Late Game Mechanics

[GloryGloryByzantium]

Discussion around late game / large empire mechanics has always had a role in this game as in other PDX games. The mechanics and marketing indicated that this was always going to be part of the show piece of this particular game. [insert sadness about game status]. I've always considered a focus on imperial decay since I started modding around Fall 2020. It was intended as part of a reboot of a Late Republic mod. Recent attention on this period by others has been found in the forums and reddit and a further conversation has been started in the context of Invictus [see here and messages following]. Invictus for now will continue to focus on flavor and some of the imperial decay changes are significant in their mechanical innovation or rebalancing ramifications.

So here's a dedicated thread to discussing imperial decay and decadence. What challenges to you want to see as a player once you've hit blob stage? What kind of collapses do you want to see occur to the AI? Should the focus be on character interactions, rogue military leaders, economic collapse which requires emergency powers and sweeping reform? How avoidable should it be? How threatened by a game over do you want to be before there's a transition from fun to frustrating? Maybe you purists out there want frustrating civil wars?

Let's hear your thoughts.

 

[isaacssv558]

I mostly want the ability for large empires to fragment or lose large chunks of territory to revolts. e.g. The crisis of the third century with two multi-region power blocs splitting off from the main empire.

The existing province revolt mechanic is good for early game, but late game they lead to piecemeal revolts which an empire can crush in detail. These revolts are boring and run up the tag count.

 

[DukeLeto42]

Sure, I'll join in with some thoughts:

Provincial Revolts:
As it stands, the main late-game roadblock is handling dozens of revolting provinces. None of them are a threat, but putting them down is just tedious because it can become a constant issue. It's better than it used to be (revolts not triggering below a threshold), but replaced "extremely rare revolts" with "constant small revolts." I'd love to see each province in revolt have a small negative impact on global provincial loyalty, which would drive more to go in a wave, followed by a fairly hefty loyalty bonus for putting down a revolt. This would make revolts more likely to be a real threat, but less likely to be just a constant source of micro and a crisis that can have a firm end.

Civil Wars:
Ironically, because more and more power base is split up, larger realms have more solutions to Civil Wars, instead of internal conflict becoming the main challenge. Still, it's not all bad. The main issue Civil Wars have is just how chaotic the "carpet siege and immediately flip" can be, especially when a small army gets into a "safe" region, takes a couple territories, and then spawns a second army from them (this also applies to the similar CBs). There's also the whole issue of your capitol getting plopped into random territories when captured, which should not really be how it works. This system feels like it needs another pass but isn't a terrible idea (realm split up, having to consolidate into opposing blobs, and then fighting it out).

Economic Collapse:
Drawing from Rome itself, one of the big challenges it faced was maintaining its resource supply. The current trade good system starts off in a nice place - local food supplies along with food and other trade goods surpluses and bonuses conferred by them - but the province-province trade system kind of ignores everything in the middle. A system using trade flows that could be disrupted would help you visualize the empire you're building, but also create weaknesses an empire or an internal crisis could exploit.

 

[GloryGloryByzantium]

I mostly want the ability for large empires to fragment or lose large chunks of territory to revolts. e.g. The crisis of the third century with two multi-region power blocs splitting off from the main empire.

The existing province revolt mechanic is good for early game, but late game they lead to piecemeal revolts which an empire can crush in detail. These revolts are boring and run up the tag count.

Provincial Revolts:
As it stands, the main late-game roadblock is handling dozens of revolting provinces. None of them are a threat, but putting them down is just tedious because it can become a constant issue. It's better than it used to be (revolts not triggering below a threshold), but replaced "extremely rare revolts" with "constant small revolts." I'd love to see each province in revolt have a small negative impact on global provincial loyalty, which would drive more to go in a wave, followed by a fairly hefty loyalty bonus for putting down a revolt. This would make revolts more likely to be a real threat, but less likely to be just a constant source of micro and a crisis that can have a firm end.

