Revolution Design for MNAI-U and Extramodmod

[lfgr]

(Revised on August 14th, 2022, with input from Alexeyev_)
(Minor revisions May 2023)

In this post, I'll lay out my plans for the revolution component in MNAI-U and EMM. This is the only area where I'll make substantial balance changes, and I welcome everybody's feedback on my suggestions. Since the whole post is very long, I put some things in spoilers that might be less interesting for some.
I plan to roughly proceed in the following order:

• Improved UI.The most important aspect of this is to be more transparent about the various stability modifiers. I consider this not only an improvement by itself, but it will also make it much easier to provide well-founded feedback. This is mostly done as of version MNAI 2.9u, except the following:
  • Replace the old RevWatchAdvisor by a shortcut to the new "Revolution" page of the Customizable Domestic Advisor
  • Improve the city screen revolution bar and its tooltip
  • Use an "Instability" icon in place of the various occurrences of "stability" and "instability" in the game text.
• Balance/Improve stability. This is about when a revolution occurs, and how you can prevent revolutions. I already made several small simplifications and improvements in version 2.9u. More on this below.
• Balance/Improve revolutions. This is about what a revolution is. Currently, there are peaceful revolutions that simply consist of demands to e.g. change a civic or join some other civilizations, and violent revolutions that spawn units. I don't really know how all of this works currently, so I'll go through it and see if there are any improvements to be made. This includes e.g. balancing the number of units in a revolution.
• Additional flavor.The revolution system as a whole still doesn't feel very FfH-y yet, and I think there is a lot of potential for variation depending on civilization and religion. Areas of possible changes include:
  • Types of demands by peaceful revolutionaries
  • Types of units spawning (depending on the type of revolution; e.g. nationality-based or religion-based)
  • Civilization of a newly spawned revolutionary player (probably usually the same as the mother civ or a previous owner)
  • Revolutionary leaders (maybe depending on the type of revolution; also, when the pool of leaders is depleted, maybe generate a random name and use a generic portrait)
  • All kinds of messages; generic messages could be made more fitting to the setting, and special messages depending on the type of revolution could be added.

Below is a more detailed plan for part 2, which I consider most important for gameplay. I will not post more detailed thoughts on 3 and 4 yet; I need to get a better understanding of how things currently work to even decide what I want exactly.

Spoiler General guidelines :

• Simplicity. It should be easy to understand what is going on in your cities, and what you can do about it. UI is huge factor here, of course, but this also means that there is an upper bound on the complexity of stability calculations.
• Variation. There should be different strategies to deal with revolutions, each with their own strengths and weaknesses. For example, I'll rather have civics push into a general direction instead of providing flat boni or penalties.
• Noticeable gameplay impact. This mostly concerns expansion. I think it is important to note that a game with revolutions is intended to work differently than a game without; it's not simply an extra annoyance you'll have to deal with. Instead, it fundamentally changes your ability to expand; rushing through your enemies with an über-stack while also keeping their cities should not be viable most of the time.
• Additive modifiers. In civ, almost all modifiers are added up, i.e. +10% and +10% means +20% instead of +21% (the latter corresponds to 1.10 times 1.10). This makes designing negative modifiers harder, since we have to take care when it is possible reach -100%, but I think it is worth the consistency and clarity.

Spoiler Local vs. national (in)stability :

There are currently two ways of providing (in)stability for e.g. civics and buildings. A local (in)stability effect applies to a single city and is accumulated every turn. If this accumulated value raises above some threshold, a revolution happens. National stability effects do two things: First, it is added every turn to each city, just like local (in)stability. Second, it also accumulates, for your whole empire instead of a single city. However, as far as I can see, this accumulated value is not used anywhere except in the stability graph of the demographics screen.
You may have noticed that now a civic having "+1 local instability in every city per turn, +1 national instability per turn" doesn't make much sense; that's just +2 instability in every city every turn. There are a few instances in the code where the current (in)stability per turn in a city is used, and this does not use the national (in)stability effects. But I'm pretty sure the omission of national stability effects is not significant in any way. Thus, barring any surprises when checking the code more carefully, I plan to remove the distinction between local and national stability.

Proposed concrete effects
Below I describe (to a varying degree of detail) how the various modifiers currently (MNAI v2.9u) work, and how I plan to change them. Civic and building effects are at the end; some of those are mostly based on flavor and are definitely subject to balance changes.

THIS IS OUTDATED as of MNAI 2.9.2u.

Happiness

Spoiler How it currently works :

• If the city is unhappy (ignoring unhappiness from war weariness), then the base instability per turn is 15 * (excess unhappiness)^0.8.
  Then, several modifiers apply; -30% if the city was recently acquired (within 12 turns, gamespeed-adjusted), -30% from unhappy production (Governor's Manor and Pillar of Chains), and a more complicated bonus if the population is larger than the excess unhappiness: (excess unhappiness) / population * 100 - 100. In the latter equation, the population is capped at 12. Instability from unhappiness is theoretically unlimited.
• If the city is happy: First, excess happiness is capped by 10 or the population, whatever is lower. Then, the formula is (1.2 + (excess happiness) / pop) * (excess happiness)^0.6. If the owner of the city is a rebel and at war with the cultural owner of the city, a -50% malus applies.
  Stability from happiness is therefore no more than 2.2 * 10^0.6 ~= 9.

• War weariness is still ignored for excess unhappiness, but counted for excess happiness.
• Base instability from unhappiness is simply 100 * (excess unhappiness)/pop, capped at 100 or 10*pop, whatever is lower.
• Base stability from happiness is 10 * (excess happiness)/pop, capped at 10.
• -30% directly after acquiring. This decays by 2pp/turn, to 0% after 15 turns.
• -50% effect from unhappiness and happiness from unhappy production buildings.
• Further civic effects (see below)

Rationale: This is mostly a simplification. Unhappiness is capped by pop to ensure small cities don't go too crazy. Note that growing the population can never improve things on its own. The unhappy production effect is now neatly symmetric.

Distance

Spoiler How it currently works :

It's a mess. First, a "communications bonus" is computed.

• Base communication is 0
• Each trade route but the first gives +17
• If the city is connected to the capital: +50, then +17 if engineering is researched, then add the city's trade route modifier, add the culture rate modifier, and add half of the gold rate modifier
• Coastal cities get +25 with the coastal trading tech (Fishing) or +50 with the ocean trading tech (Sailing)
• The city's maintenance modifier is subtracted
• +100 per airlift capacity (+100 from Obsidian gate)
• The bonus cannot be negative

Then we calculate the euclidean city distance to the capital or nearest Summer/Winter palace. The formula for the effective distance is:

    307 * (city distance) / (map diameter) / ( 1 + (comm. bonus) / 100 ) ) - 666 / (map diameter)

Then the effective distance is modified by civics and buildings. Then it is multiplied by a factor between 0.5 and 2 depending on civ size. This final value (rounded) is the base location effect.
Finally, if the city is not connected to the capital, there is an additional flat 3 instability for rebels, or 5 + era + pop/3, capped at 10, for normal civs.

Proposed: First take the step distance (not euclidean distance) to the shortest government center (capital or Summer/Winter palace), or the farthest city if there is no government center. This value is modified by the DistanceMaintenancePercent of the map size, divided by the maximum plot distance (diameter for flat maps). This value, between 0 and 1, is the adjusted distance.
Base distance instability is 135 * (adjusted distance) * (0.7 + 0.1*population).

• +100% if not connected to the capital
• -50% if airlift is possible (Obsidian gate)
• Various civic modifiers
• Instability is capped at 100

Spoiler Rationale and further ideas :

• The new formula is based on the base distance maintenance of the city (see CvCity::calculateDistanceMaintenanceTimes100())
• The constants I chose try to model the typical current values, which go up to roughly 200 instability for cities in opposite corners of the map with average communication.
• The communications bonus is weird and too complicated; it seems to be an attempt at modeling more efficient connections, without computing it in any concrete way.
• The flat 666/diameter bonus is also weird, it basically subtracts 3 from the distance (assuming communication of roughly 67). Thus, it's actually very dependent on the map size.
• The use of euclidean distance is arbitrary and doesn't fit in with other game mechanics.
• The empire size effect has been outsourced.

