Towards a fun late game dynamic

[heraklonas]

1. Large revolts. The last thing that EU4 needs is another mechanic that bogs down progress in the late game by pure chance. What we need is not an event that randomly punishes a player for holding a culturally diverse empire. I have another suggestion to the change of rebel mechanics here.
One of the things that make EU4 so much better than EU3 is to turn many originally random mechanics (such as religious conversion and diplo-annexation) into a streamlined process. (You cannot imagine how frustrating it is to try 20+ times to diplo-annex a vassal with an "Unlikely" chance, if you haven't tried it) I think rebels can change in a similar manner. Positive revolt risk leads to the accumulation of rebel strength, and negative revolt risk reduces it. When rebel strength accumulates to a certain extent, then a large scale rebellion breaks loose all at once. This will make rebellion actually challenging without resorting to events with 10-20 years of MTTH.

2. Targeted coalitions. Instead of making things too convoluted, I think the issue can be solved by not allowing countries to join a coalition war on the side of an unknown country. Once the native Americans make contact with Europeans, it is fair game to have them join forces against a common threat. (Sorry for listing a RL example, but I think Seven Years' War should be observed as an example of such kinds of co-operation. This example is to show that long-distance coalition is not completely unacceptable as a game mechanic)

3. No capital core decay. France is the perfect example to show that why this mechanic is potentially troublesome, as someone has already mentioned. I think a better solution is to make cores reappear or reset when a nationalist or patriot rebel uprising occurs in a province, instead of after a successful siege. This is to make cores re-emerge even the great powers can stomp the rebels before they can finish a siege. This will have the positive side effect of making espionage more useful, as the idea group makes creating new cores in foreign territory easier.

really great ideas - especially the first one.

 

[unmerged(80402)]

I agree, that's a no no. Different cultures maybe, but same culture groups nations should disappear forever

Primary tags for each culture should never dissapear. IE ukraine, belarus etc. If there is no primary tag(example french minors), then they can still just dissapear. But not poland, bohemia, hungary, or ukraine.

 

[neaiskink]

Hello,

great thread idea!
1. Warscore calculation could be changed to a mechanic resembling Victoria:
my post here http://forum.paradoxplaza.com/forum...-surrender&p=16266584&highlight=#post16266584

I believe additional wargoal mechanic - where being opportunistic would be very costly in terms of relations and stability - would make mid-game more interesting (i.e. eating smaller isolated enemies should be harder).
2. I think linear function of AE decrease is BOGUS (as in "they made separate peace 30 years ago"). It should be exponential with slow AE decrease just after an action and faster decrease later. This would solve the "permanent coalition" problem, where even if you do nothing from 1650 to 1750 you still have that giant coalition against you.
3. Coalition mechanics need tweaking:
a) coalitions should have geographic limits (as in trade or colonization)
b) initiating offence "coalition CB" should be available to the main 2 or 3 members only.
c) it should be possible to AVOID joining the coalition war for a large (for example 6) stability and relations hit
d) the entry/exit from coalition should be free, without the 5 y restriction
e) one should be able to force-dismember a coalition

I believe these changes would force the player to focus on wider geographical expansion (figting manchu would not mean fighting france), would stop the idiotic coalition offensives with only minors joining in (you beat them bloody, they pay, go home, try again in 5 years?), would allow for flexibility in case another coalition war is not an option.

 

[Bud_Backer]

I would like to respond to the OP suggestions here:

1. Large revolts. The last thing that EU4 needs is another mechanic that bogs down progress in the late game by pure chance. What we need is not an event that randomly punishes a player for holding a culturally diverse empire. I have another suggestion to the change of rebel mechanics here.
One of the things that make EU4 so much better than EU3 is to turn many originally random mechanics (such as religious conversion and diplo-annexation) into a streamlined process. (You cannot imagine how frustrating it is to try 20+ times to diplo-annex a vassal with an "Unlikely" chance, if you haven't tried it) I think rebels can change in a similar manner. Positive revolt risk leads to the accumulation of rebel strength, and negative revolt risk reduces it. When rebel strength accumulates to a certain extent, then a large scale rebellion breaks loose all at once. This will make rebellion actually challenging without resorting to events with 10-20 years of MTTH.

