Another thread about capturing cities..

[EsoEs]

Conquest by its nature is going to be more powerful than any other strategy available to the player because it is the only strategy where the players actions directly affect the AI. Everything else you do in this game affects solely the player, while warring affects the player and the AI. There is no peaceful way to damage/slow down an AI economy. So while some people may be averse to conquest, the truth of the matter is that it is inherently going to be more powerful. You cannot balance conquest/ non conquest strategies atm, it is simply not going to be possible. We can try to limit conquest (happiness, low gold/science yields), but then again we've gone so far down this road I think it was overnerfed, which was why I even began the thread in the first place.

 

[Seek]

Just to clarify; I haven't followed the latest discussion very closely, but in general I was in favor of reducing the food bonuses for the AI, and giving them other bonuses instead.

I would much prefer a situation where the AI cities are the same size as human cities, but are more productive per person, than one where the AI just gets larger cities.

This would be fine with me, but Thal cut the AI food without compensating science, production, or anything else. This is essentially the crux of the matter for me; since more population means more of everything else, the food cut was a general nerf to the AI and I don't think that it was necessary or desirable.

 

[Txurce]

This would be fine with me, but Thal cut the AI food without compensating science, production, or anything else. This is essentially the crux of the matter for me; since more population means more of everything else, the food cut was a general nerf to the AI and I don't think that it was necessary or desirable.

I know Thal wants to let the beta play out for a while this time around, but this seems to need addressing before the next GotVEM.

The reason I love the idea of using gold as the primary horizontal limiter rather than happiness is that an increasing gold deficit would start disbanding your army while not causing some bizarre slowdown of population growth in core cities. Again, the AI could handle this relatively easily as it has the funds to absorb the costs, and as it is a passive mechanism, it also would not require really any work in terms of coding AI behavior.

I wish Thal had tried this approach (which seemed to be where the thread was headed) instead of the food nerf. I think it worked out that way because of the city-capture issue*. The main reason I would have wanted to try it is that the gold approach offers the human more choice, while remaining an effective passive tool for the AI. It keeps the happiness mechanic intact, just allows gold to be used (as it already is) to buy happiness. It would not make the game much easier for the human, but it would make it less frustrating. And it might give the AI a welcome, technically non-exclusive edge: occupying a city and paying through the nose to have it be productive more quickly. That's the sort of needle worth trying to thread.

Now that we are where we are, I don't know if it's easier to buff Science and production, or to try the gold fix.

* It's worth remembering that Civ5 is not meant to be a game about conquering large numbers of cities. That is specifically why the devs changed Conquest from being about land to being about capitals. When you start altering this core concept - including adding the 30% land rule for the hell of it - you are on your way down a slippery slope of re-balancing a lot of peripheral factors.

 

[Ahriman]

I have mixed feelings about decoupling conquest from domination.

Maybe we have different interpretations of what decouple means, but in my view, it is impossible to not decouple conquest from domination.
What I mean by that is: if you attempt to make domination easier by making conquest easier, then you will also make other victory conditions easier. This isn't a choice, this is inherent to how the game works.
* * *

Conquest by its nature is going to be more powerful than any other strategy available to the player because it is the only strategy where the players actions directly affect the AI. Everything else you do in this game affects solely the player, while warring affects the player and the AI. There is no peaceful way to damage/slow down an AI economy. So while some people may be averse to conquest, the truth of the matter is that it is inherently going to be more powerful. You cannot balance conquest/ non conquest strategies atm, it is simply not going to be possible. We can try to limit conquest (happiness, low gold/science yields), but then again we've gone so far down this road I think it was overnerfed, which was why I even began the thread in the first place.

I would much rather have a game where conquest was roughly balanced as a playstyle with other playstyles and the Domination victory condition was hard to achieve than a game where the Domination victory was equally easy to achieve as other victory conditions but warfare and conquest was clearly the most effective playstyle to adopt.

I think it is weird to argue that conquest is inherently always going to be the best way to play - your main point in this thread was that you thought domination was too hard to achieve because conquering wasn't advantageous. So it is easy to weaken conquest as a strategy through city capture mechanics.
* * *

This would be fine with me, but Thal cut the AI food without compensating science, production, or anything else. This is essentially the crux of the matter for me; since more population means more of everything else, the food cut was a general nerf to the AI and I don't think that it was necessary or desirable.

I'd be fine with boosting science and production bonuses in rough proportion to the food bonuses that were removed.
I agree that a general nerf wasn't called for.

 

[Jorlem]

For the new 30% land requirement for a domination victory, is that 30% of all owned land, or 30% of the total available landmass?

 

[Txurce]

And is there any way to estimate besides eyeballing?

 

[tlaurila]

Could there also be an OR domination to capital sniping: Something like own as much land as all other players combined and twice as much as the next biggest Civ, and you've won.

 

[Thalassicus]

Think of the AI growth bonus as a "tallness factor." Lower means they build wider empires, and higher leads to tall empires. Since population is limited by happiness either way, it does not strictly make the game harder or easier, other than the AI's inherent aptitude at playing a tall or wide empire. Reducing the bonus makes wide and conquest AIs more successful. I changed the growth modifier primarily for this reason -- it helps the AI expand. I'm not too concerned about how large/small AI cities are when captured, since we can raze cities down to a size we want. Improving science is different from improving growth, because it's a direct and unrestricted AI boost regardless of what playstyle they follow or how much happiness they have.