It would be neat to compute a more accurate distance with pathfinding. But since maintenance doesn't use that either, it's probably fine.

Religion

Spoiler How it currently works :

The effects are split into good and bad effects, each with a base (in)stability and some possibly modifiers.

• State religion in city:
  • Base stability: 4
  • Holy city: +5 base stability
  • Civic boni if we own the holy city (doubled if this is the holy city):
    • Theocracy: +4
    • Religion: +2
    • Sacrifice the Weak: +5
  • +50% if at war with at least 2 other players
  • +50% from Theocracy
  • +20% from Religion, Sacrifice the Weak, Crusade
• Non-state religions in city, if we have a state religion:
  • +2.5 base instability from the first two non-state religions, +1 for each further non-state religion.
  • If heathens own the holy city: +4 from Theocracy, +1 from Religion
  • +4 if non-state-religion holy city
  • +50% from Theocracy

Proposed: Split up effects further and add some flavor.

• State religion in city:
  • Base stability: 4 for non-holy, 8 for holy city
  • +50% from war against one more heathen players
• Various civic modifiers
• Stability if we own the holy city: +4 from Theocracy
• Non-state religions in city:
  • Base instability mostly depends on state religion vs. present religions (penalties are additive):
    • Order vs. Veil: +5
    • Order vs. CoE/OO: +3
    • RoK/Emp vs. Veil: +3
    • RoK/Emp vs. CoE/OO: +2
    • All other combinations, including no state religion vs. something: +1
    • Non-styte Order and non-state Veil in city: +1
    • Double effect for holy cities (e.g. Veil holy city +10 for Order player).
  • Various civic modifiers
• Instability if heathens own the state holy city: +4 from Theocracy

Further ideas:

• Possibly remove all boni and penalties caused by CoE as state religion
• Maybe treat CoE as neutral religion

Culture rate

Spoiler How it currently works :

Take the culture rate (i.e., culture per turn), and modify it as follows, depending on era:

• Ancient: -33%
• Classical: -50%
• Medieval: -60%
  Base stability is this modified rate to the power of 0.7, capped at 10. The cap is attained at a modified culture rate of 25.
  If the player is at war with the majority culture in the city, the base stability is halved.

Proposed: Remove the era modifier and the "war with majority culture" modifier, and take the culture rate to the power of 0.6. This puts the effect somewhere between the previous ancient-era and classical-era effects. Maximum value of 10 is attained at culture rate 42.

Further ideas: Maybe buff this? Culture is rarely worth it otherwise. Might have to check what culture rates are typically attained.

Nationality

Spoiler How it currently works :

If more than 60% own nationality: Base stability is (ownPercent-60)/8, to a maximum of 5.
If more than 40% foreign nationality: Base instability is (foreignPercent-40)^2 / 200. Cap is 13, which is attained at 91% foreign nationality. The current culture rate is subtracted from this, but cannot remove more than 90% of it (this doesn't mean much due to rounding).

• If the city is recently acquired, -40% effect, and capped at 6.
• If we are at war with the majority culture, and we're not a rebel, +50% effect.
• If the majority culture is our vassal, -30% effect, and capped at 5.
• City states: +50% effect
• Liberty, Decentralization: +20% effect
• Republic, Military state: -20% effect

Proposed: The base formula is now foreignPercent^2 / 500 - 3 instability, meaning 3 stability at up to 20% foreign culture, 2 at 30%, 1 at 35%, 0 at 40%, and going up to 17 instability at 100%. However, the following modifiers apply when calculating "foreign culture", depending on the foreign player:

• -30% if the player is your vassal
• -50% if the player is the same civ as you, and you're not at war with them

After calculating base (in)stability, various civic modifiers apply.

Unhealthiness
How it currently works: First compute excess unhealthiness, where temporary unhealthiness only counts half. Base instability is 2 * (excess unhealthiness)^0.6. Rebels get -67%. Instability from unhealthiness is theoretically unbounded.

Proposed: Scale linearly: +1 instability per excess unhealthiness (temporary counts half), with a cap of 10. This is roughly lower than before in the range of 1-5 and above 15 unhealthiness, and higher between 6 and 14.

Starvation
Current and proposed: +100 instability when starving (i.e., +100 instability whenever a population point is lost).

Garrison
How it currently works: Base stability is 2 * (# defenders / 2)^0.6, which is roughly 1.3 * (# defenders)^0.6. If the city has X% defense from buildings, the base stability is increased by X% (up to 100%). Final stability is capped by 8.
Maximum stability is attained with 19 defenders at 0% defense, and with 6 defenders at 100% defense.

Proposed: Simply +1 stability per military unit, capped at 10. Each 5 population points reduce stability by 1 (but not below 0; i.e., in a 20-pop city, you need 5 defenders for +1 stability, but only 1 defender does not cause instability).
The building and civic modifiers apply.

Rationale: Linear scaling is nicer, and requiring more units for larger cities is intuitive and in line general balancing goals. It is not clear how walls are keeping the peace in the city, so it makes more sense to tie the percentage bonus to specific buildings.

City Size
How it currently works: Take the city size and subtract -3 in ancient, -4 in classical, and -5 in medieval era. If the result is positive, that's the instability per turn.

Proposed: Completely scrap this. City size has an effect on other modifiers.

Empire Size
How it currently works: The size of an empire is defined as the number of cities in the empire divided by the "target" number of cities from the world size (4 for Duel, 8 for Huge).

• Size > 2.0: +4 instability
• Size > 1.6: +3 instability
• Size > 1.4: +2 instability
• Size > 1.2: +1 instability
• Size > 1.0: +0 instability
• Size > 0.7: -1 instability
• Size <= 0.7: -2 instability

Proposed: As before, just that settlements only count 25%.

Disorder
How it currently works: If the city is in disorder, instability per turn is

• 75 normally
• 11 (15%) for recently acquired cities (12 turns at normal game speed) and rebel players
• 15 (20%) if a revolution occurred already in this city.

Proposed: If the city is in disorder, instability per turn is

• 75 normally
• If acquired within the last 15 turns, it's 5*(turns since acquired) instead
• -90% for rebel players
• -80% if a revolution occurred already in this city.

Rationale: Smoother effect for conquered cities to make it more intuitive; scrap game speed adjustments, since revolt turns are not affected by game speed either.

Crime
How it currently works: The formula is (crime - 20) / 10, where crime above 100 is not counted. This means crime below 0 gives a positive effect.
Proposed: No changes.

Culture spending
How it currently works: Base stability is the square root of the culture slider value (0 for 0%, 10 for 100%). Creative leaders get a +70% bonus.
Proposed: Scrap this. It seems like a minor effect and the culture slider already affects stability from culture rate in a more intuitive way.

Golden age
How it currently works: -20 instability if in a golden age.
Proposed: No changes.

Handicap
How it currently works: Revolutions are not affected by Handicap.
Proposed: -45% instability at settler, -30% at chieftain, -15% at warlord, and no changes from noble onwards

Civic effects
Proposed: Entering anarchy now always gives a +50 one-time instability penalty in each city.

