attractive. Kind of like what BNW introduced into the late game, but stretched right through the game. Not sure of your thoughts on this, but that’s my 2 cents worth.

In my experience not playing science is playing suboptimally, thus making things harder for no reason, and THAT is a big problem. If science buildings and other stuff didn't impact your science output so badly, then you could think about not going education from some time while going for medieval units and such, something unthinkable on the higher diffs.
 
but with current mechanisms whitch hurt wide and conquer, you take little risk with huge reward.

Yes so true. It is done that way because of the weakness of the AI at 1UPT.

The best way that I can see to fix the domination of tech is via tech diffusion from high-teching civs to low-teching civs. It means that high teching civs can get a badly needed leading techs but at the cost that other civs acquire the lower techs by chance. The lower teching civs run the risk of not getting an important leading tech, but might pick up lagging techs by chance.

The benefit of tech diffusion are great:
  • Brings in some excitement to low teching civs
  • has an opportunity cost to both low and high tech rates
  • brings in more diversity of game play
  • is realistic

The difficulties are:
  • getting the tech diffusion rate balanced
  • deciding which techs the low teching civs get
  • adjusting the AI not to always be a high teching civ

The interesting thing in terms of programming is to decide on what the dominant civs in the region are teching (top-middle-bottom tiers) and grant lower techs of the same tier to low teching civs. So if the dominant tech is at the bottom with military, the high tech civs are getting say longswords but at least the low tech civs are getting swords.

Tech diffusion has a long history in the civ series and would seem to work for civ5 it seems to me. Has not already someone done it? I think communitas mod does it but would have to check.
 
Nice! This addresses many of my balance grievances :)

Religion is probably still in need of better balancing. Tithes still seems to be king......
 
what about just decreasing city bombardment radius to 1 hex? it will resolve siege units problem. lower str of the archery units, give siege units cover...

without dll modding, iirc its possible to put damage limits on ranged units.
or just 'refund' some health after bombardment via lua script
different formula idea is great.. say in ranged combat you inflict base*(1/(100-hp)) damage:
base = 20 damage, enemy unit has 100 hp, resulting damage is 20
base = 20 damage, enemy unit has 50 hp, resulting damage is 10
base = 20 damage, enemy unit has 10 hp, resulting damage is 1
:goodjob:

about Honor, why should it give happiness at all? moreover, for the buildings you dont need in every city normally (and speeding up those super-useless military buildings)? why bonus for garrisons? it seems honor is designed as a tech tree for turtling, not for offensive war. what i suggest is to make it more agressive - you should get happiness, culture etc from war, not from building some crap and forcing your soldiers to sit on their asses.
1. finisher gold effect to some early policy ('Spoils of War'), plus a City Raider promotion for melee units (get gold from attacking cities)
2. military caste: scrap culture and happiness, give 5 maint free units
3. free barracks in first 4 cities
4. culture from conquest - it was an autocracy policy before bnw iirc, when you capture a city you get culture = its culture output * 10
5. why increase GG rate and give a free GG? i'm going to have many of them anyways as i'm going fight other civs, otherwise i wouldnt choose honor. Give some instant military units instead!
6. experience and discipline promotion... as other guy said earlier, human has advantage over AI in combat, no need to increase it further. Faster healing in own territory (lets name it 'Volunteers') would be much funnier. You capture a city, heal and go for another. Win a battle slower (maybe even lose a unit??), heal faster - much better than win fast and heal slowly.

Liberty
to combat budget deficit, maybe its better to move free garrisons from Tradition to some liberty policy ('Militia') - Tradition dont need it anyways. And what about some gold per city (gold from walls - safer trading space)? Tradition's free buildings may be balanced by faster production of buildings like granary, lighthouse, walls ('Public Goods'). i think it would fit the theme if liberty was given tile-purchase discount (tradition expands through culture while liberty does it through gold).

about science,
1. allow academy build at a later stage (architecture?)
2. remove pikes and infantry from the upper tech tree branch (pikes to guilds, infantry to combined arms)
3. +2 science per library is a great idea to help wide empires, but wont it be better to nerf science buildings instead? library effectively gives +50% to science, no other building is comparable (except NC). btw NC may be pushed further too.
4. tech diffusion.. or better rate of stealing techs which is easier to mod i assume. it also should be easy to increase a bonus to research for neightbours knowing a tech.
 
Top Bottom