To put it another way...

• Altering the population, resistance time, and buildings in captured cities changes city capture.
• Altering AI bonuses changes AI leaders.

These two things do have some overlap, but the recent changes are targeted at different things (puppet/occupy strength, and AI wide/tall favorability).

These are the AI initial bonuses:

Spoiler :

local isBold = (GameInfo.Leaders[player:GetLeaderType()].Boldness > 5)
local teamID = player:GetTeam()
if isBold then
    player:ChangeGold(10*handicapID)
end
for _, adjPlot in pairs(GetPlotsInCircle(plot, 1, handicapInfo.AICapitalRevealRadius)) do
    adjPlot:SetRevealed(teamID, true)
end

-- 1 settler
-- 2 chieftain
-- 3 warlord
if handicapID >= 4 then -- prince
    if isCoastal then
        log:Debug("Is Coastal")
    end
    M_InitUnitClass(player, "UNITCLASS_WORKER", plot)
    M_InitUnitClass(player, "UNITCLASS_ARCHER", plot)
end
if handicapID >= 5 then -- king
    if isCoastal then
        M_InitUnitClass(player, "UNITCLASS_WORKBOAT", plot)
        M_InitUnitClass(player, "UNITCLASS_TRIREME", plot)
    elseif isBold then
        M_InitUnitClass(player, "UNITCLASS_WORKER", plot)
    end
    capitalCity:SetNumRealBuilding(GetUniqueBuildingID(player, "BUILDINGCLASS_MARKET"), 1)
end
if handicapID >= 6 then -- emperor
    if isCoastal then
    end
    if isBold then
        M_InitUnitClass(player, "UNITCLASS_WORKER", plot)
    end
    M_InitUnitClass(player, "UNITCLASS_ARCHER", plot)
end
if handicapID >= 7 then -- immortal
    if isCoastal then
        M_InitUnitClass(player, "UNITCLASS_WORKBOAT", plot)
        M_InitUnitClass(player, "UNITCLASS_TRIREME", plot)
    elseif isBold then
        M_InitUnitClass(player, "UNITCLASS_WORKER", plot)
    end
end
if handicapID >= 8 then -- deity
    if isBold then
        M_InitUnitClass(player, "UNITCLASS_CATAPULT", plot)
    else
        M_InitUnitClass(player, "UNITCLASS_ARCHER", plot)
    end
end

AIs also have other bonuses which depend on difficulty level. These are the handicap settings on Emperor:

Spoiler :

0
050 StartingLocPercent
080 AdvancedStartPointsMod
010 HappinessDefault
100 NumCitiesUnhappinessMod
100 PopulationUnhappinessMod
000 Gold
000 GoldFreeUnits
100 ProductionFreeUnits
050 ProductionFreeUnitsPopulationPercent
002 ProductionFreeUnitsPerCity
100 RouteCostPercent
100 UnitCostPercent
100 BuildingCostPercent
100 ResearchPercent
100 PolicyPercent
100 ImprovementCostPercent
002 CityProductionNumOptionsConsidered
002 TechNumOptionsConsidered
002 PolicyNumOptionsConsidered
0-1 AttitudeChange
030 NoTechTradeModifier
025 BarbCampGold
000 BarbSpawnMod
010 BarbarianBonus
010 AIBarbarianBonus
030 EarliestBarbarianReleaseTurn
007 BarbarianLandTargetRange
018 BarbarianSeaTargetRange
000 AIStartingDefenseUnits
000 AIStartingWorkerUnits
000 AIStartingExploreUnits
100 AIDeclareWarProb
075 AIWorkRateModifier
100 AIUnhappinessPercent
085 AIGrowthPercent
057 AITrainPercent
100 AIWorldTrainPercent
057 AIConstructPercent
100 AIWorldConstructPercent
057 AICreatePercent
100 AIWorldCreatePercent
057 AIBuildingCostPercent
050 AIUnitCostPercent
040 AIUnitSupplyPercent
050 AIUnitUpgradePercent
0-4 AIPerEraModifier
160 AIAdvancedStartPercent
021 AIFreeXP
003 AIFreeXPPerEra
PROMOTION_HANDICAP_I AIFreePromotion
006 AICapitalRevealRadius

It is not at all clear to me that this is true. Can't you just have a global science yield modifier for AI players, based on difficulty levels (and maybe, as you suggest, by era)?

The handicap list does not have a variable to alter AI science rates. I can still do it, but it requires a few hours of work instead of a few minutes. I've got it on my todo list for the next beta cycle.

 

[Ahriman]

Think of the AI growth bonus as a "tallness factor." Lower means they build wider empires, and higher leads to tall empires. Since population is limited by happiness either way, it does not strictly make the game harder or easier,

This makes no sense to me at all.
Removing a significant bonus will absolutely make the AI weaker. Removing a growth bonus means less population means less science, production, gold, etc.
It is just not true to say that the growth bonus only controls tall vs wide.

