A core set of balance changes

Celevin · Oct 17, 2010

Ahriman said:
Because the fixed happiness bonus from your luxuries, many social policies (protectionism, cultural diplomacy, legalism), difficulty level and wonders is being spread over more and more cities, so the average happiness per city is declining.

Well, my point was that when you pass the threshold of +3 additional happiness per city, you should always expand from a happiness perspective. You can grab that through policies, or just having enough gold to build a colloseum in every new city. All it is is an opportunity cost. There's no more barriers to expanding your empire, gaining a higher science and production. In Civ4, you needed more than just build time to make a city profitable: you needed the write techs or civics, like additional trade routes. The moment you unlock colloseums in Civ5 it's a good idea to start rapidly expanding.

(I'm ignoring policies for now - the goal is to have two working slowdowns to ICS, not just policies)

Ahriman said:
Still, its hard to get a formula that doesn't feel arbitrary and forcing (Thou Shalt Have This Many Cities).
Even supposing we could deal with fractional happiness (which is messy), what kind of formula would you propose?
Maybe something like marginal unhappiness from the Nth city = 1.5 + N/K
Where K depends on map size.

Anything where the rate of change of our function is positive will work. Yours works nicely, after quickly graphing it. If we wanted to modify it, we could slow it by doing (N/K)^.8 or something. I like yours because it sticks to rationals, and is really easy for the player to understand. We could even put in the UI happiness screen: "1 additional city will add x happiness".

Ahriman · Oct 17, 2010

I would imagine something like K = 6 on standard size, 9 on large, 12 on huge. Might be a bit too low.
Use the same relative ratios as the policy costs modifiers (which IIRC are 0.3 on standard, 0.2 on large, 0.15 on huge).

The change in the marginal impact still reasonably easy for the player to interpret; every new city has a marginal happiness cost 1/K higher than the previous one.
But the cumulative unhappiness is still a bit confusing for a lot of users.

Celevin · Oct 17, 2010

The value most players will need is the marginal value. It's hard to get a good system that has good marginal and cumulative values.

I'm concerned over the N part. Comparing going to 3 cities and going to 6 cities, it's quite the leap in cost. It might punish big empires too much.

In Civ4, commerce was very easy to get. In Civ5, there might be a point on a map that even with a lot of policies and buildings opened up, marginal unhappiness will be larger than marginal happiness.

Maybe we should make it always increase, but make it asymptotic to some high unhappiness number? That way eventually marginal unhappiness has a small cap so we never get a situation where people just aren't able to build new cities.

Ahriman · Oct 17, 2010

Comparing going to 3 cities and going to 6 cities, it's quite the leap in cost.

Well, it depends on K.

If K= 12, then with less than 12 cities, total unhappiness from # cities is less under this system than it is under the standard one. [And the 6th city only costs 0.25 more happiness than the 3rd city).

So you can still get any slope you like by setting K appropriately.

Assuming K = 12
city marginal cumulative
1 1.583333333 1.583333333
2 1.666666667 3.25
3 1.75 5
4 1.833333333 6.833333333
5 1.916666667 8.75
6 2 10.75
7 2.083333333 12.83333333
8 2.166666667 15
9 2.25 17.25
10 2.333333333 19.58333333
11 2.416666667 22
12 2.5 24.5

Celevin · Oct 17, 2010

I tried another function in order to make it asymptotic. To clarify from scratch the goal of this:
- Marginal unhappiness increases with number of cities
- Marginal unhappiness has a soft cap of X

We want the unhappiness soft cap to be big enough such that someone can't do ICS, but small enough that they never feel they can't profit from another city.

I chose the function with the following parameters. The values beside them are the parameters I used as an example for the graph. The base exponent "smoothes" the function by lowering each marginal unhappiness by a bit as it's lowered. -1/4 seemed like a good starting number.

Base Exponent: -1/4
Max - Min Unhappiness Spread: 4
Max Unhappiness: 6

The equation used was:
MARGINAL UNHAPPINESS = (Max Unhappiness) - (Max - Min Unhappiness Spread)*(# of cities)^(Base Exponent)

Here are the results:

Number of Cities Marginal Comulative
1 2.000 2.000
2 2.636 4.636
3 2.961 7.597
4 3.172 10.769
5 3.325 14.094
6 3.444 17.538
7 3.541 21.079
8 3.622 24.700
9 3.691 28.391
10 3.751 32.142
11 3.804 35.945
12 3.851 39.796
13 3.893 43.689
14 3.932 47.622
15 3.967 51.589
16 4.000 55.589
17 4.030 59.619
18 4.058 63.677
19 4.084 67.761
20 4.109 71.870
21 4.131 76.001
22 4.153 80.154
23 4.173 84.328
24 4.193 88.521
25 4.211 92.732

As said before, as the number of cities approaches infinity, the highest value the marginal unhappiness can be is 6 in this example.

Thalassicus · Oct 17, 2010

Thank you for linking this thread Ahriman, I hadn't noticed it before and will take a look.