There's a couple of ways to address the revolts. You could make them have civil war mechanics, with obvious problems. Or, you can set the loyalty of neighboring areas with similar culture composition to drop dramatically / immediately, leading either to the regions being placated or else they join the revolt. Here's where it gets tricky, if they join as the same nation, then you can potentially have a blob revolt that's so large you cannot suppress it all in a single war (unless we further dropped the war cost for revolts). Or you have them ally the revolting nation in which case the whack a mole issue just becomes lots of moles at once.

Civil Wars:
Ironically, because more and more power base is split up, larger realms have more solutions to Civil Wars, instead of internal conflict becoming the main challenge. Still, it's not all bad. The main issue Civil Wars have is just how chaotic the "carpet siege and immediately flip" can be, especially when a small army gets into a "safe" region, takes a couple territories, and then spawns a second army from them (this also applies to the similar CBs). There's also the whole issue of your capitol getting plopped into random territories when captured, which should not really be how it works. This system feels like it needs another pass but isn't a terrible idea (realm split up, having to consolidate into opposing blobs, and then fighting it out).

Civil War mechanics can be modified. Agamidae made some code that allows entire areas to flip if forts and capital are taken.
There are some interesting flavor events for vanilla that allow for eastern and western Roman empire, but it's quite rare to see them.

Economic Collapse:
Drawing from Rome itself, one of the big challenges it faced was maintaining its resource supply. The current trade good system starts off in a nice place - local food supplies along with food and other trade goods surpluses and bonuses conferred by them - but the province-province trade system kind of ignores everything in the middle. A system using trade flows that could be disrupted would help you visualize the empire you're building, but also create weaknesses an empire or an internal crisis could exploit.

This would be great, but fixing trade without changes to vanilla code is quite handicapped sadly.

 

[DukeLeto42]

There's a couple of ways to address the revolts. You could make them have civil war mechanics, with obvious problems. Or, you can set the loyalty of neighboring areas with similar culture composition to drop dramatically / immediately, leading either to the regions being placated or else they join the revolt. Here's where it gets tricky, if they join as the same nation, then you can potentially have a blob revolt that's so large you cannot suppress it all in a single war (unless we further dropped the war cost for revolts). Or you have them ally the revolting nation in which case the whack a mole issue just becomes lots of moles at once.

I think the second option is probably the best, provided that war score costs are kept in check - I had at least one instance soon after release when a massive revolt nation was too much (this was before the revolt rework). I also think we need a significant loyalty boost post-revolt, as nothing's more tedious than beating a pitiful revolt and finding it's at -.5 loyalty and will just go again in a few years.

Civil War mechanics can be modified. Agamidae made some code that allows entire areas to flip if forts and capital are taken.

Ooh, very interesting.

This would be great, but fixing trade without changes to vanilla code is quite handicapped sadly.

Yes, I was definitely thinking more on the lines of "where could development improve the game."

 

[Herennius]

As posted in the Imperator Invictus mod thread, I see increased corruption as a possible way to simulate the difficulty of keeping a big empire together internally. Similar to how a bigger state rank increases income of office holders directly, for major and world power there could be additionally an increase in corruption by applying a monthly modifier on all office holders (e.g. 0.05 for major powers and 0.1 for world powers). It would create an interesting choise between coping with higher corruption levels, fighting them directly (you still have several tools to directly offset even a 0.1 monthly corruption modifier) and evading it by regularly appointing new office holders. This "corruption managment" is currently nearly absent in Imperator - you might occasionaly bribes or give free hands to make disloyal characters loyal again, but you never really have do to deal with the downsides of widespread corruption (as single characters being highly corrupt isn't a real problem - on contrary, you basically have those people in your hands, as trials are highly likely to suceed vs. them)

 

[GloryGloryByzantium]

As posted in the Imperator Invictus mod thread, I see increased corruption as a possible way to simulate the difficulty of keeping a big empire together internally. Similar to how a bigger state rank increases income of office holders directly, for major and world power there could be additionally an increase in corruption by applying a monthly modifier on all office holders (e.g. 0.05 for major powers and 0.1 for world powers). It would create an interesting choise between coping with higher corruption levels, fighting them directly (you still have several tools to directly offset even a 0.1 monthly corruption modifier) and evading it by regularly appointing new office holders. This "corruption managment" is currently nearly absent in Imperator - you might occasionaly bribes or give free hands to make disloyal characters loyal again, but you never really have do to deal with the downsides of widespread corruption (as single characters being highly corrupt isn't a real problem - on contrary, you basically have those people in your hands, as trials are highly likely to suceed vs. them)