Spoiler Specific civics :

• Despotism
  • Current: +4 instability per turn, +25% from distance, +100 instability when adopting
  • Proposed: +10% instability from distance
• City States
  • Current: +1 instability per turn, +10% from distance, +50% from foreign nationality, -50 instability when adopting
  • Proposed: +1 instability per turn, -50% from distance, +50% from nationality
• God King
  • Current: +2 instability per turn, +20% from distance, -50 instability when adopting
  • Proposed: +25% from distance
• Aristocracy
  • Current: +5 instability per turn
  • Proposed: +5 instability per turn, -25% from distance
• Theocracy
  • Current: -4 instability per turn if owning state religion holy city, +4 instability per turn if heathens own holy city, +50% effects from religions, +50 instability when adopting
  • Proposed: -4 instability per turn if owning state religion holy city, +4 instability per turn if heathens own holy city, +50% effects from religions
• Republic
  • Current: -4 instability per turn, -20% from distance, -30 instability when adopting
  • Proposed: -4 instability per turn, +25% from distance, +15% violent revolution threshold
• Religion
  • Current: -2 instability per turn if owning holy city, +1 instability per turn if heathens own holy city, +20% bonus from state religion
  • Proposed: +25% positive effects from religions
• Pacifism
  • Current: -2 instability per turn, -20 instability when adopting, -50% violent revolution threshold
  • Proposed: -5 instability per turn, -50% stability from garrison, -50% maximum stability (cap) from garrison, +20% violent revolution threshold
• Nationhood
  • Current: -1 instability per turn, -20 instability when adopting
  • Proposed: -1 instability per turn, +25% from nationality
• Sacrifice the weak
  • Current: +3 instability per turn, +20% from distance, -5 instability if owning Veil holy city, +20% stability from state religion, +200 instability when adopting, +50% violent revolution threshold
  • Proposed: Extra +150 Stability from adopting, +20% effects from religions, -25% effects from happiness/unhappiness, +25% stability from garrison, +5 maximum stability (cap) from garrison, +2 instability per turn, -30% violent revolution threshold
  • Rationale: The instability is meant to be countered by garrison effects. So you get more out of a big garrison, but are also punished more by a small garrison.
• Social Order
  • Current: +100 instability when adopting, -30% violent revolution threshold
  • Proposed: +25% effects from happiness/unhappiness, +2 maximum stability from happiness, +1 instability per turn, -10% violent revolution threshold
  • Rationale: Recall that this civic gives +1 happiness per garrisoned unit. Similar to Sacrifice the weak, this amplifies garrison effects.
• Consumption
  • Current: No effect
  • Proposed: +25% effects from happiness/unhappiness
• Scholarship
  • Current: No effect
  • Proposed: No effect
• Liberty
  • Current: -2 instability per turn, +20% penalty from foreign nationality, -50 instability when adopting, +30% violent revolution threshold
  • Proposed: +50% from culture rate, +5 maximum stability (cap) from culture rate, +20% violent revolution threshold
• Crusade
  • Current: -3 instability per turn, +20% bonus from state religion, +50 instability when adopting
  • Proposed: -3 instability per turn, +20% bonus from state religion, +25% bonus from garrison, -20% violent revolution threshold
• Tribalism
  • Current: +10% instability from distance
  • Proposed: +10% instability from distance
• Apprenticeship
  • Current: No effect
  • Proposed: No effect
• Slavery
  • Current: +4 instability per turn, +50 instability when adopting, +50% violent revolution threshold
  • Proposed: +2 instability per turn, +2 maximum stability from garrison, -30% violent revolution threshold
• Arete
  • Current: -2 instability per turn, -25% instability from distance
  • Proposed: +25% stability from state religion
• Military State
  • Current: -20% instability from nationality, -30% violent revolution threshold
  • Proposed: -25% instability from nationality, +25% stability from garrison, +2 maximum stability from garrison, +2 instability per turn, -10% violent revolution threshold
• Caste System
  • Current: +4 instability per turn, +20 instability when adopting
  • Proposed: No effect
• Guilds
  • Current: +2 instability per turn
  • Proposed: No effect
• Decentralization
  • Current: +5 instability per turn, +20% instability from distance, +20% instability from nationality, +40 instability when adopting, +50% violent revolution threshold
  • Proposed: No effect
• Agrarianism
  • Current: No effect
  • Proposed: No effect
• Conquest
  • Current: +1 instability per turn
  • Proposed: +1 instability per turn, -10% instability from nationality
• Mercantilism
  • Current: No effect
  • Proposed: No effect
• Foreign Trade
  • Current: -20 instability when adopting
  • Proposed: No effect
• Guardian of nature
  • Current: -2 instability per turn
  • Proposed: +25% stability from state religion
• No Membership
  • Current: No effect
  • Proposed: No effect
• Overcouncil
  • Current: No effect
  • Proposed: +25% stability from happiness, +10% violent revolution threshold
• Undercouncil
  • Current: No effect
  • Proposed: -25% instability from crime

Buildings

Spoiler Buildings :

• Archery Range
  • Current: -1 instability/turn
  • Proposed: +25% stability from garrison
• Asylum
  • Current and proposed: -2 instability/turn
• Basilica
  • Current: No effect
  • Proposed: +50% stability from garrison
• Adventurers' Guild
  • Current: No effect
  • Proposed: -25% instability from distance
• Breeding Pit
  • Current: No effect
  • Proposed: -25% effect from happiness/unhappiness
• Chancel of Guardians
  • Current and proposed: -1 instability/turn
• Citadel of Light
  • Current: No effect
  • Proposed: -1 instability/turn
• Command Post
  • Current: No effect
  • Proposed: +25% stability from garrison
• Courthouse
  • Current and proposed: -1 instability/turn, -10% instability from distance
• Governor's Manor
  • Current: -1 instability/turn, -10% instability from distance
  • Proposed: -1 instability/turn, -10% instability from distance, -50% instability from unhappiness
• Dungeon
  • Current: -1 instability/turn, -10% instability from distance
  • Proposed: -1 instability/turn
• Elder Council
  • Current and proposed: -1 instability/turn
• Market
  • Current: -1 instability/turn
  • Proposed: No effect
• Monument
  • Current: -1 instability/turn
  • Proposed: No effect (but note the culture rate effect)
• Pagan Temple, Desert Shrine, Temple of the Hand
  • Current: No effect
  • Proposed: -1 instability/turn
• Smugglers' Port
  • Current: +1 instability/turn, +5% distance instability
  • Proposed: +1 instability/turn
• Temple of ...
  • Current: No effect
  • Proposed: -1 instability
• Training Yard, Arena, Sculptor's Studio
  • Current: No effect
  • Proposed: +25% stability from garrison

Spoiler Wonders :

• Altar of the Luonnotar
  • Current: No effect
  • Proposed: -1 instability/turn
• Altar of the Luonnotar (anointed)
  • Current: No effect
  • Proposed: -2 instability/turn
• Altar of the Luonnotar (blessed)
  • Current: No effect
  • Proposed: -2 instability/turn, -1 instability/turn in all cities
• Altar of the Luonnotar (consecrated)
  • Current: No effect
  • Proposed: -3 instability/turn, -1 instability/turn in all cities
• Altar of the Luonnotar (divine)
  • Current: No effect
  • Proposed: -4 instability/turn, -2 instability/turn in all cities
• Altar of the Luonnotar (exalted)
  • Current: No effect
  • Proposed: -8 instability/turn, -2 instability/turn in all cities
• Altar of the Luonnotar (final)
  • Current: No effect
  • Proposed: -10 instability/turn in all cities
• Bazaar of Mammon
  • Current: No effect
  • Proposed: -25% effect from crime in all cities
• Bone Palace
  • Current: No effect
  • Proposed: -1 instability/turn in all cities
• Deruptus Brewing House
  • Current: No effect
  • Proposed: +10% effect from happiness/unhappiness
• Code of Junil
  • Current: No effect
  • Proposed: -50% instability from disorder in all cities
• Dies Diei
  • Current: No effect
  • Proposed: -5 instability/turn
• The Eyes and Ears Network
  • Current: No effect
  • Proposed: -1 instability/turn in all cities
• Summer Palace, Winter Palace
  • Current: -2 instability/turn, -20% instability from distance
  • Proposed: No effect (already reduces distance instability in nearby cities)
• Heron Throne
  • Current: No effect
  • Proposed: No effect
• Heroic Epic
  • Current: No effect
  • Proposed: -3 instability/turn
• National Epic
  • Current: No effect
  • Proposed: -3 instability/turn
• Nox Noctis
  • Current: No effect
  • Proposed: -100% effect from crime
• Palace (and replacements)
  • Current and proposed: -5 instability/turn, -2 instability/turn in all cities
  • Proposed: -5 instability/turn (all cities bonus moved)
• Pillar of Chains
  • Current: No effect
  • Proposed: -50% instability from unhappiness
• Shrine of the Champion
  • Current: No effect
  • Proposed: -3 instability/turn
• Song of Autumn
  • Current: No effect
  • Proposed: +25% stability from culture rate in all cities
• Stigmata on the Unborn
  • Current: No effect
  • Proposed: +2 maximum stability from garrison in all cities
• Tablets of Bambur
  • Current: No effect
  • Proposed: -1 instability in all cities
• The Necronomicon
  • Current: No effect
  • Proposed: -25% effect from happiness/unhappiness in all cities
• Tower of Complacency
  • Current: -10 instability/turn, -1 instability/turn in all cities, -50% instability from distance
  • Proposed: No effect (note the excess happiness bonus)