Update: I agree with most of your points, have done about half this and will get the other half done. Thank you for the detailed list.

1. Weaken puppets.
Option A
Block puppet AI from building military structures, forge, or defensive structures. They build only culture, happiness, gold, food/growth and science boosts. Puppets never construct buildings with strategic resource requirements.
Add a flat penalty (probably 30%) to the culture, science, hammer and gold output of puppet states. This makes them much less efficient, per unhappiness point, than regular cities.

Option B
Make puppets stay in permanent unrest unless they have a military unit stationed in the city.

This might be possible with Lua changes, though it will be much easier once we have C++ access. I'll put it on my todo list.

2. Make great scientists produce a fixed number of beakers, that increases slightly with tech era. The effects of other great people doesn’t scale nearly so much.
Make research agreement costs increase by era proportionally to how beaker costs increase. Its lame that a cheap tech costs 250 gold in the early game, and an expensive tech costs a similar amount of gold in the late game.

#1 is not possible before we have C++ access.

I increased research pact cost 25% in Gradual Research. They do increase each time you use them. Changing research pacts by era would be significantly easier with C++, but I think I can do it with Lua and the DB.Query() function... I'll put it on my todo list.

3. Increase tech beaker costs by era for the later eras, with later eras increasing by more. Eg: industrial by 10%, modern by 20%, future by 30%.
Change the tech tree so that replaceable parts requires rifling, and Electronics requires Combustion (no mech-inf before tanks please).

Did #1 with Gradual Research. The tech tree changes are not possible with the current structure of the tree, links don't cross.. though if you skip stages of units your earlier units cannot upgrade, so that does place a "soft" restriction on tech progress.

Tanks are buffed in Balance - Units.

4. Unit balancing.
i) Reduce the strength of the horseman and companion cavalry by 1, reduce their movement speed by 1.
ii) Increase the stats of the ironclad. It is pathetically weak.
iii) Nerf artillery somehow, they are far too strong. Either reduce ranged strength (probably the simplest) or reduce their range (but leave with indirect fire) or remove indirect fire (but keep 3 range).

i) Reduced horseman and CC strength by 11% in Balance - Units. I'm trying to take a gentle approach with them at first... if there's a feeling more of a changes is needed feedback would be helpful.
ii) Did this in Balance - Units.
iii) One point to consider with Artillery is that by going that tech path, you lose out on having Factories sooner. Siege are also of questionable usefulness compared to simple melee units, though that might be due to a poor combat AI. I might add a change to Artillery... possibly -10%

.

5. Improve defensive structures. Defenses should be meaningful. At the moment, they aren't. This would also tend to reduce the problem where as soon as an AI loses its army, you can rapidly conquer all its cities.
Option A: Increase Strength of all defensive buildings by 50%

Option B: Remove maintenance cost from all defensive structures.

Option C: Have military base give +1 range to city bombardment.

Afforess did some parts of this in Active City Defense, I did other parts in Balance - Buildings and Wonders.

6. Change maritime city states. Currently, cultural and military city states provide fixed bonus irregardless of empire size, but maritime give a bonus per city, and that bonus is large. It favors large empires and makes small cities grow very rapidly.

Option A: Make MCSs give +X% food in all cities. Eg X = 5%, increases 0.5% each era. (Double this if allied).

Option B: Make MCS give a flat X food, allocated across all your cities somehow. Eg X = 6 food, increases by era. (Double this if allied)

Option C: Make MCS give +X food across the largest Y cities (cycling back to the original cities if Y > number of cities). Eg X = 2, Y = 3, Y increases +1 per era. (Double this if allied.)

Option D: Make MCS give +X good per population point. Eg +0.15 food per population (double this if allied).

7. Rebalance cultural city states, so their effect is not destroyed so rapidly as you gain cities.
Rather than flat +X culture, have them give +X culture +Y/2 culture per city, where Y is the % increase in social policy cost from adding cities (30% on standard map size, 20% large, 15% huge).

These aren't possible without C++ access (or would be very difficult to implement without it) but I did reduce the Maritime bonus and increased the Military spawn rate in Balance - City States. I'll revisit this once we have C++ access.

8. Make it easier for cities to grow to larger sizes, so it is easier to have a large population with a few megacities, rather than just lots of medium sized cities.

Option A:
Reduce the amount of food needed to grow to larger city size at the higher end. Currently it takes roughly 4x as much cumulative food to reach population 14 as it does to reach population 7, this is too punitive.

Option B: Reduce the hospital storage bonus from 50% to 25%, and reduce hospital hammer and maintenance cost by half. Add an Aqueduct building at Engineering tech that has 25% food storage, and costs half as much as a hospital.

Interesting idea in option B, and it can be done with the currently available tools. I'll put it on my to-do list for a new mod.

9. Rebalance strategic resources so that they are actually rare, and meaningfully binding. Cavalry and heavy infantry should feel elite, and not be the core of the army.

Option A
Make all horses produce only 2 copies, they are far too common. All other resources provide either 2 or 4.

