So after playing numerous Progress Peaceful Wide type games, I've noticed that happiness tends to fall into pretty predictable patterns for me. Here are some notes (for Immortal play):
1) Expansion Restriction (3 cities): Unhappiness starts to kick in, affects satellites and can even get to the capital in some cases. 3 luxs are needed for ideal (50%) happiness. Communitas_79 provides 2 nearby luxs as a standard, +1 lux from AI trading can usually cover me. Aka its important to get those luxs up and running for continued expansion.
2) Expansion Push (4-7,8,9 cities): I feel like this is the phase where happiness is truly "doing its job". Generally I have to cap growth at 4 pop (sometimes 3 depends on circumstances) as I am expanding, and an ignorance of infrastructure will tilt my happiness into the 34% "pit". So I have to balance expansion with proper management. There is a lot of variance during this time period, as a nice CS quest to give you an ally, or a good yield pantheon can make tremendous difference (like the difference between happiness comfortable and being in full rebellion, it can be that swingy). For example, I have noticed that my happiness tends to be significantly higher with Renewal and Spirit of the Desert games, because both of them give nice base yields that don't require workers to activate.
3) The medieval build phase: So a consequence of an early focus on expansion is that you now have a LOT to do in Medieval. Your capital probably has wonders it wants to build, you still have lots of basic infrastructure to do, and now you need to get your military really ready, as Knights come online this is the time that many AIs like to be aggressive, and its very easy to just flat out die on Immortal if you are not prepared.
Its at this point I find the happiness penalties to military units can make or break you. Local happiness can be a problem during this stage, effectively saddling satellite cities with a -35 to -50% penalty to military production (aka don't bother) I often find it important to maintain growth caps on one key city so it can build military forces without local unhappiness. Also rush buying military units in the capital has become more common for me, you have to get your troop count up quickly....and you simply can't do that with hammers when your facing a -25% or higher penalty to production.
Part of the issue here as well as that you can be a bit lux reliant at this stage in the game, as your still behind on infrastructure. This means a war declaration performs double duty, as it can often strip away your luxs, which can literally cut your military production in half. Again the stress here is on speed, you must must must get your military in order as quickly as possible before those penalties really start to kick in.
4) The "Ping Pong" phase (Renaissance - Public Works): So having survived Medieval, I now generally have all of my core infrastructure built, and Global Happiness starts to solidify more. This begins the most annoying part of happiness for me, the ping pong phase. Almost every city I have other than the capital does the following:
- Gains a Pop, suffers like 4-5 extra unhappiness.
- Growth slows to a crawl, I focus on infrastructure to correct the unhappiness.
- Growth returns to reasonable, city gains a pop, suffers like 4-5 unhappiness, etc
One day I really need to record how many turns a satellite actually spends in unhappiness vs happiness, I bet good money the city is unhappy a good majority of the time.
5) The Public Works phase (Public Works - Landmarks)
Once you get public works, you get a new lever against unhappiness. So the -15% to unhappiness I find pretty worthless, it lowers your unhappiness by a good amount....until you grow and the ping pong kicks in again. But the +1 happiness remains rock solid and ever useful. Its what I'm "actually buying" with the public works.
6) The "No Big Deal" phase (Landmarks - Game End)
So depending on the amount of Landmarks and Public Works you have built (CS quests can also be a factor here) an interesting thing happens. Its quite possible for your cities happiness to become equal to its population. Because unhappiness (other than urbanization) can never go beyond your population, it is possible to enter "Perma Happiness". Your cities all move into 50% unhappiness (aka max unhappiness countered by happiness equal to your population).
At this point, happiness is no longer a major factor in terms of city development and growth. Now happiness is really a function of war weariness combined with ideological pressure. In general though, I find unless I am going full war monger and just ignore war weariness completely, I don't normally find the happiness effects at this point in the game strong enough to bring down my Global Number. Normally its the supply drops for war weariness that give me more pause.
A Note on Difficulty
I have made this point before but I will reiterate, that I really do see a strong effect on difficulty when it comes to happiness. There is an "Anti-snowball" that occurs. Compared to Emperor, on Immortal, the AI is able to get enough bonuses quickly enough that the median yield numbers start to infect you with greater unhappiness. Its not a lot, but it can make the difference between keeping your civ at 50% vs dropping to 45% global with -1 local unhappiness in a city (which is a -35% reduction in growth and settler production). This means on Immortal I expand slower, and grow slower, than I did on Emperor. Less growth means less production = slower infrastructure = fewer yields = more unhappiness.
Its subtle but actually quite noticeable when you change difficulties. The last time i played Emperor, happiness was a "breeze" in comparison (and an Emperor player might feel the same if they dropped to King). Meanwhile on my Deity games I notice the effect even more strongly.
This note is to showcase that its perfectly reasonable to see two players at different difficulties with very different Happiness viewpoints. If an Emperor player popped up and said "Stalker I don't know what your talking about, I find happiness very reasonable", that would not surprise me at all. So that's part of the challenge of balancing it, its not a basic yield where we can all agree that it should be increased or decreased. Happiness is highly variable depending on lux availability, difficulty, CS quests (and access to early CS allies), and of course....AI performance.