Flexible Happiness

[Ahriman]

Civ4 happiness works very differently than in Civ5, so I'm not sure Civ4 comparisons are very helpful. Civ4 happiness is city-based not empire based, Civ4 has war weariness (significantly rising happiness during wartime), and in Civ4 excess happiness has zero marginal benefit.
And of course, in Civ4, happiness is not the main limit on expansion (maintenance costs are), so relaxing happiness constraints does not allow for unchecked expansion.
So I don't think that Planetfall or Alpha Centauri are good guides.

I would oppose reducing the base happiness yield of buildings; I already think that happiness from buildings is too low relative to happiness from other sources.
I would also oppose forcing players into using specialists for happiness (which is what would happen if you reduced building happiness further).
You'd also start getting a position where, with the freedom policy that reduces specialist unhappiness, specialist citizens could provide positive net unhappiness, which I think is undesirable.

 

[Sneaks]

The issue with happiness in my mind is the odd hybridization of local and global happiness.

Unhappiness penalties are always global, and none are city based. I think this needs to change. If local unhappiness is 10 even if we are globally happy, I have profound issues with this. Perhaps once a city dips into the face, it should drop into resistance. Just being mildly unhappy should perhaps cause random tiles around the city to become pillaged.

In return, I think global happiness needs to be recalculated. Instead of being the aggregate sum of all the happiness and unhappiness, it should take into account the % of cities in each state.

 

[Ahriman]

The issue with happiness in my mind is the odd hybridization of local and global happiness.

I find this a bit odd. There is no hybridization, there is only global happiness. What you are suggesting is a hybrid system, where local and global happiness both exist and matter. This seems very complex and confusing for no real gameplay gain. I think it is good that happiness is global and simple; that it doesn't matter where the luxury good comes from, or which city I build that happiness wonder in.

Calculating and managing local happiness in each city would become very confusing given all the various modifiers (-5% from Liberty, etc.) and throwing low happiness cities (eg: those newly conquered, which have lost many of their happiness or happiness-with-social-policy buildings) into unrest would be not fun either.

I think people need to be more clear about what problem they think they are trying to solve, or what their design goal is.

 

[Dunkah]

You could use a separate mechanic to define the Puppets Loyalty/ State of mind/ Influence...

The higher your influence with them the more happines/ Gold/ Hammers they provide.

Influence could be adjusted by number of units stationed within borders, a new Entertainer unit that you can spend (like CSD), or cash that is dumped in to the city. Larger cities would need larger amounts of Influence.

 

[tlaurila]

I would oppose reducing the base happiness yield of buildings; I already think that happiness from buildings is too low relative to happiness from other sources.

Buildings are the only souce of happiness that scale directly with number of cities, ie. the number of unhappiness. Relatively more happiness from buildings would increase the slope of the Empire Size Manageable vs Happiness Technology Level - possibly quite sharply so. This would require advanced science before conquering, probably decreasing the Conquest victory's appeal. And make early-game conquests all the more harder to leverage to victory. Poor Romans.

I suppose it's a matter of taste, but personally I much like that one could fight conquest wars in the early eras as well. One thing a more flexible happiness would allow, as in early game your other happiness options are severely more limited.

Why not just let the unemployed citizen be free of unhappiness, but remove its yield entirely?

This sounds like a simple and brilliant solution, far as I can tell. (Or perhaps reduce the unhappiness by factor X) It also goes back to civ1&2, where "unemployed people" were automatically Entertainers. It makes sense too, people only gripe if they're force to work for "The Man", producing resources to their Civ.

With the unemployed unhappiness reduction factor set right, it's hardly too cheap a solution - when your cities start starving and produce near nothing to keep them from revolting (just ask North Koreans how'd this work)... AND it would let you starve your conquests, and loose any resources they would bring, in exchange for them not bringing (that much) unhappiness.

Happiness might be a well-working system for empire size control, but as it stands, you need to plan it just right, or you end up in pretty much unrecoverable s***t. One mistake, like say trusting a trading partner that ended up backstabbing you, you got 15 , and You Lose. Perhaps I'm just not on top of my game for not really being able to recover from that, but isn't my idea of a fun game mechanic. It should cost to have happiness go awry, but a Restore/Restart/Quit -button just isn't fun. And as said, the AI surely suffer from this even much worse. For example, I recall steamrolling a chinese army that was stuck with the unhappiness combat penalty and just couldn't work its way out until I had killed most of their army, and taken several cities.

 

[Txurce]

This sounds like a simple and brilliant solution, far as I can tell. (Or perhaps reduce the unhappiness by factor X) It also goes back to civ1&2, where "unemployed people" were automatically Entertainers. It makes sense too, people only gripe if they're force to work for "The Man", producing resources to their Civ.