Option B
Limit resources to 2 but have them increased slightly by tech. Eg: Steel (or Metallurgy?) increases iron by +2, economics increases horses by +2, railroad increases coal by +2, combustion increases oil by +2, nuclear fission increases uranium by +2.

Interesting ideas. A would be easiest, but I think I can do B in Lua. I'll put it on my to-do list.

9. Rebalanec improvement yields.
Some possibilities:

i) Make civil service apply only to farms on open terrain next to rivers/lakes. No more super hillfarms. Fertilizer boosts them as normal.

ii) Increase mine yields +1hammer at dynamite.
Make farm yield boosts also affect pastures, so that bonus resources are actually decent bonuses.
Maybe fishing boats +1 gold at refrigeration?

iii) Increase the yield of great person improvements with techs. Eg: Education increase Academy yield +2 science, Metallurgy increases Manufactory yield +2 hammers, Acoustics increases Landmark yield +2 culture, Banking increases customhouse yield +2 gold. These things scale really badly.

i) This will require C++ access.
ii) While a Dynamite bonus is tempting from a realism perspective, I feel it overpowers the tech due to already containing Artillery. This is why I went with a +1

bonus at Engineering for Balance - Terrain Improvements, and detailed more reasons http://forums.civfanatics.com/showthread.php?t=385083#post9696532 here. Slight buffs to bonus resources are also included in Terrain Improvements, and a slight buff to water tile gold with the Harbor in that balance mod.
iii) Tech bonuses are interesting, I'll think about it a while. Overall, I feel a simple buff is somewhat less complicated while still achieving the desired effect, which is done in Terrain Improvements.

10. Rebalance the luxury goods, so that dyes, sugar, silk, spices, furs aren't inferior.
Add a weaver building (at what tech??) that gives +2 gold to dyes, silk, furs, like the +3 the mint gives for gold and silver.
At +1 gold with currency tech to the tile yield of sugar and spices.

Interesting ideas. I'd avoided boosting luxury resources because they're already quite useful, but these are good ideas. I'll put it on my to-do list.

11. Social policy rebalancing.
Increase collective rule bonus from 50% to 75% (or maybe just size 2?).
Increase trade unions to 25% cost reduction.
Humanism gives +2 happy with university, not +1. Or gives +1 with university, +1 with public school.
[Universal suffrage is also weak, but might be ok with the boost to walls).
[Fascism is weak, but will be fixed by strategic resource rarity reduction.]
[Puppet nerfs boost Police state.]

Did this and some more in Balance - Social Policies.

12. Various AI changes.
a) Reducing AI willingness to pay for excess luxuries, strategic resources, and cities.
b) Improving combat AI to be more careful about exposing its troops to enemy fire (especially flat terrain and tiles exposed to ranged fire), and to keep a reserve guard, and to heal units better, and to target embarked units with ranged fire.
c) Increase AI willingness to spend gold pursuing city state alliances, including out-bidding the human player.

These don’t solve all the problems, but they’re cleaner than totally rewriting the whole economy system. Any other changes (like changing core tile yields) will tend to require the entire economy system to be changed.

Elvince did some of this in AI tweaks - Battle AI. Improving the tactical AI will require C++ access. I don't remember where I saw it, but another AI mod did some changes to AI trade weighting.

In addition to the points you made, it seems civilizations could use a little balance as well, so I did that in Balance - Civilizations.

Ahriman · Oct 17, 2010

I tried another function in order to make it asymptotic. To clarify from scratch the goal of this:
- Marginal unhappiness increases with number of cities
- Marginal unhappiness has a soft cap of X

An asymptotic soft-cap seems reasonable.
I would add a third though: simple to understand.

This one isn't *too* bad, but its not simple to people who don't understand negative exponents.

It is pretty painful though; at 10 cities its 12 unhappiness worse than vanilla. Thats a very large difference. At 15 cities its 22 unhappy worse than vanilla. At 20 cities its 32 worse.
That seems too much. I'd drop the base exponent to -0.2, maybe even less.

[At -0.2, 10 cities is 10 unhappy worse than vanilla, 15 cities is 18 worse than vanilla.]

Which parameter would you adjust to change map size, the Max unhappy or the exponent?

Celevin · Oct 17, 2010

Ahriman said:
An asymptotic soft-cap seems reasonable.
I would add a third though: simple to understand.

This one isn't *too* bad, but its not simple to people who don't understand negative exponents.

It is pretty painful though; at 10 cities its 12 unhappiness worse than vanilla. Thats a very large difference. At 15 cities its 22 unhappy worse than vanilla. At 20 cities its 32 worse.
That seems too much. I'd drop the base exponent to -0.2, maybe even less.

[At -0.2, 10 cities is 10 unhappy worse than vanilla, 15 cities is 18 worse than vanilla.]

Which parameter would you adjust to change map size, the Max unhappy or the exponent?

