New Beta Version - September 25th (9-25)

[Gazebo]

The main reason of sudden unpredictable drop is that happiness is a "per yield per city integer", which mean that if you have 10 cities, you can have +/- 50 happiness in one turn without real reasons. (though it would be quite unlikely to have a drop THAT big)

The solution "stop using integers for happiness" would smooth up everything and really improve the experience, but basically ask to recode everything linked to happiness and track for every use of it everywhere (which would need to a lot of works, and a lot of bugs, so even more work). Not to say it could break any mod relying on the Community Patch or Vox Populi.

The solution "average happiness trough turns" is an alternative way to force a smooth up, but would lead to the unintuitive "your global happiness is not the sum of happiness of all your cities, since it is averaged on multiple turn". On the plus side, it would make unhappiness far more predictable.

I’ve tinkered with top value smoothing but it causes exactly the problem you mentioned.

To head off complaints: There is no bug in the code. The smoothing of global averages helps but ultimately as you noted no amount of smoothing changes the fact that there is a hard conversion to single digit integers in civ’s deep happiness code.

Two possibilities I haven’t explored:

Use top level smoothing _plus_ an ‘overflow’ pool of ‘becoming unhappy’ citizens. So happiness can only change by 1 per turn, but the overflow pool of ‘pressure’ holds the other changing values. Make sense?

I could also have the unhappiness penalties only start to trigger once you are out of golden age points (much like running out of gold affects science). We’d probably need to reduce golden age points or something to compensate.

G

 

[CppMaster]

I like second option more. Golden age points will be more valuable.

 

[Almonds]

I could also have the unhappiness penalties only start to trigger once you are out of golden age points (much like running out of gold affects science). We’d probably need to reduce golden age points or something to compensate.

Off the cusp here but that just sounds like it'd be an absolute mess to balance. I'd go so far as to say it fundamentally changes that yield and how happiness would function. You've already put a lot of hard work into adding sources of GA points all over the game, balanced for a completely different purpose. I wouldn't mess with that.


I've also never experienced the extreme dips that others seem to be plagued by. I think the key is to be thinking of ways to reduce unhappiness while your people are still happy. At least half the time a city needs a new building I check the Needs, regardless of my empire's happiness level. I do dip into negatives but there's always a clear reason for it, like conquering too quickly. Sometimes if I feel like I'm in a rut of unhappiness, I reset all of my specialists. Getting people off those slots and then re-adding only where appropriate for my current predicament helps pull me out. I think some people think their solution is to pile on Scientists or Merchants etc to alleviate Poverty or Illiteracy, but I have found that that just creates a bubble of inflated yields that drives up those Needs artificially.

 

[Moi Magnus]

Alternative suggestion:
Scale the different floors with your number of cities.

Ex:
- You need 2 happiness per city to have +10% to yields
- You need -4 happiness per city to have -20% to yields
- (Use a linear progression in between)
- You need -4 happiness per city during 10 turns to start having revolutions.

It means that if you have 10 cities, a drop of 40 happiness has the same effect as a drop of 20 happiness with 5 cities.

(Another idea is "global happiness is the average of the happiness of all your cities", instead of the sum. I find it more elegant, but it means all the global bonuses to happiness have to be rebalanced, which isn't great)

 

[TwoPodRay]

I feel like I'm the only person who never experiences these huge happiness swings in my games. I don't even particular try to combat unhappiness, either. I purposely DON'T buy every lux that is available because I want to make it easier to obtain whatever lux is needed for WLTKDs, so my empires could easily be happier. The only time I ever seem to dip into unhappiness is right at the beginning of the game when I first expand to 4 or 5 cities and often get down to like -5. After that I am always comfortably positive and sometimes overwhelmingly positive (like 50+ or even 100+). I typically don't have to consider happiness when choosing what buildings to construct. I would probably have to find myself in some sort of strange situation for happiness to become a big problem- like getting Dow'd by the whole world or something.

This is the case regardless of the type of game I'm playing. Puppet/authority as a conqueror, wide/progress, tall/tradition- almost always swimming in happiness. Maybe it's inherent to the way I play? Maybe I prioritize production and have weak growth so my population never balloons and this just naturally leads to excess happiness?

I am the same way. I’ve learned to dynamically adjust my play to combat any potential happiness problems before they ever crop up. Outside of the first 100 turns of initial expansion, I don’t even look at the happiness number except to see how many more turns until my next golden age.

I could also have the unhappiness penalties only start to trigger once you are out of golden age points (much like running out of gold affects science). We’d probably need to reduce golden age points or something to compensate.

G

I also strongly support this. I like this approach.

I also wonder if a similar approach could be used to how War Weariness has a two pronged effect on game dynamics. I’ll make a post in General Balance on the topic later this evening when I can check a few numbers.

 

[CrabHelmet]

I don't think any changes are needed. You just have to be pre-emptive.

 

[tu_79]

Maybe if I'd understand better how unhappiness drops so strongly for some people...