Happiness might be a well-working system for empire size control, but as it stands, you need to plan it just right, or you end up in pretty much unrecoverable s***t. One mistake, like say trusting a trading partner that ended up backstabbing you, you got 15 , and You Lose. Perhaps I'm just not on top of my game for not really being able to recover from that, but isn't my idea of a fun game mechanic. It should cost to have happiness go awry, but a Restore/Restart/Quit -button just isn't fun. And as said, the AI surely suffer from this even much worse. For example, I recall steamrolling a chinese army that was stuck with the unhappiness combat penalty and just couldn't work its way out until I had killed most of their army, and taken several cities.

To say that unemployed people are not as unhappy as those working for "The Man" is unsupportable.

If you go from being happy to -15 because of one trading partner's actions, you had too many eggs in one basket. I wouldn't redesign the game to fix that.

I read just about every post in the VEM threads, and unhappiness has not been much of an issue. What people do complain about is the unhappiness and stasis of the long resistance period. It might be a lot simpler and easier to have another, more specific poll asking people whether they prefer resistance at the present length, or half like in vanilla.

 

[Ahriman]

Buildings are the only souce of happiness that scale directly with number of cities, ie. the number of unhappiness.

Actually they aren't, many social policies do this too.

Relatively more happiness from buildings would increase the slope of the Empire Size Manageable vs Happiness Technology Level

Yes, this is by design. Currently, many of the happiness techs are very low value, because they do not significantly boost your economy. You would never beeline for techs enabling theaters or stadiums.

This would require advanced science before conquering, probably decreasing the Conquest victory's appeal. And make early-game conquests all the more harder to leverage to victory. Poor Romans.

That makes no sense to me. I'm not talking about some compensating thing where we increase unhappiness too, I'm just talking about returning to the old version of this mod (and, I think but am not sure, vanilla?) where Colosseums give 3 happiness, and similar for Theater and Stadium. So how is Rome worse off in an absolute sense? Rome is better off, because now they can conquer a bunch of cities and then still handle them by building/buying Colosseums in each. If anything, more happiness from infrastructure makes conquest easier, because you can gold-purchase infrastructure to boost happiness, and happiness is a check on conquest.

This sounds like a simple and brilliant solution, far as I can tell.

Just the opposite; it would mean that there is no actual limit on empire expansion except food, because you can just make every new citizen unemployed.
It gets worse, because even setting the base unemployed citizen yield to zero, it will still get bonuses from all the civics and such that boost specialist yields (I think? AFAIK unemployed citizens are specialists - and so half food and half unhappiness Freedom policies would apply).

but as it stands, you need to plan it just right, or you end up in pretty much unrecoverable s***t

I have never found myself in an unrecoverable situation, I think it is not nearly as harsh as you say. If nothing else, you can give away/sell cities.
And yes, it requires planning... but isn't that a good thing? Incentivizing forward thinking is the kind of strategy we should be trying to encourage.

And as said, the AI surely suffer from this even much worse

I find the AI never gets caught in unhappiness spirals (they get a lot of free happiness), and I think the AI would be unable to make use of unemployed people giving happiness without access to AI code.

* * *

What people do complain about is the unhappiness and stasis of the long resistance period. It might be a lot simpler and easier to have another, more specific poll asking people whether they prefer resistance at the present length, or half like in vanilla.

I also really like the resistance length, it is a great check on lightning conquest.

 

[Txurce]

I also really like the resistance length, it is a great check on lightning conquest.

I agree with this (and everything else you wrote, for once). I'm trying to put the genie back in the bottle via focus on one underlying factor.

 

[Ahriman]

I agree with this (and everything else you wrote).

Hmmm.... I might have to quote this in every post I make here now

I agree that it is sensible to try to pin down what, if at all, people think the problem is. My problem with flexible happiness is basically that it acts to reduce the need for planning and to make the game more forgiving and more tolerant of "anything goes". But I don't think these are good things. So this seems like a solution to a problem that doesn't (IMO) exist. IMO, inflexibility can be a good thing. If you want to conquer, then invest in happiness techs and infrastructure.

 

[Txurce]

Hmmm.... I might have to quote this in every post I make here now

Not if you quote the revised statement accurately!

I agree that it is sensible to try to pin down what, if at all, people think the problem is. My problem with flexible happiness is basically that it acts to reduce the need for planning and to make the game more forgiving and more tolerant of "anything goes". But I don't think these are good things. So this seems like a solution to a problem that doesn't (IMO) exist. IMO, inflexibility can be a good thing. If you want to conquer, then invest in happiness techs and infrastructure.

Agreed. There are too many players on Emperor who don't have this problem for it to be a problem that requires game change.

 

[Sneaks]

I find this a bit odd. There is no hybridization, there is only global happiness. What you are suggesting is a hybrid system, where local and global happiness both exist and matter. This seems very complex and confusing for no real gameplay gain. I think it is good that happiness is global and simple; that it doesn't matter where the luxury good comes from, or which city I build that happiness wonder in.