Exponent. Max unhappy really hurts to adjust, and is mainly set for how much you want to hurt max-sprawled people mid game. If we set it too high or too low, it'll cause either no expansion or ICS.

I agree it's hard to understand, sorta like maintenance in Civ4. The two key qualities to making it easier are there though: that people can bet on marginal unhappiness always increasing, and that it will never go beyond 6.

We solve the world scaling issue if we make more or less unique luxuries popping depending on world size.

I'm going to do up a before-and-after theoretical tier list to see if we're adjusting something that's really going to send a civ out of whack.

Ahriman · Oct 17, 2010

Thalassicus, thanks for checking this out, and great that you're considering implementing some of this.

The tech tree changes are not possible with the current structure of the tree

Does the game hardcode having to have arrows for all AND required techs? In Civ4 it could still work with just one arrow (and the other tech image would appear as a requirement).

They do increase each time you use them

Are you sure? It feels like they increase depending on how many active you have at a time, but I could be imagining that. Regardless, they're all far too cheap. A 25% change isn't going to make much difference. The cost needs to scale massively over the course of the game, like tech costs do.

Reduced horseman and CC strength by 11% in Balance - Units. I'm trying to take a gentle approach with them at first... if there's a feeling more of a changes is needed feedback would be helpful.

I've come to the conclusion that the problem is the movement speed, not the strength.
Compare Cavalry and Riflemen, then compare Horsemen and Swordsmen.

i) This will require C++ access.

It could be fudged by creating an entirely separate Hill Farm improvement.
Hill Farms are only buildable on hills by rivers, and are boosted only by fertilizer. Normal farms are never buildable on hills.

Slight buffs to bonus resources are also included in Terrain Improvements,

I think some of these are too much, in particular the +1 food for fresh water farms with fertilizer. 5 food yield on farms is very high, and rivers are already too powerful I think.
The grocer is a decent way of implementing this, but I worry it might be too powerful in the early game. 4 food/1hammer cows in the ancient era?
I would have probably created a medieval era thing instead. I think the bonus resources are balanced in the pre-civil-service era.

While a Dynamite bonus is tempting from a realism perspective, I feel it overpowers the tech due to already containing Artillery. This is why I went with a +1 bonus at Engineering

I think engineering is much too early for 4 hammer mines (5 with golden age). I do worry about overloading dynamite, but I also think artillery really need a nerf, they are far too dominating as it stands. Rifles + artillery will annihilate infantry + cannons, it won't even be close.

Tech bonuses are interesting, I'll think about it a while. Overall, I feel a simple buff is somewhat less complicated while still achieving the desired effect,

I worry about making them too dominating in the early game, if the full effect is available then.
I'd be tempted to even give them 2 boosts, one midgame and another late-game.

Did this and some more in Balance - Social Policies.

I saw your changes. I worry your new landed elite is too powerful. There has to be a limit on the total boosts to each specialist all game. With this, you could have all specialists getting +3science+1hammer, while taking half food and half population. That's too much I think.
I think Mandate of Heaven at 75% is overpowered for India.
The others are reasonable.

Paeanblack · Oct 17, 2010

Ahriman said:
I think engineering is much too early for 4 hammer mines (5 with golden age). I do worry about overloading dynamite, but I also think artillery really need a nerf, they are far too dominating as it stands. Rifles + artillery will annihilate infantry + cannons, it won't even be close.

I think the problem with Artillery is that Dynamite has no prereqs in the north half of the tree. You don't even need Pottery. If Dynamite required Military Science or Scientific Theory, it would significantly delay the Artillery beeline.

Thalassicus · Oct 17, 2010

@Paeanblack
A Military Science requirement is an interesting possibility, and would be easy to do, given the relative position of the two on the tech tree.

@Ahriman

All techs in Civ V have links for 'and' requirements, none overlap or cross-link from other locations. It would be possible to do an 'and' requirement and indicate it in the tech tooltip, and might be possible to have links overlap, though these would be a little out of place compared to the rest of the tech tree.

I don't remember the exact amount and there doesn't appear to be a place to alter it in the xml files, but Research Agreements do go up in cost each time you complete one. I think it's +50

each time on Epic speed. Something to remember is these replace tech trading, which did not increase in cost with era... though there's the point advanced civs get a bigger bonus now than they could in IV. Afforess's Tech Diffusion helps with that somewhat.

You're right that Cavalry have 3

, that's interesting.

The first thought that comes to mind is context: at the time of Cavalry, roads are generally all built up and are faster than in the classical era. There's still a lot of terrain without roads when horsemen are around, and the road speed bonus doesn't come until mid-medieval period. The movement difference might be to compensate for this, giving "mobile" units about the same mobility in both eras. Though it doesn't apply to slow units, probably because of the second problem:

A tricky thing about changing movement speeds is its coarse granularity. Even -1

to horsemen is a 25% nerf. In my first pass at attempting unit balance I'm trying to keep changes in the small 10-20% range... though if needed I could go with a larger change.