Yes, I really like this idea. There's a couple of details to work out though. Does it need to be a monthly tick? As anyone who has given away too many free hands, those ticks can climb quickly. Also, does it need to be for everyone? I imagine some characters with certain traits would be immune (gotta have our Cato's too).
Using corruption as a level will also have other ramifications, such as more characters becoming populares. Perhaps this is a desirable outcome in the long run of the game. It makes me wonder if the devs originally had intended something like this? @Trin Tragula @Snow Crystal I also keep uncovering Roman events where a populist should naturally rise to power and the scenario described here would only make it more likely.

 

[Herennius]

Yes, I really like this idea. There's a couple of details to work out though. Does it need to be a monthly tick? As anyone who has given away too many free hands, those ticks can climb quickly. Also, does it need to be for everyone? I imagine some characters with certain traits would be immune (gotta have our Cato's too).
Using corruption as a level will also have other ramifications, such as more characters becoming populares. Perhaps this is a desirable outcome in the long run of the game. It makes me wonder if the devs originally had intended something like this? @Trin Tragula @Snow Crystal I also keep uncovering Roman events where a populist should naturally rise to power and the scenario described here would only make it more likely.

Thank you for taking the ball up and refining my idea

The reasons I suggested a monthly tick...well, personaly free of any modding skills, I drawed from existing modifiers and hoped to suggest something thats easy to implement. Also, a gradual monthly increase is more safe against players inevitable attempts to cheese around with things. The thrown out numbers of +0.05/0.1 were just examples of course (and I followed the development philosophy to first make a bold change, investigating the effect afterwards and then nerfing things again, if necessary).

Making it fir for everyone isn't necessarily mandatory. Restricting it on governors (or maybe Generals/Admirals, too) could be enough. Thinking about, it might no be necessary to penalize the Augur or a research office that way.

For traits...I had the tought as well, but wasn't sure, if such sub conditions are possible (and in case how most modder friendly). Unless missing something, there are three trait pairs affecting corruption (Honest -0.05 vs. Deceitful +0.05 ; Guileless -0.1 vs. Crafty and Righteous -10 vs. Corrupt +10) - while the the first two are pretty self explaining and a corrupt character will likely get just a one-time corruption infusion after getting the trait, I'm unsure how Righteous` -10 are applied (if characters start out with negative "corruption score", it isat least not visible UI-wise). I would be ok with not applying the extra penalty for the three positive ones (would make such characters extra desireable for big empires); further increasing it for the negative ones might sound "fair" then on the first glance, but that easily might be piling too much on them.

Faction shift towards Populares/Democrats is something I hadn't on the radar when making my suggestion, so that's a good point. My last republican world empire is probably some playthroughs old, so I can't jugde ATM, if the game could and should cope with even more faction leaning in that direction. In case this sideeffect would be undesireable - could it be counteracted by reducing corruption impact on faction leaning by some amount? Of course, not desireable, but restricted to the affected state ranks. And yes, now we have crossed already the line from "Why not make just that esay change X?" towards rebalancing things like you said in the original post of this thread

 

[CapnMunch]

So, beyond the lack of really meaty resistance to rule, I think the other big thing preventing any true collapse in I:R is that there don’t seem to be a lot of opportunities for “death spirals,” where systems interlock to actually punish failure in the long term.

Now, I don’t have a lot of experience in I:R, so maybe some of this exists already and I’ve just never seen it, but a few humble suggestions:

• Have Stability decrease monthly while fighting a civil war or undergoing a revolt, so that violent instability has more momentum, and even a quickly suppressed uprising leaves an impact.
• Make Tyranny decay scale with Religious Unity. Religious minorities, naturally, are distrustful of steps towards authoritarianism.
• Similarly, make Aggressive Expansion decay scale somehow with the number of Unintegrated Cultures. The more peoples brought under one yoke, the more suspicion outsiders will have.
• Prevent larger powers from benefiting from Revanchism. It’s explicitly a bounce-back mechanic and, sorry Rome, you don’t need help to bounce back when you already own half the world.

