Revolution: what's your experience?

[Vokarya]

We would need DLL work to get the AI to keep bigger garrison forces with Revolutions on. The last few games I played on Noble I found that getting a religion meant the difference between early revolts and a well-knit civilization.

I would like to know exactly what elements of Revolutions are giving the AI's trouble. I feel like that is something we can work on once I get all the rest of my plans implemented.

 

[Samson]

I really love this mod, still coming back to it over and over. However I still never finish games in it because it is just too easy. I have noted a couple of people mentioning the AI can't handle revolutions, so I am going to try again now with that disabled. But normally I play with variable difficulty and I end up at deity, the AIs are all on settler and still I can easily outscore and conquer them by 500AD or so. Can anybody advise on what settings to get the AI to actually be aggressive in expansion? I have tried ruthless, I have tried setting only the conquest and domination victories, I have made sure to have stack attack off (iirc AI can't deal with it), but still none of them really seem to get rolling through their neighbours. I'm sure at some point in this mods history they could do that and you would get a properly dominant AI in most games and it would be a challenge for the player even at normal difficulties.

The setting that makes the most difference to me in having the AI competitive is the AI flexible difficulty OFF. Having this on makes it much harder for the bigger AI's, which are the ones you want to be successful to have some competition.

 

[PsychopathicWarmonger]

May I ask, how often do you all experience empire-wide revolutions?
I can't ever recall experiencing one myself and maybe only once or twice for the AI in all the time I've played this mod.
I feel as if there's strange differences in how many of us are experiencing revolutions, despite playing on similar difficulties.
I myself am playing on Monarch with Rev difficulty at 5.0.
Perhaps I'm just quite good at managing revolutions (in that case I should bump up to Emperor) but I'm currently playing a game where I haven't experienced rebellions at all, only having 1 city go into revolt once.

 

[Vokarya]

Does anyone think there is a particular time when revolutions are particularly bad for the AI? There is a buried trigger regarding democracy levels on civics that increases the negative effects of certain civics and it kicks in much earlier in AND; mid-Classical (Politics tech) vs. late-Renaissance (Constitution tech) in BTS. AI's should be running Monarchy in the medieval/Renaissance eras unless they are small, and I wonder if that is having too many effects on the AI.

 

[Omirta]

Early game revolutions don't seem to happen in my games anymore. I noticed that when the revolution meter reaches 600, it doesn't increase anymore unless the city is actually unhappy.

 

[45°38'N-13°47'E]

Early game revolutions don't seem to happen in my games anymore. I noticed that when the revolution meter reaches 600, it doesn't increase anymore unless the city is actually unhappy.

What if you increase the human revolution offset to 5 or 10?

 

[hoon9732]

The current version of revolution is giving me a decent experience. I've played ROM: AND back in rev 700 and in those times the revolutions easily got out of control. Maybe revolutions are supposed to be abrupt and unexpected, but that simply makes the game annoying :-(

But I sometimes wonder if there are hidden factors that affect revolution. The factors shown in the city screen(garrison, civic choices, happiness, etc.) alone doesn't seem to explain abrupt fluctuations in revolution.

 

[Samson]

But I sometimes wonder if there are hidden factors that affect revolution. The factors shown in the city screen(garrison, civic choices, happiness, etc.) alone doesn't seem to explain abrupt fluctuations in revolution.

Have you looked at the Revolution Advisor (the Che Guevara symbol in the top right corner)? Particularly useful is the Revolution Cheats page. I do not know if it is considered cheating, but the numbers there make revolutions predictable.

 

[hoon9732]

Have you looked at the Revolution Advisor (the Che Guevara symbol in the top right corner)? Particularly useful is the Revolution Cheats page. I do not know if it is considered cheating, but the numbers there make revolutions predictable.

Yeah I just checked that. It seems that the number of units in the city affects the rev values! I thought the presence is the only factor. What a surprise. Now placing doomstacks in cities likely to revolt would be a great help.

 

[Vokarya]

Yeah I just checked that. It seems that the number of units in the city affects the rev values! I thought the presence is the only factor. What a surprise. Now placing doomstacks in cities likely to revolt would be a great help.

Would you believe it's actually an exponentially decreasing effect? And that Walls and Castles increase the effectiveness of garrison forces? There are a lot of effects that are hidden away in the Revolutions code and very badly documented if at all.

 

[Vokarya]

Here are some things I am currently thinking about Revolutions. I'm playing a test game on Noble difficulty, Epic speed, Standard size map, human offset of 0, and some of the Era numbers tweaked to make things a little less frustrating in the early eras. I'm in the early Renaissance and I have destroyed a couple rival civilizations.

• I don't think that Secular/Atheist/Personality Cult need to give an inherent stability bonus, although I'm mixed about this. Mostly I don't think religious civics should be playing in the realm of general stability. They have +/- to state religion stability/non-state religion instability as well as holy city control modifiers.
  
• Revolutions seem to spawn an Explorer unit along with the other units that appear. Does this unit really serve a purpose? I play with Require Complete Kills and all it turns into is an additional unit that you have to hunt down and kill. It can't make a difference either on offense or defense.
  