You're right about the possibility of a separate farm improvement for hills. It would be a serious pain though... lots of xml entries, and linking in with the AI too. Another thing to consider is consistency... there's advantages to finding solutions that retain consistency with similar game concepts (like finding a solution to the tech tree oddities re. riflemen/infantry/mech mentioned above without crossing links).

A big reason for changes to farms and food resources is to make a specialist economy viable. In the vanilla game specialists are very underwhelming for yield purposes, yet I didn't want to solely boost them alone, so I spread the bonuses across several areas of the game (farms, resources, buildings, policies, specialists, GP tile improvements).

You have a point that rivers might be too powerful, though they seem more common in Civ V than in IV (and you almost always start next to a river) so it doesn't seem quite as much of an issue. Rivers also got a nerf since the watermill improvement was removed from IV, and replaced with a weaker building version. Still, I'll think it over and see if there's an alternative way to improve farms.

A big reason for a mine bonus at Engineering is to provide more variety than TP spam. TP spam somewhat trumps production, especially since a TP on a hill gets a double bonus out of golden ages. In addition, another reason is to address concerns over production rates being too slow in the early game. A third reason is to provide a more viable alternative to the civil service slingshot route, on the lower half of the tech tree.

One thing to consider about GP tile improvements in the early game is the exponential effect of early gains for late-game potential. A quick, instant wonder or tech amplifies your empire dramatically, especially in the first one or two hundred turns.

Still, you have a very good idea to buff GP improvements incrementally at various points in the tech tree, I think it's an excellent idea.

Something to consider about Mandate of Heaven is

from happiness does not affect borders, and isn't affected by modifiers. In addition, even if your civ has 30

excess in the Industrial era (shortly after Theaters become available), the vanilla Mandate of Heaven provides with less culture than a single cultural CS ally, a few percent of overall culture in a culture-victory game.

Regarding Landed Elite, a common feeling seems to be the Tradition tree is underwhelming compared to others. Specialists in general could also use some benefits to make them a viable alternative to direct yields, as mentioned above, since in the vanilla game they are significantly worse. In addition, a specialist economy is not very effective until the medieval period or later, so the early-game usefulness of this policy is limited. This is a corollary to the earlier point about the exponential effect of early-game benefits - since the effect is not terribly useful until mid to late game, it's not as powerful as other effects at the same tier in the Liberty and Honor trees. Production, culture and happiness bonuses are useful immediately, and the upgrade cost reduction can provide early gold savings when upgrading to Knights or Longswordsmen.

In addition, a

bonus to specialists in the Tradition tree helps smaller empires keep up with research compared to large ones, due to the reasons discussed in depth here and elsewhere (primarily, this bonus can best be utilized by a non-rex, non-conquering game opener).

Another reason for this is it lessens the relative value of Scientists and Great Scientists, since other specialists can provide research bonuses.

Ahriman · Oct 17, 2010

I think the problem with Artillery is that Dynamite has no prereqs in the north half of the tree.

Agreed. Military science would make a lot of sense, and could even work with the existing tree.
But the other problem is that the marginal value of techs is not really equal. Some techs give massive advantage (Rifling, dynamite, civil service) while others are very mediocre (masonry, metallurgy, engineering, metal casting, compass, economics, refrigeration, penicillin, computers, ecology).

* * *

in the vanilla game specialists are very underwhelming for yield purposes

I think its a mistake to try to boost specialists by buffing yields. This can get out of control very rapidly, when compared to tile yields (and with Maritime city states). A better solution IMO is to buff great people.

though they seem more common in Civ V than in IV (and you almost always start next to a river)

I don't think this is true. It depends a lot on mapscript and map size. On Large Continents, I start with a river maybe half the time (the other half is usually coastal).
The main reason why rivers are more powerful is that the +1 gold is a massively larger relative benefit. It used to be that cottages provided up to 4-5 commerce, so +1 was no big deal. But now trading posts provide +2 gold, so +1 is a very big deal. And another +1 from golden age on every riverside tile.
There is no fix to this other than removing the gold from being riverside (or moving it a building), but that would be a very big shift.

In addition, even if your civ has 30 excess in the Industrial era

I agree that its near useless in the industrial era, but so what? So are a lot of other early game policies.

Regarding Landed Elite, a common feeling seems to be the Tradition tree is underwhelming compared to others

But this is largely because expansion (including through puppets) is so favored, and because its so hard to get a big city.
If expansion is restricted (through puppet nerfs, and maybe by having the unhappiness from #cities increase with the #cities), while city growth is encouraged (though aqueduct building or food cost tweaks) then this will already tend to favor Tradition.

I'm trying to have a set of balance changes that work together, rather than a whole bunch of separate ones that are on or off separately.

Thalassicus · Oct 17, 2010

The reason for different utility of techs is actually to provide big tech / small tech alternation. This is a design decision by Meier since the first Civ, and I think he chose this to provide variety in the tech tree, though I don't remember all the other detailed reasons. I can't find the interview where I saw this, it was years ago (release of IV). Developers of Civ have always followed this practice, though.