 

[WARenie]

Thing which I was thinking about for some time is an event which fires during civil war or any other big war for big empires, which invites all neighbors of an empire to attack it without taking stability hit. If you are big, you have many neighbors most likely and it will be not easy to defend all your land.

 

[isaacssv558]

There's a couple of ways to address the revolts. You could make them have civil war mechanics, with obvious problems. Or, you can set the loyalty of neighboring areas with similar culture composition to drop dramatically / immediately, leading either to the regions being placated or else they join the revolt. Here's where it gets tricky, if they join as the same nation, then you can potentially have a blob revolt that's so large you cannot suppress it all in a single war (unless we further dropped the war cost for revolts). Or you have them ally the revolting nation in which case the whack a mole issue just becomes lots of moles at once.

I think that having nearby provinces join poorly suppressed revolters seems good. Dropping the war cost seems like the best solution. A lowered war cost and bigger revolts might also give incentives to grant revolters autonomy by requesting vassalization instead of annexation. e.g. If all of Brittania revolts while Rome is fighting the Persians, they might accept vassalization rather than spend the time reconquering the entire isle.

Ideally, revolters would delay their revolt until they could expect a reasonable size relative to the parent country. I am tired of watching AI great and major powers crush tiny revolters. Maybe provinces which would revolt, but find it hopeless, can lower the happiness of neighboring provinces? Basically, simulate unrest and civil disobedience spreading until it breaks into open revolt. Otherwise, I worry that empires will simply continue to crush revolts one by one.

Civil War mechanics can be modified. Agamidae made some code that allows entire areas to flip if forts and capital are taken.
There are some interesting flavor events for vanilla that allow for eastern and western Roman empire, but it's quite rare to see them.

Flipping entire areas would be great. Carpet sieging is the worst part of civil wars. Would it also be possible to prevent undefended provinces from flipping without conquering the local fort? It is really annoying when armies retreat into my territory and flip some random rural province.

 

[Herennius]

Ideally, revolters would delay their revolt until they could expect a reasonable size relative to the parent country. I am tired of watching AI great and major powers crush tiny revolters.

Is your impression based on games with the AI loyalty boost being active (which gives them everywhere +30 loyalty, as soon as a single province revolts - which makes bigger revolts impossible)? Then I would recommend you to use a mod removing it - it may still not be perfect (as unruly provinces still don't intelligently coordinate when they strike), but it allows at least the existing join-neraby-revolt-of-same-culture-feature to fire occasionally and the AI struggles more with revolts then overall (how much depends on whether you play just with this mod - then especially Seleukids get often in troubles - or if you use Invictus with its improved AI...then even the computer copes quite good with bigger revolts)

 

[GloryGloryByzantium]

Restricting it on governors (or maybe Generals/Admirals, too) could be enough. Thinking about, it might no be necessary to penalize the Augur or a research office that way.

But what if the augur's wife likes to eat oysters for breakfast?

if the game could and should cope with even more faction leaning in that direction. In case this sideeffect would be undesireable - could it be counteracted by reducing corruption impact on faction leaning by some amount?

I'm ok with long-term transitions to populism. In the old version of the game, you had to have high populists counts plus high corruption to enact princeps civitates.

So, beyond the lack of really meaty resistance to rule, I think the other big thing preventing any true collapse in I:R is that there don’t seem to be a lot of opportunities for “death spirals,” where systems interlock to actually punish failure in the long term.