Calculating and managing local happiness in each city would become very confusing given all the various modifiers (-5% from Liberty, etc.) and throwing low happiness cities (eg: those newly conquered, which have lost many of their happiness or happiness-with-social-policy buildings) into unrest would be not fun either.

I think people need to be more clear about what problem they think they are trying to solve, or what their design goal is.

On this, you're actually wrong. Happiness exists locally and universally as it exists in game right now, and unhappiness as well has several subcategories. Most buildings are examples of local happiness. A Colosseum may only provide happiness within the confines of unhappiness per population within a city. It cannot overflow to other cities. It also cannot combat the unhappiness that exists for that city existing (the -2 per city).

As that paragraph demonstrates, happiness is an extremely complex system as it exists now. I think your confusion comes from the fact that nowhere in the game does it display in tooltips or otherwise your city by city happiness levels. These DO exist, however, and do affect global happiness.

I frankly find the current system unnecessarily complicated, unrealistic, and unenjoyable. Why should building a colosseum on the west coast of a continent affect the happiness and growth of a city on the east coast? The funny thing is it won't if that initial city is small in population.

To make clear some things:

Global Happiness sources:
Luxuries
Wonders
Policies
National Wonders

Global Unhappiness sources:
Per city unhappiness
Resistance/Occupation

Local Happiness sources:
Colosseum, Theater, Stadium

Local Unhappiness sources:
Population

 

[Maniac]

The global happiness system allows a tension between vertical and horizontal expansion (wide and tall empire in Civ5). It's the first game in the civ series which succeeds in doing this. All Civ4's maintenance system did was slow down ICS, rather than make it unprofitable.

 

[Ahriman]

Happiness exists locally and universally as it exists in game right now

I'm aware that there is very mild hybridization in that there are some kind of caps on how much net happiness you can get from buildings in a city (though not from wonders), but I find in practice that never binds. Whereas your proposal seems like a significantly hybrid system (to what end??).

So the current system is pretty simple, in practice. It mattered more in the oldest versions of the game, where you might really have a size 1-2 city with a +4 happiness Colosseum, and thus you could ICS. But that has been gone even in vanilla for a while.
So in practice the slight hybridization that was introduced is now obsolete, and the problem it was designed to fix is gone.

The funny thing is it won't if that initial city is small in population.

Colosseums now provide only 2 happiness. So this will hardly ever come up.

Why should building a colosseum on the west coast of a continent affect the happiness and growth of a city on the east coast?

Basically you're asking, why global happiness.
The answer is, because local happiness (a la Civ4) didn't work well. You got zero reward for excess happiness, and you had horrible situations where you might have one city with +8 happiness that was useless but suffer from another city that was -2 happiness. So you'd have to micromanage like crazy all over the place to get everything right, and many happiness sources were very often useless.
It is much easier to just manage a single number, and to have any extra happiness from anywhere have the same value.

The global happiness system works better.
You could also remove all the residue of the local happiness changes to vanilla and still be ok, and revert to the initial pre-patch vanilla game where all happiness was global.

I agree that you are technically correct, that there is a slight residual hybridization and that this is is slightly odd, but it would seem the solution to that would be to remove it (and go purely global) rather than to go into an even more hybrid system.

* * *

The global happiness system allows a tension between vertical and horizontal expansion (wide and tall empire in Civ5). It's the first game in the civ series which succeeds in doing this. All Civ4's maintenance system did was slow down ICS, rather than make it unprofitable.

Yes, this too.
It was a clever design. They didn't have it right initially, but it is much better now (and in this mod). 3 unhappiness per city was a big improvement, as were some of the other changes (like Forbidden Palace nerf).

 

[Baleur]

I like the idea, but i think there should be more cost than simply a lost worker turned to a specialist.
Perhaps, if it is possible, have the "happiness specialist" produce -1 gold, or something negative that acts as a cost.

 

[tlaurila]

[happiness specialist/unemployed reducing unhappiness] would mean that there is no actual limit on empire expansion except food, because you can just make every new citizen unemployed.

That's what happens if the happiness bonus is taken too far. Like everything, it's not whether or not, but how much. If for every population point your empire is "too big", you'd have to put 3 (or pick your number) population to work for happiness, the cost of growing too big will just as surely grind your empire to dust. And you'd have to start with somewhat less baseline happiness if the flexibility is there.

It'll just be a lot less of "100 happiness and everything is perfect, but if you only manage 99 all hell breaks loose". You could argue the hard need to stay above 100 is strategy and planning, and it'll even start to feel natural part of the game once you've gamed long enough. But it's gamey and counterintuitive to must have theaters to keep on conquering, and cheap game design compared to a more gradual cost response where you can weigh costs and benefits.