I agree completely that buffing specialist yields is a tricky process. This is why I kept each buff small, +1

for engineers (now in the next patch), +1

for merchants (to make them viable compared to trading posts, discussed in detail here and other places) and +1

with a Tradition policy (for reasons discussed here, here, and some other posts.) Other buffs to specialists are in the other realms of gameplay.

It seems river starts are about three out of four games on large and huge continents maps. I've started several hundred games due to testing mods frequently (usually after making a change I test start new game, check the changes took effect, and quit) so this seems to be statistically likely. I'm not as sure if rivers are more frequent, but the game definitely favors a river start, so in that aspect everyone on the map is at least relatively equal.

In addition, here's something else to consider about rivers: bonus resources are now used to balance zones of the map in V. The map generator picks regions of the map, and if yields are unequal it adds resources or otherwise compensates, so river regions are generally equal to non-river regions for overall usefulness (though with different purposes, such as higher food or production in some areas).

Regarding happiness in the industrial era, this is when Theaters become available, and is generally the timeperiod where you can conceivably get 30

at around emperor difficulty or up. Happiness will be lower in earlier eras, but the benefit of cultural city-states is also lower, so the benefit of Mandate of Heaven is still generally weaker than 1 city state unless you have much more happiness. For cultural victory games as Siam, it's proportionally worse, and weaker than a city state even at 60

. I agree it's more powerful for Ghandi, though the issue there is more with his UA in general (it's actually a very powerful UA for expansive empires).

I agree it's best to have balance changes that work together, this is the reason for the comprehensive specialist economy buffs discussed before (terrain, resources, policies, yields, etc) spread across several aspects of the game.

I also agree expansion might be encouraged a bit too much in CiV currently, and this is why I'll pursue the aqueduct and puppet state ideas. (The limiting factor is any major changes to puppet states will require C++ access.) Still, tying in with the previous point of multiple small buffs that work together, a buff to Tradition seems like a good idea. It seems unlikely it could actually become overpowered compared to Honor or Patronage, but if that happens I can always reduce it slightly.

Ahriman · Oct 17, 2010

This is why I kept each buff small,

Thats the problem; in a game where tile yields are 2, and improvement yields are 1-2, increasing specialist bonuses by 1 is not small, it is huge. Increasing merchants from 2 to 3 is not small, it is a 50% increase.

With all your changes, a merchant specialist could give 3 gold, 3 science and 1 hammer at the cost of 0.5 food and 0.5 happy, along with 3 GPPs.

As compared to say a grassland trading post, which could give 2 food 2 gold 1 science, at the cost of 1 food and 1 happy.

These are not balanced.

You should always be getting smaller yields from specialists than from working tiles, because specialists give great people.

Regarding happiness in the industrial era, this is when Theaters become available

Theaters are available in the first tech of the Renaissance era (Printing Press), not the Industral era. Wildly different.

I also agree expansion might be encouraged a bit too much in CiV currently, and this is why I'll pursue the aqueduct and puppet state ideas.

Check the last few posts in this thread before you joined; we're considering tweaking happiness from # cities, so rather than a flat 2 it increases with more cities, either linearly or towards an asymptotic upper limit.

Thalassicus · Oct 17, 2010

The terrain yields are indeed better than specialists, as shown in the post I linked earlier. Remember those posted yields are with hypothetical +2 bonuses as part of the discussion in context, not the +1 bonus I eventually settled on. Tiles also get golden age bonuses in addition to those yields, unlike specialists.

Theaters are not active at the first tech of the Renaissance era unless the player has a large quantity of gold stored upon researching the tech, remember they have to be built or purchased. In general, with normal construction practices a theater can be in every city around the time of the late Renaissance or early Industrial era.

Again though, the point is not specifics (which depends on the context of a particular game) but the overall fact that Mandate of Heaven is about equal to or weaker than one cultural city state, and only a few percent of overall culture even with a very high excess happiness. It doesn't matter what the timeframe is, I simply picked the industrial era as an example.

In addition, there's another point discussed in the policies thread:

Thalassicus said:
It does not convert all happiness from all buildings into culture, only excess happiness. For example, you might have:

30 from resources
5 from natural wonders
25 from buildings
8 from difficulty level

10 from number of cities
50 from population

In a situation like this, if you consider buildings "on top" of all other bonuses, only 32% of the they provide is being converted into culture. In the default game it would be 16%. If your civilization is unhappy, the social policy has no effect at all and is wasted space.

It's an intriguing idea to increase unhappiness depending on number of cities. I like the idea, but didn't comment because it's such a big change affecting so many realms of gameplay I want to think about the implications for a few days. I'm getting sorta sleepy too.

In addition to your idea, you're probably already using Killmeplease's Emigration mod? It also helps balance the game's unhappiness curve, and is something I've given a lot of thought to, tried out and tweaked some, and would probably integrate well with the idea of dynamic unhappiness from #cities. I created some tools to track average un/happiness and weight factors like city puppet status I integrated with it... this actually sorta ties in with your thoughts about penalties to city states.