2. Targeted coalitions. Instead of making things too convoluted, I think the issue can be solved by not allowing countries to join a coalition war on the side of an unknown country. Once the native Americans make contact with Europeans, it is fair game to have them join forces against a common threat. (Sorry for listing a RL example, but I think Seven Years' War should be observed as an example of such kinds of co-operation. This example is to show that long-distance coalition is not completely unacceptable as a game mechanic)

3. No capital core decay. France is the perfect example to show that why this mechanic is potentially troublesome, as someone has already mentioned. I think a better solution is to make cores reappear or reset when a nationalist or patriot rebel uprising occurs in a province, instead of after a successful siege. This is to make cores re-emerge even the great powers can stomp the rebels before they can finish a siege. This will have the positive side effect of making espionage more useful, as the idea group makes creating new cores in foreign territory easier.

I did not think comprehensively enough to give a detailed argument on the whole issue, but I can say that I have some different views on what constitute the offending mechanics in EU4. I whole-heartedly agree with the warscore part, but I have reservations to the vision on coalitions, core decay and rebels. I will elaborate if I have time.

I particularly like the first suggestion offered, for the same reasons as the poster stated. I don't like the idea of simply random events being a way of increasing the challenge. Something where my actions will significantly contribute to the challenge seems more effective. The question then becomes what actions, and what do they trigger? Balancing that is what the designers can work on but the idea seems very sound to me.

With respect to culture conversion, as a mechanic I found it to be fairly limited in utility to me and yet, it appears the consequences for late game are profound - disabling the breakup of larger nations in war, etc. Would it not be better to eliminate culture conversion entirely and replace it with a more limited rebellion chance reduction (much like harsh measures, but permanent) as well as a tax or production boost? This would keep the culture, and all it entails mechanically, intact, but give a reason use the mechanic on troubled or low income provinces?

 

[unmerged(798670)]

Why is removing cores still considered the goto/only option if the problem is that people who have been french for 200 years want to be french?

Reunification rebels + rebel sentiment on release would make it a massive pain to stop, throw in a higher tendency to offer to be a vassal of france and/or an event to skip vassalization and start annexing and you have a method of breaking blobs that will be more difficult than normal but still a viable option with enough resource commitment.

Honestly I'd like to see more events where dual-cores want to switch from one core to the other periodically.

 

[Bud_Backer]

I'm not sure how this could be implemented from a mechanics point of view, but I might find the late game more challenging if say all the below were to be true:

-There was no mechanic to convert culture,
-There was a mechanic that made secession a real threat dependent on the number of cultures, particularly non-accepted cultures.
-The threat of secession was reduces by government investment in the provinces in question. Ie: buildings.

So to put this in a sort of game-setting example, if I as Castile occupy the Iberian peninsula and unite it all, I have several different cultures. It might be interesting (maybe not, would it be tedious?) if I had to worry Grenada might develop sufficient dissatisfaction that it might declare independence. Now this I mean would be not the same as a mere rebellion. Rather, a country is formed, We are at war, and that country can ally with other countries for support.

Suddenly my nice empire has some hotbeds of discontent that I cannot simply resolve with an army on rebel auto-suppression...

Mechanically, to reduce that discontent, I need to show (in game terms) that these people matter to me. That means construction of infrastructure (buildings). The more I build the reduced the risk. So I can play my expansion game, but have to balance my budget for the military against the need to keep a populace content.

 

[Bud_Backer]

Honestly I'd like to see more events where dual-cores want to switch from one core to the other periodically.

I like this idea as well. With presumably some actions on the player's part to attempt to mitigate it, I presume?

 

[QuorumOf4]

Offending mechanics (non-exhaustive)

• Empires are too stable: Whether under player or AI control, rebel mechanics do not pose a serious threat to anyone, except maybe Russia and Timurids. Large, heterogeneous empires rarely see revolts in many provinces, nor do revolts tend to spread. This means that large countries tend to exert gravity and gradually pull in smaller countries, growing perpetually throughout the game with few setbacks. This goes triple for player empires.