Now, I don’t have a lot of experience in I:R, so maybe some of this exists already and I’ve just never seen it, but a few humble suggestions:

• Have Stability decrease monthly while fighting a civil war or undergoing a revolt, so that violent instability has more momentum, and even a quickly suppressed uprising leaves an impact.
• Make Tyranny decay scale with Religious Unity. Religious minorities, naturally, are distrustful of steps towards authoritarianism.
• Similarly, make Aggressive Expansion decay scale somehow with the number of Unintegrated Cultures. The more peoples brought under one yoke, the more suspicion outsiders will have.
• Prevent larger powers from benefiting from Revanchism. It’s explicitly a bounce-back mechanic and, sorry Rome, you don’t need help to bounce back when you already own half the world.

Death spirals good. Although the stability decrease during civil war might be harsh, but realistic. The question is whether players will enjoy the challenge or hate it when they get game over'ed from these sequences?
Note sure how easy it is to calculate the AE on pop integrated cultures. I think it could be done, but I'm seeing some lag as a result.
Yeah, how to turn off Roman antagonism and still get them to appropriately expand?

Thing which I was thinking about for some time is an event which fires during civil war or any other big war for big empires, which invites all neighbors of an empire to attack it without taking stability hit. If you are big, you have many neighbors most likely and it will be not easy to defend all your land.

Good idea. But would have to be done in an event rather than a natural progression. How do you weight the AI's decision for them to decide to join or refrain?

I think that having nearby provinces join poorly suppressed revolters seems good. Dropping the war cost seems like the best solution. A lowered war cost and bigger revolts might also give incentives to grant revolters autonomy by requesting vassalization instead of annexation. e.g. If all of Brittania revolts while Rome is fighting the Persians, they might accept vassalization rather than spend the time reconquering the entire isle.

Not a bad idea. But what's the limit you set? How large does the revolt have to be before the offer is sent?

 

[isaacssv558]

Is your impression based on games with the AI loyalty boost being active (which gives them everywhere +30 loyalty, as soon as a single province revolts - which makes bigger revolts impossible)? Then I would recommend you to use a mod removing it - it may still not be perfect (as unruly provinces still don't intelligently coordinate when they strike), but it allows at least the existing join-neraby-revolt-of-same-culture-feature to fire occasionally and the AI struggles more with revolts then overall (how much depends on whether you play just with this mod - then especially Seleukids get often in troubles - or if you use Invictus with its improved AI...then even the computer copes quite good with bigger revolts)

Probably, I haven't changed anything related to the AI beyond installing Invictus.

Not a bad idea. But what's the limit you set? How large does the revolt have to be before the offer is sent?

I figure it can use the existing war resolution mechanics. i.e. a player or AI who is winning a revolt, but only moderately, might choose to select the vassalization option in the peace deal rather than suppress the revolt entirely. A revolt which is winning should press on in hopes of achieving full independence.

An event might be better for AE and AI reasons, but I think it should still be based on warscore or the revolters' expected chance of success. It should be able to happen for any size revolt which the overlord has at +10 warscore for and is likely to win. The overlord should accept only if it cannot afford the cost of a full suppression due to a current or imminent crisis. e.g. If Brittania revolts from Rome during a major civil war, it should make the offer after Rome wins a few battles due to balance of power. Rome should accept the offer to free up legions for the civil war. This way, civil wars or external pressure combined with revolts can have a compounding effect.

 

[Gurkhal]

This is a small input into the part of the game about scaling costs. And that is that perhaps a new building of "mint" could also be introduced in order to allow for temporary reducing the scaling of costs in some way? But that if you build to many mints you will start to increase scaling?

 

[GloryGloryByzantium]

Should this be the bonus / malus of going from major power to great power?

 

[WARenie]

Should this be the bonus / malus of going from major power to great power?
View attachment 831166

I would remove civil war threshold modifier - lowering it makes civil wars less dangerous.

 

[Jiben]

Civilized countries should have a really hard time conquering nomadic peoples. The steppe should be next to impossible for the bosporan kingdom to hold (atleast directly) without huge investments.

Give tribes the ability to raid. If i play as rome i should have to worry about raids across the Rhine. If i play rome and use 100% of all possible troops in a war against Persia the tribes should know that local resistance is gone and raid the frontier.

Give ambitious generals or govenours the ability to start a small scale civil war on their own that could ballon if it can't be dealt with fast.