Spoiler Special buildings :

• Academy
  • Current: -3 instability/turn
  • Proposed: -2 instability/turn
• Hope
  • Current: No effect
  • Proposed: -1 instability/turn
• Unyielding Order
  • Current and proposed: -4 instability/turn, -20% instability from distance

Misc. effects
Current:

• Infernals are exempt from all revolution effects

Proposed:

• -2 base instability in all cities (previously bound to palace).
• Infernals are exempt from all revolution effects
• No revolts while Elohim Sanctuary is active
• No revolts while affected by Stasis
• A player with rebel status gets -90% instability (how long does the rebel status last?)

 

[Alekseyev_]

I only recently started using the revolution game option, so I can't really tell much about my experience with it, but thank you very much for this detailed post! This really helps a lot with understanding the current mechanics.

When I have more experience, maybe I can comment on your ideas, but I am sure you will do great either way.

 

[UnderKingDuck]

This looks really good. I do look forward to trying this system out.

 

[Artiss]

As I played with revolutions "on" (as it is the only way for limiting of snowballing conquests), internal mechanic of this feature needs some polishing of course, but main problems are:
1. Leaders. It's a nonsense when the spawning leaders are major leaders (especially of another nation or even race to boot). In my last game in sheaim cities conquered by orcs broke out a rebellion led by Varn Gosam. It's so... critically... unlore.
In my opinion there must be some number of generic portraits with random names for every nation.
2. Adapting AI. Computer opponents can't grasp this feature and its mechanic of different modifiers at all.

 

[Alekseyev_]

Here are some first opinions on your proposed mechanics:

• Improved UI.
  
• Balance/Improve stability.
  
• Balance/Improve revolutions.
  
• Additional flavor.

Spoiler General guidelines :

• Simplicity. It should be easy to understand what is going on in your cities, and what you can do about it. UI is huge factor here, of course, but this also means that there is an upper bound on the complexity of stability calculations.
• Variation. There should be different strategies to deal with revolutions, each with their own strengths and weaknesses. For example, I'll rather have civics push into a general direction instead of providing flat boni or penalties.
• Noticeable gameplay impact. This mostly concerns expansion. I think it is important to note that a game with revolutions is intended to work differently than a game without; it's not simply an extra annoyance you'll have to deal with. Instead, it fundamentally changes your ability to expand; rushing through your enemies with an über-stack while also keeping their cities should not be viable most of the time.
• Additive modifiers. In civ, almost all modifiers are added up, i.e. +10% and +10% means +20% instead of +21% (the latter corresponds to 1.10 times 1.10). This makes designing negative modifiers harder, since we have to take care when it is possible reach -100%, but I think it is worth the consistency and clarity.

All of these sound like good endeavours/ideas.

Spoiler Local vs. national (in)stability :

There are currently two ways of providing (in)stability for e.g. civics and buildings. A local (in)stability effect applies to a single city and is accumulated every turn. If this accumulated value raises above some threshold, a revolution happens. National stability effects do two things: First, it is added every turn to each city, just like local (in)stability. Second, it also accumulates, for your whole empire instead of a single city. However, as far as I can see, this accumulated value is not used anywhere except in the stability graph of the demographics screen.
You may have noticed that now a civic having "+1 local instability in every city per turn, +1 national instability per turn" doesn't make much sense; that's just +2 instability in every city every turn. There are a few instances in the code where the current (in)stability per turn in a city is used, and this does not use the national (in)stability effects. But I'm pretty sure the omission of national stability effects is not significant in any way. Thus, barring any surprises when checking the code more carefully, I plan to remove the distinction between local and national stability.

Could some sort of national instability be used to have a country fracture into multiple regions when the stability collapses? A very locally stable city could reliably remain in control of the player when an otherwise instable nation breaks apart. But then again, the same effect might be achieved by just using local instability... National (in)stability feels like something that should be there, but I can see how it might not actually make a lot of sense from a gameplay/implementation perspective

Proposed concrete effects
Happiness

Spoiler :

Spoiler How it currently works :

• If the city is unhappy (ignoring unhappiness from war weariness), then the base instability per turn is 15 * (excess unhappiness)^0.8.
  Then, several modifiers apply; -30% if the city was recently acquired (within 12 turns, gamespeed-adjusted), -30% from unhappy production (Governor's Manor and Pillar of Chains), and a more complicated bonus if the population is larger than the excess unhappiness: (excess unhappiness) / population * 100 - 100. In the latter equation, the population is capped at 12. Instability from unhappiness is theoretically unlimited.
  
• If the city is happy: First, excess happiness is capped by 10 or the population, whatever is lower. Then, the formula is (1.2 + (excess happiness) / pop) * (excess happiness)^0.6. If the owner of the city is a rebel and at war with the cultural owner of the city, a -50% malus applies.
  Stability from happiness is therefore no more than 2.2 * 10^0.6 ~= 9.

• War weariness is still ignored for excess unhappiness, but counted for excess happiness.
• Base instability from unhappiness is simply 100 * (excess unhappiness)/pop, capped at 100 or 10*pop, whatever is lower.
• Base stability from happiness is 10 * (excess happiness)/pop, capped at 10.
• -30% directly after acquiring. This decays by 2pp/turn, to 0% after 15 turns.
• -50% effect from unhappiness and happiness from unhappy production buildings.
• Further civic effects (see below)

Rationale: This is mostly a simplification. Unhappiness is capped by pop to ensure small cities don't go too crazy. Note that growing the population can never improve things on its own. The unhappy production effect is now neatly symmetric.

Distance

Spoiler :

Spoiler How it currently works :

It's a mess. First, a "communications bonus" is computed.

• Base communication is 0
• Each trade route but the first gives +17
• If the city is connected to the capital: +50, then +17 if engineering is researched, then add the city's trade route modifier, add the culture rate modifier, and add half of the gold rate modifier
• Coastal cities get +25 with the coastal trading tech (Fishing) or +50 with the ocean trading tech (Sailing)
• The city's maintenance modifier is subtracted
• +100 per airlift capacity (+100 from Obsidian gate)
• The bonus cannot be negative

Then we calculate the euclidean city distance to the capital or nearest Summer/Winter palace. The formula for the effective distance is:

    307 * (city distance) / (map diameter) / ( 1 + (comm. bonus) / 100 ) ) - 666 / (map diameter)

Then the effective distance is modified by civics and buildings. Then it is multiplied by a factor between 0.5 and 2 depending on civ size. This final value (rounded) is the base location effect.
Finally, if the city is not connected to the capital, there is an additional flat 3 instability for rebels, or 5 + era + pop/3, capped at 10, for normal civs.