---

- Edit -

I recalculated the yields from the linked post when considering the final changes, and they come out to:

18 :commerce:

: 6 grass river TPs
12 :commerce:

: 3 Grass river farm + 3 Merchant
14 :commerce:

: 2 Grass river farms + 4 Merchant (Civil Society)

12 :hammers:

6

: 3 Grass river farm + 3 River mine
..6 :hammers:

3

: 3 Grass river farm + 3 Engineer
..8 :hammers:

2

: 2 Grass river farm + 4 Engineer (Civil Society)

12 :commerce:

: 6 grass TP's
..6 :commerce:

: 4 Grass farms + 2 Merchant
12 :commerce:

: 2 Grass farms + 4 Merchant (Civil Society)

8 :hammers:

: 4 Grass farms + 2 mines
4 :hammers:

: 4 Grass farms + 2 Engineers
6 :hammers:

: 3 Grass farms + 3 Engineers (Civil Society)

Unless there's an error in my math (please check) specialists on their own are still worse than tiles even with a buff to their yields and another buff to farms. Other considerations are how Golden Ages tilt the game away from a specialist economy, while the +

policies tilt it towards them, though these factors depend more on the context of an individual game and aren't as easy to quantify.

Celevin · Oct 17, 2010

Thalassicus said:
It's an intriguing idea to increase unhappiness depending on number of cities. I like the idea, but didn't comment because it's such a big change affecting so many realms of gameplay I want to think about the implications for a few days. I'm getting sorta sleepy too.

In addition to your idea, something thing I've given a lot of thought to, tried out and tweaked some is Killmeplease's Emigration mod. You've probably tried it, it's an interesting way to make unhappiness a little more significant, and would probably integrate well with the idea of dynamic unhappiness from #cities.

That's, um, interesting. I don't know how much I'd enjoy having a happy happy empire with plans to use Mandate From Heaven / Golden Ages well, then have a billion people immigrate and start starving to death.

What I don't like is there's no inner-empire immigration. Wouldn't your big cities start receiving people at a rate faster than you actually feeding them? Can you decide which cities gain the people? It sounds like it makes food growth too less of a factor.

I don't know if we discussed it here. We need unhappiness penalties to be more continuous. While it's drastic, I think we need to take out the current ones and put in a per-unhappy-face penalty. I've written on exploiting the crap out of the current penalties, with part of the problem being about how it only switches the penalty at -10 so it's easy to bob around it, and about how the penalty at unhappy and very unhappy aren't steep at all.

Thalassicus · Oct 17, 2010

Remember AI's are generally happy, due to the AI's obsession with happiness. City states in particular generally have excessive happiness. As a result it's very unlikely the player will ever receive emigrants -- it hasn't happened in any of my games.

The Emigration mod is basically designed to solve the "ignore happiness" issue for human players. Ignoring happiness is one of the best strategy for military domination in vanilla CiV, since the penalties are relatively minor compared to the massive tech bonuses you can get from rapid conquest. Emigration also helps curb out-of-control human expansion even when not following the ignore happiness strategy. I agree completely a smoother unhappiness curve would be a good thing to help solve this, as detailed below, something on my todo list for when we get c++ access. Overall, I think a distributed combination of more unhappiness with #cities, emigration, and a smoother curve would provide a nice balance of effects without concentrating it all in one area.

Thalassicus said:
I haven't looked into it but I'd highly suspect smoothing the unhappiness curve will require c++ access. It's a rather core element of the game.

Honestly, it feels weird that happiness smoothly curves up in golden-age value, but unhappiness has jerky hops in effect. It's just annoying to be honest, puppet some city and you might drop from happy to angry, then back up to unhappy, jitters around too much. The transitions aren't consistent in effect either. Should be smooth... like -10% population growth (<=-100%), -3% unit strength (<=-75%), and -3% production (<=-75%) for each . Overall effect would be the same at 10, and the stability of your empire would keep deteriorating from there.

Ahriman · Oct 17, 2010

The terrain yields are indeed better than specialists, as shown in the post I linked earlier.

?
In the post you link to:
6 grassland river trading posts (which takes 6 river tiles!) gives less gold than 2 grassland farms plus 4 merchants, AND you are ignoring the GPPs, and the many other potential boosts to specialists (there are many fewer potential boosts to trading posts).

Trading posts have 1 other booster. Specialists have 4-5.

All your calculations ignore the GPPs, which are valuable. Suppose an average great person cost of 1000 (the first several are much less than this), then a great merchant gives you 500 gold and 30 influence, for net value 750 gold. So a GPP is worth at least ~0.5 gold.

In your design, I worry you are making Freedom/Rationalism with a Specialist economy a total no-brainer that far outclasses others, especially if you snag Statue of Liberty.

Theaters are not active at the first tech of the Renaissance era unless the player has a large quantity of gold stored upon researching the tech, remember they have to be built or purchased.