 

[GloryGloryByzantium]

If Brittania revolts from Rome during a major civil war, it should make the offer after Rome wins a few battles due to balance of power. Rome should accept the offer to free up legions for the civil war.

Sounds great, but this kind of dynamic event scripting in unfortunately tedious and hard to control. You can scope to a war and check warscore, but the specifics of warscore cost and which lands to give out are hard to generalize.

This is a small input into the part of the game about scaling costs. And that is that perhaps a new building of "mint" could also be introduced in order to allow for temporary reducing the scaling of costs in some way? But that if you build to many mints you will start to increase scaling?

Better yet, I wish we had an inflation system and a type of treasury mechanic different from the amount of gold in the taskbar. Whenever I see expendable gold up top, I start building. If I sunk it into a temple instead, I would do a better job storing it.

I would remove civil war threshold modifier - lowering it makes civil wars less dangerous.

I think the lowering the threshold makes it easier for civil wars to fire, so if anything we can make the value more negative - although as a great power, it can fire with basically only 15% of your nation disloyal. So the magnitude is balanced. What I am considering if removing the countdown timer to war. Make it like CK factions - the leader can trigger it at any time if they feel their faction is strong enough. Or if you have someone with only 10% powerbase but is really disloyal, they can trigger it along with themselves and friends and family members?

Civilized countries should have a really hard time conquering nomadic peoples.

Yes. Although in history, I think the difficulty has to do with terrain issues of supply limits and attribution but most especially due to ambushes and large numbers. The game seems to have no problem stomping outnumber armies at lower tech. Sometimes in history this happened, but not always.

Give tribes the ability to raid.

Would be nice. Can't really make migratory units do anything from a modding standpoint. Could just buff barbarians and make them more likely to trigger during war.

Give ambitious generals or govenours the ability to start a small scale civil war on their own that could ballon if it can't be dealt with fast.

This exists in vanilla event sequences. The scripts for it look well thought-out. It just doesn't fire all the time or is easy to avoid. I think I had this happen to me quite recently when Carthage became a great power. See these screenshots:

Spoiler: Ambitious General Sequence

In all it was quite fun, and even felt dangerous at times, but in the end I was able to avoid it rather easily due to it triggering with a suffete who had good attributes and a loaded treasury.

In other news, I noticed that revolts do, in fact, make nearby pops of nearby regions (or similar culture groups) less happy, it's just that the modifier is a mere -5%:

This could be increased. Due also to an event about an influential senator advocating for minority rites plus war exhaustion, I ended up with about 5-6 simultaneous revolts in Hispania. It still didn't feel that daunting because I was rich and could run around with 4 mercenary armies to put them down one by one. Re: the discussion above about cascading revolts, it may be best to try to find a way to join them into one nation so they are coherently fighting. But it might be hard to code given the potential for areas in similar superregions to revolt but might not be neighboring. Maybe if they're in the same region? I also might add that it was probably easy to defeat the revolts because their civ levels are low.

 

[WARenie]

I think the lowering the threshold makes it easier for civil wars to fire, so if anything we can make the value more negative - although as a great power, it can fire with basically only 15% of your nation disloyal. So the magnitude is balanced. What I am considering if removing the countdown timer to war. Make it like CK factions - the leader can trigger it at any time if they feel their faction is strong enough. Or if you have someone with only 10% powerbase but is really disloyal, they can trigger it along with themselves and friends and family members?

The problem is that if you have civil war on 15% of nation being disloyal, you get only 15% strong revolt which will be destroyed immediately. 50% civil war is much stronger and more dangerous than 15% civil war. Real countdown to war is that won civil war increases loyalty of everyone and you can get situation of "state constantly fights in very small civil wars and it makes everyone very loyal. New disloyal people immediately start preparing another suicidal civil war and get eliminated.". I would say that "50%" is a threshold of "faction is strong enough to start a war". Maybe it should be decreased during long wars, from high war exhaustion, when state is weaker, as revolts of smaller percentage can have chances when state is at war already. And increased during peace, when state can deal with it easier, so revolters need to gather more forces to have a chance to win.