Proposed: First take the step distance (not euclidean distance) to the shortest government center (capital or Summer/Winter palace), or the farthest city if there is no government center. This value is modified by the DistanceMaintenancePercent of the map size, divided by the maximum plot distance (diameter for flat maps). This value, between 0 and 1, is the adjusted distance.
Base distance instability is 135 * (adjusted distance) * (0.7 + 0.1*population).

• +100% if not connected to the capital
• -50% if airlift is possible (Obsidian gate)
• Various civic modifiers
• Instability is capped at 100

Spoiler Rationale and further ideas :

• The new formula is based on the base distance maintenance of the city (see CvCity::calculateDistanceMaintenanceTimes100())
• The constants I chose try to model the typical current values, which go up to roughly 200 instability for cities in opposite corners of the map with average communication.
• The communications bonus is weird and too complicated; it seems to be an attempt at modeling more efficient connections, without computing it in any concrete way.
• The flat 666/diameter bonus is also weird, it basically subtracts 3 from the distance (assuming communication of roughly 67). Thus, it's actually very dependent on the map size.
• The use of euclidean distance is arbitrary and doesn't fit in with other game mechanics.
• The empire size effect has been outsourced.

It would be neat to compute a more accurate distance with pathfinding. But since maintenance doesn't use that either, it's probably fine.

These sound sensible, the current system seems way too complicated and your simplification will be much easier to understand - as well as hopefully more balanced.

Religion

Spoiler How it currently works :

The effects are split into good and bad effects, each with a base (in)stability and some possibly modifiers.

• State religion in city:
  • Base stability: 4
  • Holy city: +5 base stability
  • Civic boni if we own the holy city (doubled if this is the holy city):
    • Theocracy: +4
    • Religion: +2
    • Sacrifice the Weak: +5
  • +50% if at war with at least 2 other players
  • +50% from Theology
  • +20% from Religion, Sacrifice the Weak, Crusade
• Non-state religions in city, if we have a state religion:
  • +2.5 base instability from the first two non-state religions, +1 for each further non-state religion.
  • If heathens own the holy city: +4 from Theocracy, +1 from Religion
  • +4 if non-state-religion holy city
  • +50% from Theology

Proposed: Split up effects further and add some flavor.

• State religion in city:
  • Base stability: 4 for non-holy, 8 for holy city
  • +50% from war against one more heathen players
• Various civic modifiers
  
• Stability if we own the holy city: +4 from Theocracy
• Non-state religions in city:
  • Base instability depends on state religion vs. present religions:
    • Order vs. Veil: +5
    • Order vs. CoE/OO: +3
    • RoK/Emp vs. Veil: +3
    • RoK/Emp vs. CoE/OO: +2
    • All other combinations, including no state religion vs. something: +1
  • Double effect for holy cities (e.g. Veil holy city +10 for Order player).
  • +50% from Theology
  • +20% from Crusade
• Instability if heathens own the state holy city: +4 from Theocracy

Further ideas:

• Possibly remove all boni and penalties caused by CoE.
• Maybe also depend instability on the interaction between non-state religions? For example, Order and Veil together could cause lots of instability, even if the state religion is e.g. FoL. Maybe only for Grigori?

With "+50% from Theology", do you mean that after researching that tech, instability from religions becomes a bigger issue? This seems like quite an interesting concept, I like it. Or were you referring to Theocracy?
As for CoE, this one could probably cause instability. After all, CoE means that everyone is out for their own and trickery and treason are rampant.
I definitely think that opposite religions should cause instability even if neither of them is a state religion. Also, the religion clashes are additive, right? So Order + Veil + OO would mean +8 instability.
Religion civic should certainly affect religion (in)stability.

Culture rate

Spoiler :

Spoiler How it currently works :

Take the culture rate, and modify it as follows, depending on era:

• Ancient: -33%
• Classical: -50%
• Medieval: -60%
  Base stability is this modified rate to the power of 0.7, capped at 10. The cap is attained at a modified culture rate of 25.
  If the player is at war with the majority culture in the city, the base stability is halved.

Proposed: Remove the era modifier and the "war with majority culture" modifier, and take the culture rate to the power of 0.6. This puts the effect somewhere between the previous ancient-era and classical-era effects. Maximum value of 10 is attained at culture rate 42.

Further ideas: Maybe buff this? Culture is rarely worth it otherwise. Might have to check what culture rates are typically attained.

Culture rate means culture per turn, right? I feel like in FfH it's quite often the case that you have either fairly low rates (monument, maybe theater) or pretty high ones (wonders), due to base culture being fairly rare and temples multiplying the output. I think this system as proposed might work, but I don't really have enough insight. Might be worth just testing.

Nationality

Spoiler How it currently works :

If more than 60% own nationality: Base stability is (ownPercent-60)/8, to a maximum of 5.
If more than 40% foreign nationality: Base instability is (foreignPercent-40)^2 / 200. Cap is 13, which is attained at 91% foreign nationality. The current culture rate is subtracted from this, but cannot remove more than 90% of it (this doesn't mean much due to rounding).

• If the city is recently acquired, -40% effect, and capped at 6.
• If we are at war with the majority culture, and we're not a rebel, +50% effect.
• If the majority culture is our vassal, -30% effect, and capped at 5.
• City states: +50% effect
• Liberty, Decentralization: +20% effect
• Republic, Military state: -20% effect

Proposed: The base formula is now foreignPercent^2 / 500 - 3 instability, meaning 3 stability at up to 20% foreign culture, 2 at 30%, 1 at 35%, 0 at 40%, and going up to 17 instability at 100%. However, the following modifiers apply when calculating "foreign culture", depending on the foreign player:

• -30% if the player is your vassal
• -50% if the player is the same civ as you, and you're not at war with them

After calculating base (in)stability, various civic modifiers apply.

No real opinion here, Except maybe that a few civics might make sense to apply here, such as Military state from before. (Not the city states +50% though). Liberty giving more instability from nationality is interesting, while largely counteracted by its great boost to culture rate, it sounds like a reasonable concept - in a more liberal society, the people of foreign nationality might feel more inclined to go their own ways.
Edit: But you already did that further below, so all good.

Unhealthiness
How it currently works: First compute excess unhealthiness, where temporary unhealthiness only counts half. Base instability is 2 * (excess unhealthiness)^0.6. Rebels get -67%. Instability from unhealthiness is theoretically unbounded.

Proposed: Scale linearly: +1 instability per excess unhealthiness (temporary counts half), with a cap of 10. This is roughly lower than before in the range of 1-5 and above 15 unhealthiness, and higher between 6 and 14.

I suppose that a fresh rebellion should still get a reduced impact from this, so they don't collapse themselves for the very same reason.

Starvation
Current and proposed: +100 instability when starving (i.e., +100 instability whenever a population point is lost).

Garrison
How it currently works: Base stability is 2 * (# defenders / 2)^0.6, which is roughly 1.3 * (# defenders)^0.6. If the city has X% defense from buildings, the base stability is increased by X% (up to 100%). Final stability is capped by 8.
Maximum stability is attained with 19 defenders at 0% defense, and with 6 defenders at 100% defense.

Proposed: Simply +1 stability per military unit, capped at 10. Each 5 population points reduce stability by 1 (but not below 0; i.e., in a 20-pop city, you need 5 defenders for +1 stability, but only 1 defender does not cause instability).
The building and civic modifiers apply.

Rationale: Linear scaling is nicer, and requiring more units for larger cities is intuitive and in line general balancing goals. It is not clear how walls are keeping the peace in the city, so it makes more sense to tie the percentage bonus to specific buildings.

City Size
How it currently works: Take the city size and subtract -3 in ancient, -4 in classical, and -5 in medieval era. If the result is positive, that's the instability per turn.

Proposed: Completely scrap this. City size has an effect on other modifiers.