• This is the big one. I really think revolts need to be more spaced out. It is very frustrating to have revolutions repeating over and over with very few ways to deal with them once you have spread your religion and put in a garrison; I have held some cities for centuries of game time and they still revolt. I am playing as an Imperialistic leader, but I have so far spawned 9 Great Generals before 1600 and it is almost all due to revolutions. The fix that I would like to try is doubling the revolution index drop for putting down a revolt. Right now, the calculation is supposed to be this:
  • A starting fraction of 0.2 if the revolution was put down within 30 turns, 0.3 if the revolution was put down before any reinforcements appeared.
  • An additional +0.1 if the city's current local index change is less than 0 (meaning conditions are improving). This is almost never going to apply because of the increase to the distance factor, which is included in the local.
  • An additional +0.1 if the city's revolution index is above the Always Violent threshold. I think this is currently 1750.
  • A reduction of -0.05 if the city has revolted three or more times, -0.08 if the revolution was put down for reinforcements.
  • The final fraction is used to reduce the city's revolution index, then drop it to the Always Violent threshold.
    
  • In practice, I'm seeing about a 20%-30% drop in revolution indexes in the revolting cities after putting down a revolt. That's not nearly enough.
  • Increasing the bonus for putting down a revolution won't defuse any of the factors that lead to a revolution. I think for the level that I am playing at, there needs to be a bit more breathing room, and this will help without completely defanging revolts.

 

[<Nexus>]

• Revolutions seem to spawn an Explorer unit along with the other units that appear. Does this unit really serve a purpose? I play with Require Complete Kills and all it turns into is an additional unit that you have to hunt down and kill. It can't make a difference either on offense or defense.

A Spy would be more appropriate.

 

[hoon9732]

This is the big one. I really think revolts need to be more spaced out. It is very frustrating to have revolutions repeating over and over with very few ways to deal with them once you have spread your religion and put in a garrison; I have held some cities for centuries of game time and they still revolt.

Indeed. Personally, I thought the "Rebellious Spirit" modifier overly encourages cities to have revolts regularly. The modifier itself seems to fit the history, and makes 'revolting regions' to be predictable. However, the modifier seems to get reinforced every time that in the end the rebellious regions hardly stay pacified.

I have no idea if national/local rebelliousness works separately, but maybe breaking down a local rebellion should add up temporarily to the national stability. Maybe that could help? Several times of my game play showed my cities rebelling in a cycle. If A region rebels, then B region would surely follow next for some unknown reasons.

 

[<Nexus>]

I have no idea if national/local rebelliousness works separately, but maybe breaking down a local rebellion should add up temporarily to the national stability. Maybe that could help? Several times of my game play showed my cities rebelling in a cycle. If A region rebels, then B region would surely follow next for some unknown reasons.

IIRC "national" does not stand for "nation wide" but for "nationality" that being misleading. At least I think that's what Afforess said once.

 

[Vokarya]

A Spy would be more appropriate.

Revolutions spawns Spies as well. I'm not so concerned about that. We have a fix in the DLL that Require Complete Kills does not require you to finish off any UNITAI_SPY or UNITAI_MERCHANT units. Otherwise a civilization that had an unused Great Spy would live forever. (Great Spy is UNITAI_MERCHANT, probably to get it to use its special Infiltrate ability rather than attempt spy missions that it can't do.)

It's the recon units and also the Great General spawns that are concerns. The Great General seems to spawn on a different square so it's a sitting duck for any unit.

 

[Vokarya]

IIRC "national" does not stand for "nation wide" but for "nationality" that being misleading. At least I think that's what Afforess said once.

National stability is an incredibly hard thing to figure out. I've tried to break it down and I'm getting confused. There is a calculated value based on the cultural % slider that doesn't seem to get used for anything. I may go back and look at it again.

 

[Vokarya]

Here's something else I noticed. Monarchy is the only government civic with an increased revolutionary sentiment. I can understand the representative governments having decreased revolutionary sentiment, but I think Monarchy should be a zero and Despotism have the increased sentiment. I think this would help Monarchy be more viable among the AI. On my last playthrough, I saw every AI running Republic.

 

[IPEX-731BA5DD06]

Ahh that explains a bit about my current game I'm writing up. I wondered why they seem to be revolting so much... Still writing this episode, have been for a month.

It will be done.. (Got to promote the game)

P.S. I'm role playing, and I may end up losing my empire as a result.

 

[<Nexus>]

Here's something else I noticed. Monarchy is the only government civic with an increased revolutionary sentiment. I can understand the representative governments having decreased revolutionary sentiment, but I think Monarchy should be a zero and Despotism have the increased sentiment. I think this would help Monarchy be more viable among the AI. On my last playthrough, I saw every AI running Republic.

Are revolutions affected by whats going on in the world? I'm thinking of the others-have-slavery happiness tag.
I think Monarchies and Dictatorships could more stable until they meet civs that are democratic.
I could imagine a tag that checks for the number of contacted civs with <bCanDoElection>1</bCanDoElection> and adds -1 stability penalty for each civ. The penalty would be caped by the number in the tag.

Do we have something like this already?

 

[Vokarya]

Are revolutions affected by whats going on in the world? I'm thinking of the others-have-slavery happiness tag.
I think Monarchies and Dictatorships could more stable until they meet civs that are democratic.
I could imagine a tag that checks for the number of contacted civs with <bCanDoElection>1</bCanDoElection> and adds -1 stability penalty for each civ. The penalty would be caped by the number in the tag.

Do we have something like this already?

I couldn't find anything like that in the revolution code. I don't think we really want to add any more elements to the revolution calculations at this time. There are so many elements in play already that I would like to find ways to minimize some of them because of all the cumulative addition.