So, they have 20+ cities, then they research one or two techs and suddenly there's one more unhappy citizen in each city? I'd think that only a few cities should apply for the increased unhappiness, not all of them. I'd think that some of those cities could just stop working on a specialist or two and come back to normal.
Also, why is it in industrial age? Is it because by the time it is possible to have very large empires and not earlier? Or is it because of fertilizer?

I guess logs are needed for being sure about those things.

 

[TwoPodRay]

I don't think any changes are needed. You just have to be pre-emptive.

I think for the players that are currently preemptive, it won’t make a difference. For those who are not, they’ll either have to learn to be more preemptive or will suffer GA point loss.

And if there’s some hard to detect bug or error that isn’t getting reported or discovered, until it is found, this is a decent bandaid that could be reverted later if needed.

Thinking about it more, perhaps the revolting cities could stay, but the 20% reduction on yields could be moved to only happen after you’ve run out of GA points. One of my war strategies if not taking cities or vassals is to spike their war weariness and sit in their lands, keeping them from fixing improved tiles, using Great people, trade, or diplomatic units, and from producing a military. I just keep refusing peace. This causes huge unhappiness values for them and causes city revolts and overall keeps them from being a threat for a long time if not the rest of the game. (But more on war weariness in my future post)

 

[Almonds]

I think for the players that are currently preemptive, it won’t make a difference. For those who are not, they’ll either have to learn to be more preemptive or will suffer GA point loss.

And if there’s some hard to detect bug or error that isn’t getting reported or discovered, until it is found, this is a decent bandaid that could be reverted later if needed.

GAP as a yield is really much too integrated into many systems for an overhaul of its functionality to be treated as a band-aid. All for a mythical bug that Gazebo asserts does not exist. It's not a good use of anyone's time.

I'd rather we discuss how the happiness mechanics works as-is, rather than calling for a change. To me, it is apparent that there are a lot of things going on under the hood of an empire that contribute to let's call it "invisible instability". To be one tech or city away from a deep drop in unhappiness without any clear indicator of what is to come is understandably frustrating. It's also not very clear that reactionary procedures to alleviate unhappiness take longer, because of these underlying unhappiness sources. Maybe there should be updates to the tooltip to stress the more subtle sources of unhappiness, like tech level or specialist use. Even better, maybe some extra info could be added to the city-level Needs tooltip. Something like, your next Technology will cause x to increase. Is that possible?

 

[TwoPodRay]

GAP as a yield is really much too integrated into many systems for an overhaul of its functionality to be treated as a band-aid. All for a mythical bug that Gazebo asserts does not exist. It's not a good use of anyone's time.

I'd rather we discuss how the happiness mechanics works as-is, rather than calling for a change. To me, it is apparent that there are a lot of things going on under the hood of an empire that contribute to let's call it "invisible instability". To be one tech or city away from a deep drop in unhappiness without any clear indicator of what is to come is understandably frustrating. It's also not very clear that reactionary procedures to alleviate unhappiness take longer, because of these underlying unhappiness sources. Maybe there should be updates to the tooltip to stress the more subtle sources of unhappiness, like tech level or specialist use. Even better, maybe some extra info could be added to the city-level Needs tooltip. Something like, your next Technology will cause x to increase. Is that possible?

While I do actually think it could easily make logical/thematic sense for the adjustment he mentioned (EDIT: to clarify, GAP is already reduced by the amount of unhappiness per turn, the only change would be to delaying effects of being unhappy until you’re out of GAP), I’m also curious to hear more about ideas for improving the future happiness predictions as you alluded to, and if those are improved sufficiently, then no change is likely needed anyways.

 

[theMisterRud]

I’ve tinkered with top value smoothing but it causes exactly the problem you mentioned.

I could also have the unhappiness penalties only start to trigger once you are out of golden age points (much like running out of gold affects science). We’d probably need to reduce golden age points or something to compensate.

G

,
I think this solution still doesn't solve the fact that those drops FEEL jarring and painful to experience. Don't get me wrong, it is not a bad solution, is just that I'm afraid that the cure might be aggraveting to implement, while the result is still sub-par. The first solution has a better feel to it.

Out of curiosity, what kind of smoothing have you tinkered with ?

 

[Stalker0]

The fundamental problem is actually a UI one, a newer player can play the game in a way that the UI is telling him he’s doing well...to then suddenly fall off a cliff.

While I don’t like the happiness dips to be thst extreme, I do think they can be managed by a seasoned player. The problem is newer players. They will expand and ignore infrastructure, and seeing +20 happy they think “I’m playing good!”.

Then suddenly happiness plummets, and the player goes “but wait...what did I do wrong?” Thst generates the sort of helplessness thst causes rage quits.

Now if changes to the smoothing curve can help so be it, be even something Iike a color change of the happiness number thst is responsive to your number of cities.

At 6 cities, 20 happy should be green (I’m doing good!). At 20 ciites 20 happy should be orange (careful now, you are ok st the moment but you should make this number higher).

 

[Gazebo]

,
I think this solution still doesn't solve the fact that those drops FEEL jarring and painful to experience. Don't get me wrong, it is not a bad solution, is just that I'm afraid that the cure might be aggraveting to implement, while the result is still sub-par. The first solution has a better feel to it.