Empire Size
How it currently works: The size of an empire is defined as the number of cities in the empire divided by the "target" number of cities from the world size (4 for Duel, 8 for Huge).

• Size > 2.0: +4 instability
• Size > 1.6: +3 instability
• Size > 1.4: +2 instability
• Size > 1.2: +1 instability
• Size > 1.0: +0 instability
• Size > 0.7: -1 instability
• Size <= 0.7: -2 instability

Proposed: As before, just that settlements only count 25%.

Disorder
How it currently works: If the city is in disorder, instability per turn is

• 75 normally
• 11 (15%) for recently acquired cities (12 turns at normal game speed) and rebel players
• 15 (20%) if a revolution occurred already in this city.

Proposed: If the city is in disorder, instability per turn is

• 75 normally
• If acquired within the last 15 turns, it's 5*(turns since acquired) instead
• -90% for rebel players
• -80% if a revolution occurred already in this city.

Rationale: Smoother effect for conquered cities to make it more intuitive; scrap game speed adjustments, since revolt turns are not affected by game speed either.

Crime
How it currently works: The formula is (crime - 20) / 10, where crime above 100 is not counted. This means crime below 0 gives a positive effect.
Proposed: No changes.

Culture spending
How it currently works: Base stability is the square root of the culture slider value (0 for 0%, 10 for 100%). Creative leaders get a +70% bonus.
Proposed: Scrap this. It seems like a minor effect and the culture slider already affects stability from culture rate in a more intuitive way.

Golden age
How it currently works: -20 instability if in a golden age.
Proposed: No changes.

Handicap
How it currently works: Revolutions are not affected by Handicap.
Proposed: -45% instability at settler, -30% at chieftain, -15% at warlord, and no changes from noble onwards

All good!

Just one question regarding crime: What's the calculation of that, aside from buildings and wonders?

Civic effects
Proposed: Entering anarchy now always gives a +50 one-time instability penalty in each city.

Spoiler Specific civics :

• Despotism
  • Current: +4 instability per turn, +25% from distance, +100 instability when adopting
  • Proposed: +10% instability from distance
• City States
  • Current: +1 instability per turn, +10% from distance, +50% from foreign nationality, -50 instability when adopting
  • Proposed: +1 instability per turn, -50% from distance, +50% from nationality
• God King
  • Current: +2 instability per turn, +20% from distance, -50 instability when adopting
  • Proposed: +25% from distance
• Aristocracy
  • Current: +5 instability per turn
  • Proposed: +5 instability per turn, -25% from distance
• Theocracy
  • Current: -4 instability per turn if owning state religion holy city, +4 instability per turn if heathens own holy city, +50% effects from religions, +50 instability when adopting
  • Proposed: -4 instability per turn if owning state religion holy city, +4 instability per turn if heathens own holy city, +50% effects from religions
• Republic
  • Current: -4 instability per turn, -20% from distance, -30 instability when adopting
  • Proposed: -4 instability per turn, +25% from distance.
• Religion
  • Current: -2 instability per turn if owning holy city, +1 instability per turn if heathens own holy city, +20% bonus from state religion
  • Proposed: +25% positive effects from religions
• Pacifism
  • Current: -2 instability per turn, -20 instability when adopting
  • Proposed: -5 instability per turn, -50% stability from garrison, -50% maximum stability (cap) from garrison
• Nationhood
  • Current: -1 instability per turn, -20 instability when adopting
  • Proposed: -1 instability per turn, +25% from nationality
• Sacrifice the weak
  • Current: +3 instability per turn, +20% from distance, -5 instability if owning Veil holy city, +20% stability from state religion, +200 instability when adopting
  • Proposed: +20% effects from religions, -25% effects from happiness/unhappiness, +25% stability from garrison, +5 maximum stability (cap) from garrison, +2 instability per turn
  • Rationale: The instability is meant to be countered by garrison effects. So you get more out of a big garrison, but are also punished more by a small garrison.
• Social Order
  • Current: +100 instability when adopting
  • Proposed: +25% effects from happiness/unhappiness, +2 maximum stability from happiness, +1 instability per turn
  • Rationale: Recall that this civic gives +1 happiness per garrisoned unit. Similar to Sacrifice the weak, this amplifies garrison effects.
• Consumption
  • Current: No effect
  • Proposed: +25% effects from happiness/unhappiness
• Scholarship
  • Current: No effect
  • Proposed: No effect
• Liberty
  • Current: -2 instability per turn, +20% penalty from foreign nationality, -50 instability when adopting
  • Proposed: +50% from culture rate, +5 maximum stability (cap) from culture rate
• Crusade
  • Current: -3 instability per turn, +20% bonus from state religion, +50 instability when adopting
  • Proposed: -3 instability per turn, +20% bonus from state religion, +25% bonus from garrison
• Tribalism
  • Current: +10% instability from distance
  • Proposed: +10% instability from distance
• Apprenticeship
  • Current: No effect
  • Proposed: No effect
• Slavery
  • Current: +4 instability per turn, +50 instability when adopting
  • Proposed: +2 instability per turn, +2 maximum stability from garrison
• Arete
  • Current: -2 instability per turn, -25% instability from distance
  • Proposed: +25% stability from state religion
• Military State
  • Current: -20% instability from nationality
  • Proposed: -25% instability from nationality, +25% stability from garrison, +2 maximum stability from garrison, +2 instability per turn
• Caste System
  • Current: +4 instability per turn, +20 instability when adopting
  • Proposed: No effect
• Guilds
  • Current: +2 instability per turn
  • Proposed: No effect
• Decentralization
  • Current: +5 instability per turn, +20% instability from distance, +20% instability from nationality, +40 instability when adopting
  • Proposed: No effect
• Agrarianism
  • Current: No effect
  • Proposed: No effect
• Conquest
  • Current: +1 instability per turn
  • Proposed: +1 instability per turn, -10% instability from nationality
• Mercantilism
  • Current: No effect
  • Proposed: No effect
• Foreign Trade
  • Current: -20 instability when adopting
  • Proposed: No effect
• Guardian of nature
  • Current: -2 instability per turn
  • Proposed: +25% stability from state religion
• No Membership
  • Current: No effect
  • Proposed: No effect
• Overcouncil
  • Current: No effect
  • Proposed: +25% stability from happiness
• Undercouncil
  • Current: No effect
  • Proposed: -25% instability from crime

Removing the "adoption" stability hit makes sense in all but one case: I would definitely keep it for "Sacrifice the Weak". This civic is such a brutal change in values with an equally brutal effect on the population, that I could very much see a change to it as the cause for a few cities to try to secede. (Can either put it to 150 to get the old +200 in conjunction with the 50 from anarchy, or just leave it at 200 for a really major impact). Similarly perhaps about the +50 from slavery.
All in all these look good and reasonable.

Buildings

Spoiler Buildings :

• Archery Range
  • Current: -1 instability/turn
  • Proposed: +25% stability from garrison
• Asylum
  • Current and proposed: -2 instability/turn
• Basilica
  • Current: No effect
  • Proposed: +50% stability from garrison
• Adventurers' Guild
  • Current: No effect
  • Proposed: -25% instability from distance
• Breeding Pit
  • Current: No effect
  • Proposed: -25% effect from happiness/unhappiness
• Chancel of Guardians
  • Current and proposed: -1 instability/turn
• Citadel of Light
  • Current: No effect
  • Proposed: -1 instability/turn
• Command Post
  • Current: No effect
  • Proposed: +25% stability from garrison
• Courthouse, Governor's Manor
  • Current and proposed: -1 instability/turn, -10% instability from distance
• Dungeon
  • Current: -1 instability/turn, -10% instability from distance
  • Proposed: -1 instability/turn
• Elder Council
  • Current and proposed: -1 instability/turn
• Market
  • Current: -1 instability/turn
  • Proposed: No effect
• Monument
  • Current: -1 instability/turn
  • Proposed: No effect (but note the culture rate effect)
• Pagan Temple, Desert Shrine, Temple of the Hand
  • Current: No effect
  • Proposed: -1 instability/turn
• Smugglers' Port
  • Current: +1 instability/turn, +5% distance instability
  • Proposed: No effect (but +1 instability from crime rate)
• Temple of ...
  • Current: No effect
  • Proposed: -1 instability
• Training Yard, Arena, Sculptor's Studio
  • Current: No effect
  • Proposed: +25% stability from garrison

Smugglers' Ports are fairly useful, a good building limited to the Undercouncil. (Which also gives you -25% instability from crime with your changes.) I would keep the +1 instability on them. I like the buff to Adventurers' Guilds. Also good stuff.