They become available then. Its just not true to say "Regarding happiness in the industrial era, this is when Theaters become available".
It doesn't take an entire era to build a theater.

In addition to your idea, something I've given a lot of thought to, tried out and tweaked some is Killmeplease's Emigration mod.

It may just be personal taste, but I strongly dislike this kind of mechanic.
It is just not fun for me if I just suddenly start losing citizens. If I want to run a totalitarian police state, my citizens aren't going to abandon me. Its ahistoric (we have very, very seldom had mass exodus of population - if for no other reason that for most of history people are too poor to be able to move, and generally only from relatively free countries - no mass exodus from Great Leap Forward China or Stalinist Russia), and its (for me) unfun.
I also think it messes up food as being the limiting factor on population growth.
It is also another major mechanic that the AI won't understand at all, and it won't seem very transparent even to the human player.

I think that a mechanic like this needs to remain a very rare, optional thing, separate from a main inclusive design of balance tweaks.
Nothing against it as an option, its just not for me.

We need unhappiness penalties to be more continuous. While it's drastic, I think we need to take out the current ones and put in a per-unhappy-face penalty.

This is probably reasonable, but again it would need to be very simple, and potentially with a cap. I worry that if you start making the penalties (eg production) too punitive, you risk creating death spirals where players can't dig themselves out.
One thing that I think might work well is science penalty for unhappiness. This makes sense to me from both gameplay and realism basis.

* * *

Remember AI's are always happy, due to the AI's obsession with happiness.

Not necessarily, I've seen conquest-happy AIs go Very Unhappy.

Thalassicus · Oct 17, 2010

Great person points do indeed make specialists more valuable, remember golden ages also make them less valuable, and there's other considerations. This also ties in with the earlier point about early game vs midgame vs late game bonuses. Overall in playtesting people have found a specialist economy starts becoming viable in the midgame, if you have the modded food bonuses to farms, modded resource yields, other benefits to buildings and wonders, policies that benefit them, etc... basically, the combined package brings specialist economies up to par with a tile economy. Specialists don't ever completely outweigh tiles however, it depends on the terrain, and you have to strategize for them.

By doing this player choice is increased for small empires and there are more interesting decision-making opportunities. The player has the option of changing from a tile to a specialist economy in midgame, but doesn't have to - a tile economy is still powerful. In addition, large empires are still better off with other policy trees and a tile economy.

The point about mandate of heaven is not the era, that timeframe was just picked as an example. Happiness and cultural city state bonuses both rise and fall with era. In general MoH is weaker than 1 city state and provides a few percent of overall culture at best. There's alternative solutions to the problem we can pursue with c++ access, but in the meantime this makes MoH somewhat worthwhile.

The Emigration mod is of course up to personal taste. I like finding ways to solve problems in multiple aspects of the game. Reducing population is a natural way to deal with the ignore-unhappiness problem in a manner that avoids requiring dramatic unhappiness penalties, and works well with a smoother unhappiness curve.

Do you have ideas for how to solve the ignore-happiness problem other than curve-smoothing and population control? It's a long thread to read everything.

Depending on what the solution is, it might be possible to implement with the tools we have available... and anything will probably be possible with the full sdk later.

Ahriman · Oct 17, 2010

Specialists don't ever completely outweigh tiles however, it depends on the city, and you have to strategize for them.

This gets at my point. Specialist economy should never be even equaling yields from
Which would you rather work: grassland trading post not on a river (at best, 2 food, 2 gold, 1 science, up to 3 gold with golden age) or merchant specialist (3 gold, 3 GPPs, up to 3 gold, 3 GPPs, 1 hammer, at half food and half unhappy)?

Look at Civ4. People could use a specialist economy despite it never reaching the same yields as a cottage, because the specialists themselves were valuable.

By doing this player choice is increased for small empires

a) Player choice is not increased if specialists are superior to tiles. Player choice is decreased if one actions is clearly superior to another.
b) Why only small empires? I don't see anything in your changes that stops you running specialists in every city.

Specialist economy should be about great people, not the yields of the specialists.

In general it's weaker than a city state and provides a few percent of overall culture at best. There's alternative solutions to the problem we can pursue with c++ access, but in the meantime this makes MoH somewhat worthwhile.

I guess my problem is that its overpowered with India but underpowered with everyone else. Boosting it only fixes half of that while making the other worse.

Do you have other ideas for how to solve the ignore-happiness problem?

I think you're on the right track with the general continuous penalties, say -abX% production and science and culture and military strength, where X is the number of unhappiness points, b is maybe 3, and a is scaled by map size (1 on standard, less than 1 on large, more than 1 on small).

A core set of balance changes

King

Tyrant

King

Tyrant

King

Attachments

Bytes and Nibblers

Tyrant

King

Tyrant

Prince

Bytes and Nibblers

Tyrant

Bytes and Nibblers

Tyrant

Bytes and Nibblers

King

Bytes and Nibblers

Tyrant

Bytes and Nibblers

Tyrant

Similar threads