 

[Thalassicus]

Changing unemployed citizens into happiness specialists would mean there's no cap on how high we can go. I feel it's best to restrain the possibilities to a narrower range.

@Baleur
I'd be okay with something like -1 gold for specialists such as these.

I don't really like the idea of flexible happiness as the A.I. will not be able to utilize it well.

The AI would use it well. Specialists have flavor values (like flavor_happiness) which the AI is coded to prioritize based on circumstances.

I agree that it is sensible to try to pin down what, if at all, people think the problem is.

The two problems are bulleted in the original post. The second one is the AI cannot plan ahead as effectively as humans. Consider this sequence of events:

1. AI has 1.
2. Neighbor declares war, ending 3 luxury trade deals.
3. The AI drops to 11.

Unless the AI has unbuilt happiness buildings and enough gold to purchase them, this cripples both production and combat strength in an emergency situation. It's even harder to pull out of if we pillage their luxuries. I don't want to reduce the penalties of very-unhappy state, because we could get back to the situation where conquerors simply ignore unhappiness (like a year ago). Just increasing happiness across the board would probably make the game too easy - I want there to be serious cost-benefit tradeoffs to consider when pushing happiness higher.

Flexible happiness provides a way for the AI to deal with these situations, such as activating a few more happiness specialists to bring it out of very-unhappy status, giving it more time to adjust to the unexpected event.

 

[Txurce]

The AI would use it well. Specialists have flavor values (like flavor_happiness) which the AI is coded to prioritize based on circumstances.

The two problems are bulleted in the original post. The main thing this deals with is the AI cannot plan ahead as effectively as humans. Consider this sequence of events:

1. AI has 1.
2. Neighbor declares war, ending 3 luxury trade deals.
3. The AI drops to 11.

Flexible happiness provides a way for the AI to deal with these situations, such as activating a few more happiness specialists to bring it out of very-unhappy status, giving it more time to adjust to the unexpected event.

I don't think people are voting to help out the AI, but I may just be too cynical.

That said, if you really think this is an AI problem that needs addressing, as opposed to the breaks of the game in a fairly uncommon situation (how many AI run negatively?), then I'm not going to keep arguing against it.

 

[Ahriman]

I'd be okay with something like -1 gold for specialists such as these.

What happens though when you start stacking up +gold and science and -food and unhappiness social policies though?

The main thing this deals with is the AI cannot plan ahead as effectively as humans.

I never find that the AI runs into happiness problems, so its inability to plan ahead doesn't really matter much.
I have never seen a situation such as the hypothetical one you construct. I have never managed to trade for 3 luxuries from a single AI.
The AI normally is floating +20 happy or more. It is also usually floating very large amounts of gold, so it could easily buy itself more happiness if needed. Worst case scenario even in this hypothetical case, it will lose a city to an invader, and will then probably be on positive happiness again.

So I think you are fixing a problem that doesn't exist.

I don't want to reduce the penalties of very-unhappy state, because we could get back to the situation where conquerors simply ignore unhappiness (like a year ago)

Allowing conquerors to easily adjust happiness will also have a very negative effect.

The costs to this policy (messing up how happiness to limits growth, reducing the value of happiness techs, reducing the need for the human to plan, dramatically boosting the value of specialist-boosting policies) seem far larger than the benefits (improving AI performance in a situation that rarely arises).

 

[Thalassicus]

The two problems I listed are:

• Happiness encounters wilder swings up and down than other yields (like science, food, or gold).
• AIs have difficulty dealing with situations that drop them below 10.

I'll admit the first point is subjective. How much do we like our power level to wildly swing between one turn and the next? I feel the current range is a bit too high, but it's a personal opinion.

The second has empirical evidence - militaristic AIs often drop below 0 happiness (at least according to InfoAddict) in my games.

 

[Ahriman]

Happiness encounters wilder swings up and down than other yields (like science, food, or gold).

I don't find that happiness has wild swings, except during city conquest. Gold also has high variance, when you buy a lot of stuff or get a lot of stuff in trade or peace deal, but I don't see this as a problem.

So yes, happiness can be more variant... but so what?

The second has empirical evidence - militaristic AIs often drop below 0 happiness (at least according to InfoAddict) in my games.

Dropping below zero (but above -10) is basically harmless and is self-correcting, by design (lower pop growth when negative happiness).
I very very rarely observe an AI going into Very Unhappy territory.

The fact that AIs go below zero is not a sufficient reason to create a new mechanic to prevent them from ever doing so.
Happiness would be lame as a mechanic if you could easily avoid ever becoming unhappy.
If you could avoid unhappiness just by taking slight economy hits, then why not just rewrite the effects of negative happiness and allow it to be reached? [I don't think you should do this, I'm just saying that creating a system that by construction doesn't have negative happiness is a bit pointless.]