Spoiler Wonders :

• Altar of the Luonnotar
  • Current: No effect
  • Proposed: -1 instability/turn
• Altar of the Luonnotar (anointed)
  • Current: No effect
  • Proposed: -2 instability/turn
• Altar of the Luonnotar (blessed)
  • Current: No effect
  • Proposed: -3 instability/turn in all cities
• Altar of the Luonnotar (consecrated)
  • Current: No effect
  • Proposed: -4 instability/turn in all cities
• Altar of the Luonnotar (divine)
  • Current: No effect
  • Proposed: -5 instability/turn in all cities
• Altar of the Luonnotar (exalted)
  • Current: No effect
  • Proposed: -10 instability/turn in all cities
• Altar of the Luonnotar (final)
  • Current: No effect
  • Proposed: -20 instability/turn in all cities
• Bazaar of Mammon
  • Current: No effect
  • Proposed: -25% effect from crime in all cities
• Bone Palace
  • Current: No effect
  • Proposed: -1 instability/turn in all cities
• Deruptus Brewing House
  • Current: No effect
  • Proposed: +10% effect from happiness/unhappiness
• Code of Junil
  • Current: No effect
  • Proposed: -50% instability from disorder in all cities
• Dies Diei
  • Current: No effect
  • Proposed: -5 instability/turn
• The Eyes and Ears Network
  • Current: No effect
  • Proposed: -1 instability/turn in all cities
• Summer Palace, Winter Palace
  • Current: -2 instability/turn, -20% instability from distance
  • Proposed: No effect (already reduces distance instability in nearby cities)
• Heron Throne
  • Current: No effect
  • Proposed: -5 instability/turn
• Heroic Epic
  • Current: No effect
  • Proposed: -3 instability/turn
• National Epic
  • Current: No effect
  • Proposed: -3 instability/turn
• Nox Noctis
  • Current: No effect
  • Proposed: -100% effect from crime
• Palace (and replacements)
  • Current and proposed: -5 instability/turn, -2 instability/turn in all cities
• Shrine of the Champion
  • Current: No effect
  • Proposed: -3 instability/turn
• Song of Autumn
  • Current: No effect
  • Proposed: +25% stability from culture rate in all cities
• Stigmata on the Unborn
  • Current: No effect
  • Proposed: +2 maximum stability from garrison in all cities
• Tablets of Bambur
  • Current: No effect
  • Proposed: -1 instability in all cities
• The Necronomicon
  • Current: No effect
  • Proposed: -25% effect from happiness/unhappiness in all cities
• Tower of Complacency
  • Current: -10 instability/turn, -1 instability/turn in all cities, -50% instability from distance
  • Proposed: No effect (note the excess happiness bonus)

An Altar of the Luonnatar of lower levels is not hard to get, and on the later ones the numbers escalate quite a bit. I think the effects for all cities are too much. I would greatly reduce these, or keep them local. Or both: Full effect as above locally, and a small effect everywhere.

Bazaar of Mammon wouldn't change the cap, just make it slower to reach, right? Keep in mind it is a building that everyone will have at some point.
Heron Throne seems too good, this is already a super powerful early wonder with its +8 culture and water hammers. I'd leave the extra stability effect out.
Palace -2 in all cities represents a base stability?
Rest makes sense for now.

Spoiler Special buildings :

• Academy
  • Current: -3 instability/turn
  • Proposed: -2 instability/turn
• Dwarven Vault (Empty)
  • Current: No effect
  • Proposed: +2 instability/turn
• Dwarven Vault (Low)
  • Current: No effect
  • Proposed: +1 instability/turn
• Dwarven Vault
  • Current and proposed: No effect
• Dwarven Vault (Stocked)
  • Current: No effect
  • Proposed: -1 instability/turn
• Dwarven Vault (Abundant)
  • Current: No effect
  • Proposed: -2 instability/turn
• Dwarven Vault (Full)
  • Current: No effect
  • Proposed: -3 instability/turn
• Dwarven Vault (Overflowing)
  • Current: No effect
  • Proposed: -4 instability/turn
• Hope
  • Current: No effect
  • Proposed: -1 instability/turn
• Unyielding Order
  • Current and proposed: -4 instability/turn, -20% instability from distance

I would remove the Dwarven Vault effect. It already gives extra unhappiness/happiness, and other very strong benefits if rich. A wealthy Khazad empire would be extremely stable with this, while a poor Khazad would be punished even harder.

Misc. effects
Current:

• Infernals are exempt from all revolution effects

Proposed:

• Infernals are exempt from all revolution effects
• No revolts while Elohim Sanctuary is active
• No revolts while affected by Stasis

Very sensible.

 

[Alekseyev_]

One more thing to look at would probably be the "trend" value in the city screen's stability info. The number typically does not match the current values if there had been any large temporary effects, but its calculation is not clear.

 

[fallfgrace]

This looks really amazing, for me Revolutions is a stand out feature for Civ4. It suits FFH2 so much, as revolutions create emergent gameplay, it would be a boon for this mod to have further developed revolutions and for that I salute you. I will be keeping close watch.

 

[lfgr]

Finally found some time to revise this. Thanks everybody for the comments. I updated the main post.

1. Leaders. It's a nonsense when the spawning leaders are major leaders (especially of another nation or even race to boot). In my last game in sheaim cities conquered by orcs broke out a rebellion led by Varn Gosam. It's so... critically... unlore.
In my opinion there must be some number of generic portraits with random names for every nation.

This shouldn't happen regularly already. Though I believe there is a chance that Malakim spawn regardless of host civ when there's desert around. I guess that's a bit overengineered.
I agree that generic portraits with random names are probably best, though I would add an option to make major/minor leader spawning possible. I've been slowly working to collect fitting names. Finding nice leader portraits with permissive licenses seems surprisingly hard, though.

2. Adapting AI. Computer opponents can't grasp this feature and its mechanic of different modifiers at all.

I will try to work towards this, but don't expect too much. In the end, I expect that I'll help the AI by just lowering the penalties that it can't handle well.

One more thing to look at would probably be the "trend" value in the city screen's stability info. The number typically does not match the current values if there had been any large temporary effects, but its calculation is not clear.

This is a relict of the old revolution advisor and will be removed. The city screen stability info will be adapted to contain information that's currently in the new "revolution screen" in the custom domestic advisor.

Could some sort of national instability be used to have a country fracture into multiple regions when the stability collapses? A very locally stable city could reliably remain in control of the player when an otherwise instable nation breaks apart. But then again, the same effect might be achieved by just using local instability... National (in)stability feels like something that should be there, but I can see how it might not actually make a lot of sense from a gameplay/implementation perspective

Yeah, I agree the local-only approach feels a bit weird, but that's how it works and I'm probably not going to change it.

With "+50% from Theology", do you mean that after researching that tech, instability from religions becomes a bigger issue? This seems like quite an interesting concept, I like it. Or were you referring to Theocracy?

Yeah, I meant Theocracy.

As for CoE, this one could probably cause instability. After all, CoE means that everyone is out for their own and trickery and treason are rampant.