The AI is very often *not* bound by happiness, which is the only mechanism by which reducing growth could favor expansion, and with +4 unhappy per city, then with a given amount of happiness the AI will be noticeably weaker in a wide situation to a tall one.

So I think you're wrong here.

Lab it, if you're curious; run an AI game with the bonus, without the bonus, and look at various economy figures. I would be amazed if removing effectively a free aqueduct in every city didn't have a huge impact.

The handicap list does not have a variable to alter AI science rates. I can still do it, but it requires a few hours of work instead of a few minutes. I've got it on my todo list for the next beta cycle.

Great, I think it would be worth it, IMO this is the right move to make to compensate for smaller population.

 

[Txurce]

Lab it, if you're curious; run an AI game with the bonus, without the bonus, and look at various economy figures. I would be amazed if removing effectively a free aqueduct in every city didn't have a huge impact.

I was thinking along the same lines, and am curious to learn the results.

 

[Thalassicus]

Faster growth means more population, which lowers happiness, and gives less capability to expand. The AIs play by the same happiness rules as the human in VEM but does not appear to use "avoid growth." I tested the variable while running autoplay games for Game 2 of the GotM over the past week.

This means faster growth is a bonus for AIs which like playing tall, and a penalty for AIs who like to expand or conquer. Since most AIs like to expand or conquer, lowering the bonus helps three out of four AIs.

 

[Txurce]

Thanks for testing it. In my games India, China, Siam and Korea tend to go tall most often, but I've seen Siam in particular go wide when given an opportunity. I'd suggest buffing these civs in some way, but the RNG factor in game approach would often make it moot.

And I really can't wait to try the science buff/misc. nerf to the AI. It should be reasonably evident in King, right?

 

[Thalassicus]

I think it should be clear at King level, yes. I've posted about it on the Difficulty Levels thread.

 

[Ahriman]

This means faster growth is a bonus for AIs which like playing tall, and a penalty for AIs who like to expand or conquer.

This sounds like an assertion to me that isn't necessarily linked to evidence. In your autoplay, were most civs really further ahead in tech and gold and production and population without the growth bonus?
If so, that indicates that the city AI is really, really bad. A bonus should never make you worse off; if nothing else you could work fewer farms and more mines.

 

[Thalassicus]

The AI doesn't generally plan ahead. For example, it does not appear to be programmed to recognize it needs to limit growth now because it will run out of happiness in the future due to expansion/war. When I watch the happiness chart in InfoAddict it continues growing as fast as possible until it reaches the happiness cap. This limits successful expansion and conquest. AI expansion noticeably increased when I reduced the growth rate.

 

[EsoEs]

The AI doesn't generally plan ahead. For example, it does not appear to be programmed to recognize it needs to limit growth now because it will run out of happiness in the future due to expansion/war. When I watch the happiness chart in InfoAddict it continues growing as fast as possible until it reaches the happiness cap. This limits successful expansion and conquest. AI expansion noticeably increased when I reduced the growth rate.

So not only are we buffing AI expansion, we're giving a science bonus across all levels of play as well? I thought that was to compensate for the supposed population "nerf", when in fact we're buffing them twice, in major areas of AI play, expansion/science.

The problem with this in my mind is that the "runaway" AI is going to be receiving the same bonus to his score as all the other players, and a %bonus is more valuable the higher the score. We seem to be trying to create an AI superpower, rather than a balance of stronger/smaller nations.


The human player was compensated with easier and more profitable conquest(less resistance times/unhappiness due to decreased AI populations), but now conquest is even more of an imperative than it was before. Stopping the AI "superpower" is now going to be vital to any victory, forcing the human player to wage war.

What is going on here??

 

[Thalassicus]

v110
WWGD

• Added a science bonus for AIs based on difficulty level, and reduced other AI bonuses.

The buffs are balanced with nerfs.

The goal is an even level of challenge and importance in decision-making from start to finish. Most Civilization games start hard and get easier to the end. I don't like when the first ~50 turns determine the outcome of the game, and the rest is just playing out an assured victory.

 

[gdwitt]

Why not nerf the AI happiness bonuses directly?
For ex., I usually see that the large expansionist civs still maintain a happiness over 50 in emperor games, despite having multiple puppets and over 15 cities.
I can barely keep happiness over 10.
They must be in continual golden ages.
I would double the penalty for keeping city in puppet state.

 

[Txurce]

If the AI's happiness bonus were a problem, then this might be a good idea. But since it tends to lose most of the time to most of the players using VEM, a nerf like this one would require a compensating buff in some other area. That is part of what I think VEM is trying to do with AI performance: create a balance where no single aspect of its overall bonus stands out in any era.

 

[Ahriman]

The AI doesn't generally plan ahead. For example, it does not appear to be programmed to recognize it needs to limit growth now because it will run out of happiness in the future due to expansion/war.

On what difficulty level?
On high difficulty levels I generally notice that AIs retain high excess happiness. This is why I am suspicious that removing the food bonus makes them better off. It might make them build more cities, but that is not the same thing as making them more powerful.