Out of curiosity, what kind of smoothing have you tinkered with ?

No one likes to lose, I can't really help feelings.

The fundamental problem is actually a UI one, a newer player can play the game in a way that the UI is telling him he’s doing well...to then suddenly fall off a cliff.

While I don’t like the happiness dips to be thst extreme, I do think they can be managed by a seasoned player. The problem is newer players. They will expand and ignore infrastructure, and seeing +20 happy they think “I’m playing good!”.

Then suddenly happiness plummets, and the player goes “but wait...what did I do wrong?” Thst generates the sort of helplessness thst causes rage quits.

Now if changes to the smoothing curve can help so be it, be even something Iike a color change of the happiness number thst is responsive to your number of cities.

At 6 cities, 20 happy should be green (I’m doing good!). At 20 ciites 20 happy should be orange (careful now, you are ok st the moment but you should make this number higher).

I feel like the GAP buffer is the silver bullet to this: having the penalties only take effect once you are out of GAP to lose would be comparable to a sudden jump to negative GPT because of losing a few TRs. The latter is manageable because you can build a buffer against it and it gives you time to correct the issue before you are truly penalized. The question is whether or not the GAP pool and/or sources of GAP are too large, which is an issue we can discuss elsewhere.

It also means that empires with more golden ages (and a large GAP pool) will be more prone to stability overall, whereas civs coming out of a golden age are at slightly higher risk, as their pool will be emptied.

G

 

[pineappledan]

I feel like the GAP buffer is the silver bullet to this: having the penalties only take effect once you are out of GAP to lose would be comparable to a sudden jump to negative GPT because of losing a few TRs. The latter is manageable because you can build a buffer against it and it gives you time to correct the issue before you are truly penalized. The question is whether or not the GAP pool and/or sources of GAP are too large, which is an issue we can discuss elsewhere.

Persia is gonna be laughing all the way to the bank this next patch.

This also might shut up a few of the people who were complaining about the Taj Mahal

 

[Gazebo]

Persia is gonna be laughing all the way to the bank this next patch.

This also might shut up a few of the people who were complaining about the Taj Mahal

I mean, if your empire can generate GAP fast enough to counteract unhappiness, that seems like it opens up a small but interesting new twist in the happiness mechanic, no? Why not? GAP is a 'filler' yield right now anyways.

G

 

[pineappledan]

Why not? Though I think this means that the GAP on monopoly resources needs to get re-assessed (ie Ivory). Those resources are going to be a lot more valuable now, because empires can use a monopoly of 5 ivory to suddenly set their lower happiness buffer at -10 (the generate 2GAP per elephant tile)

You might need to reduce the GAP on monopoly bonus from +2 to +1

 

[TwoPodRay]

Also, would you be immune to the unhappiness penalties while currently in a golden age? Such that if you just started a GA and your GAP is 0, right when unhappiness tanks, you at least are safe from penalties until GA ends? Could also make wonders/policies that start GAs to be more helpful than before for unhappy empires, because it gives them breathing room to fix problems while in the GA without penalties.

 

[Almonds]

I mean, if your empire can generate GAP fast enough to counteract unhappiness, that seems like it opens up a small but interesting new twist in the happiness mechanic, no? Why not? GAP is a 'filler' yield right now anyways.

G

It it's interesting and new, sure, but also seems super easy to abuse. There are many ways to get and buff Golden Ages without accumulating any GAPs at all. Wonders, Policies, Great Artists all grant instant ones and if properly buffed can last for a super long time. You can even get permanent Golden Ages in the later eras, fully negating any unhappiness in the late-game where unhappiness might pile up from Public Opinion or heavy War Weariness. Aside from that, getting GAPs from tile yields, policies, etc effectively become free happiness points that artificially lower the happiness threshold. I worry that this change would largely negate unhappiness penalties that should actually be pretty severe if not attended to by properly managing Needs.

Aside from all that, this new mechanic would now have to be balanced against the existing production, culture, gold, and other benefits (Mosques, Baths, I know there are numerous ones) of getting GAs, which would effectively mean having nerfing these GA-centric strategies to make up for their new benefit.

 

[theMisterRud]

The fundamental problem is actually a UI one, a newer player can play the game in a way that the UI is telling him he’s doing well...to then suddenly fall off a cliff.

This is the point that I wanted to make: as a player that has not experienced these swings in a while now, I feel this is more of a presentation problem that a sistemic one. Even a notification that warns the player if a projected happines value is too far below the current one would be beneficial in terms of feedback.

That said, the GAP solution would be great as it would fix the only instance in which I got these swings: tall artistry.

 

[tu_79]

Sorry, I missed what the proposal exactly is. During Golden Ages there is not going to be a raised happiness? How so?

Changing distress to just hunger could be useful for giving a tool to the player to combat unhappiness. Put the city to work on food, and select no growth. Voilà, happy city. Not going to be useful until some more buildings are built, but useful meanwhile.