First, I clarified that I meant having CoE as state religion should maybe. My thinking was that CoE is not a real religion, so it shouldn't give boni. On the penalty side, my understanding is that CoE uses an effective, but subtle form of controlling the populace; so maybe an order follower wouldn't even know that the state religion is CoE. So maybe CoE should at least not be treated as a (semi-)evil religion. I'll keep this up in the air for now.

I definitely think that opposite religions should cause instability even if neither of them is a state religion. Also, the religion clashes are additive, right? So Order + Veil + OO would mean +8 instability.

Yes, additive. I decided to add only an extra +1 if Order and Veil is present and both are non-state religions. Many religions in the city already give a relatively high penalty, and I want to make it simple enough to display concisely. Adding a penality for every clash probably won't work well, since you that would grow quadraticly with the number of religions.

Religion civic should certainly affect religion (in)stability.

I proposed only affecting positive effects, see the civic list. I stand by that, I think it's fine to have the Religion civic as a small extra bonus, and Theocracy as a full committment to religious purity.

Culture rate means culture per turn, right? I feel like in FfH it's quite often the case that you have either fairly low rates (monument, maybe theater) or pretty high ones (wonders), due to base culture being fairly rare and temples multiplying the output. I think this system as proposed might work, but I don't really have enough insight. Might be worth just testing.

Yes, that's culture per turn. I agree we should probably discuss this after testing.

I suppose that a fresh rebellion should still get a reduced impact from this, so they don't collapse themselves for the very same reason.

I added a general -90% for all rebels to the proposal, I think that makes more sense. I'll have to check how long the rebel status is generally retained though.

Just one question regarding crime: What's the calculation of that, aside from buildings and wonders?

There are some events that permanently change crime in a city. There's also global crime, which is added to every city can be changed by Undercouncil resolutions and maybe (?) events.

Removing the "adoption" stability hit makes sense in all but one case: I would definitely keep it for "Sacrifice the Weak". This civic is such a brutal change in values with an equally brutal effect on the population, that I could very much see a change to it as the cause for a few cities to try to secede. (Can either put it to 150 to get the old +200 in conjunction with the 50 from anarchy, or just leave it at 200 for a really major impact). Similarly perhaps about the +50 from slavery.

Agreed for sacrifice the weak. I'll but it at 50+150; I'm always worried about making "evil" civics to weak. I disagree on slavery, since that in FfH (and historically, I believe) is mostly about enslaving foreign people.

Smugglers' Ports are fairly useful, a good building limited to the Undercouncil. (Which also gives you -25% instability from crime with your changes.) I would keep the +1 instability on them.

Good point, I put it back to +1 instability.

An Altar of the Luonnatar of lower levels is not hard to get, and on the later ones the numbers escalate quite a bit. I think the effects for all cities are too much. I would greatly reduce these, or keep them local. Or both: Full effect as above locally, and a small effect everywhere.

Sure, maybe I got ahead of myself here . I left the final one very strong, that should be fine.

Bazaar of Mammon wouldn't change the cap, just make it slower to reach, right? Keep in mind it is a building that everyone will have at some point.

It does change the cap, since crime above 100 is just not counter. I don't think it is very strong. At 100 crime it's -2 instability, but that will happen very rarely. Let's maybe see how it goes in playtest.

Heron Throne seems too good, this is already a super powerful early wonder with its +8 culture and water hammers. I'd leave the extra stability effect out.

Sure.

Palace -2 in all cities represents a base stability?

Yes. But mabye it's better to remove it from palace and make it explicit. Rebels and puppets don't have a palace, but that edge case doesn't justify making the effect more obscure.

I would remove the Dwarven Vault effect. It already gives extra unhappiness/happiness, and other very strong benefits if rich. A wealthy Khazad empire would be extremely stable with this, while a poor Khazad would be punished even harder.

Agreed, the (un-)happiness is probably enough.

 

[DuskTreader]

The mod Legends of Revolution claims that its AI has been taught to understand Revolutions. I don't know how accurate that is or how it works.

 

[Artiss]

If not teaching AI for calculating, stacking and targeting different stability modifiers taking into account whole picture, I can suggest two paths:

1. AI (especially conqueror with power stack) often has big treasury, so in newly conquered cities it can be forced to buy stability buildings every turn. Especially archery range (defense bonus), courthouse and dungeon, elder council (relatively cheap) and market (economy bonus).
Positive effects for garrisoning cities can be doubled for AI and negative halved.
That path difficult to name hard forceful for AI, because it's typical behavior even for human players.
If not technically buying (for too expensive for example), it can be implemented through automatically appearing some stability buildings in newly conquered AI cities for solid part of treasury. So that spending money can slow down conquering pace in the end.

2. In newly conquered cities or pacified after rebellion for 5-10 turns (according game speed) appears "unyielding order". That path if nothing else just decrease frequency of outbreaks of endless world revolution in AI cities. In addition all negative modifiers for AI can be halved and positive doubled.

 

[lfgr]

I released MNAI 2.9.2u, and the "Balance/Improve stability" phase is now finished, apart from likely further balance changes.

 

[Tielby]

I'm on turn 117 of a (settler, quick speed, Erebus continent, huge, high cohesion, low sea level) game in which I tried a couple things I've never tried before, to hilarious effect. One of those things is Revolutions (experimental), which is why I'm posting it here. I wouldn't call these bugs. They may be something to be expected, if you have experience with Revolutions, or they may be unusual consequences of a combination of factors in my specific game.

So, I chose Clan of Embers, which I've rarely played before, and then, on a lark, chose to pop For the Horde on the first turn. I ended up with seven Scorpion Clan Archers. I was expecting a MUCH heavier unit maintenance penalty than I got, but I popped a goody hut for some money on the second turn, and then three or four more in the turns soon after, so unit maintenance was not going to be a problem for a long while.

I was able to get three of them together on another continent, and two more nearish my starting city. The three were able to kill one civ pretty easily, and the two were soon joined by my starting warrior to trounce my nearest neighbor, so I had three cities very early, PLUS the money from conquering those two. Using that money, I took one of my solo archers, went to a goblin fort and hired seven goblins, then went and sacked another civ. By that point I'd built Rantine, joined up with my two archers and went south to kill a fourth civ.

I've never, ever played this way before, mind you, going on a killing spree right off the bat, plus it's just not possible under normal circumstances. AND I was completely unaware of how Revolutions was going to make this a serious problem very quickly.

Since I did not know the mechanics of Revolutions, when one of my cities declared independence, I was VERY surprised when TEN warriors immediately appeared to attack "my" city. In short order I lost another city the same way, and could see that the loss of those cities was going to cascade. EEK!

And THEN I got the option to let the AI take over my civilization for 15 turns. WHAT?!?! Totally unexpected. But I didn't seem to have much choice... and 15 turns later (turn 117), the AI returns control to me, and, honestly, I don't think I could have handled it half as well as the AI has done!

I know I'm not out of the woods yet, what with cities so far apart, and too early in the game to be able to build a lot of buildings or use civics to keep my civilization together. I'll try to keep the cities closest to me, and let go of the distant ones without another cascade.

Anyway, hilarious!

 

[DuskTreader]

Revolutions really punishes early rush games, which is realistic (early empires like the Akkadian empire or Alexander the Great's empire were extremely unstable), but punishes the Doviello and Clan of Embers more than others.

 

[Alekseyev_]

And THEN I got the option to let the AI take over my civilization for 15 turns. WHAT?!?! Totally unexpected. But I didn't seem to have much choice... and 15 turns later (turn 117), the AI returns control to me, and, honestly, I don't think I could have handled it half as well as the AI has done!

Wait what, that's hilarious! Is that an event?

 

[Tielby]

Wait what, that's hilarious! Is that an event?

It seemed to be a normal (within Revolutions) progression of a civilization in the midst of multiple revolutions. But yeah, it was great fun!

 

[DuskTreader]

You can also induce it with Gibbon Goetia