I feel the problem is at no point does an empire begin to have to deal with Corruption, There isn't a point where people far removed from the capital of the empire start skimming off the top, making continual expansion the best method for growth. There is no point of diminishing returns in empire size in terms of revolts or income. I think it needs both, Stability of an empire should decline at the fringes of the empire in foreign cultures where income lost from "corruption" directly feeds rebels in terms of nationalists or even separatists and pretenders.

The Beauty of Crusader kings two is that the larger your empire became the more unstable and internal conflicts you have to deal with. Stability would become a much more valuable stat to empires when it lowers the chances of fringe rebellions. EU4 Definitively loses something by not having to appoint local governors or local nobility. It seems there is a relatively great expansion hiding in there where different government types can support different types of ruler ship and varying numbers of advisers throughout the empire.


I'd also like to add to your list that alliances are FAR too stable, It is way to easy to be allied to two different rivals... who are in turn allied to two different rivals, creating a network of alliances that nobody wants to go to war in. The game desperately needs an event that fires when allied to two rivals that forces you to chose one and risk insulting the other.

 

[zodium]

Right, thank you very much to everyone for your reasoned responses. I'm extremely happy with the direction the thread has taken. I want to summarize the thread so far, and to do so, I will first try to put the rather diverse suggestions into a coherent framework. Broadly, I see suggestions that fall into three categories:

• Exchanges of statistical equations for dynamical equations. Bud_Backer said it best here: "I don't like the idea of simply random events being a way of increasing the challenge. Something where my actions will significantly contribute to the challenge seems more effective." Quite right: something that is random is, by definition, outside of behavioral control to the degree that it is random, and something that is outside of behavioral control can, by definition, never be said to be difficult. I see bleakie's response to my proposal for random culture-group wide rebellions as prototypical of this category.
• Exchanges of hard limits and dichotomies for trade-offs. There are many hard limits and dichotomies in the game, both transparent ones like OE and opaque ones like the province count modifier on AE, or a grey area like the difference between 100% and 101% OE. I see QuoromOfF4's proposal to implement a "corruption" mechanic for expansive empires as prototypical of this category.
• Parameter rebalancing. neaisink's suggestion to change the current linear distribution for decay over time with an exponential one, featuring slow initial decay and speedy late decay, seems prototypical of this.

While Parameter rebalancings will have to be discussed on an exemplar or case-by-case basis, the first two, I think, are best dealt with as design principles rather than via exemplars. It can be said that in general, an EU4 mechanic that depends on an underlying dynamic equation tends to affect game dynamics by rewarding strategic risk management (optimal play: taking on the correct amount of risk), while mechanics that depend on stochastic equations tend to produce challenge by punishing lack of risk minimization (optimal play: taking on the least amount of risk, or alternatively, the optimal amount of risk is always in the low regions). Are there any problems with simply concluding "Dynamical equations produce superior game dynamics except where computationally unfeasible" and "Continuous trade-offs are preferable to hard limits and dichotomies"? My gut tells me the latter is sound, but the former may need stronger qualification.

I'll give some time for comment, and if people find these to be agreeable categories, I'll summarize the thread according to this framework. I'm simply aiming to capture "most" suggestions here, and cover the remainder in an "Other" category.

 

[bleakie]

• Exchanges of statistical equations for dynamical equations. Bud_Backer said it best here: "I don't like the idea of simply random events being a way of increasing the challenge. Something where my actions will significantly contribute to the challenge seems more effective." Quite right: something that is random is, by definition, outside of behavioral control to the degree that it is random, and something that is outside of behavioral control can, by definition, never be said to be difficult. I see bleakie's response to my proposal for random culture-group wide rebellions as prototypical of this category.

• 
  
  I have to dispute this narrative, because it is very different from what I have in mind.
  My proposal is not a purely dynamical one; the trigger point of the rebellion is random (within a reasonable range of course). With this single difference, the deterministic/stochastic nature of the proposal is completely reversed. The purpose of my proposal is to discretize the revolts into larger blocks, with a single variable of randomness wielding significant power. Meanwhile, the rebel model of the current game consists of many random chances, but things get even out as empire size grows due to the Central Limit Theorem. While individual revolts are not predictable, the overall behaviour of rebels becomes too reliable to the degree of boring in the late game. Tying up the revolts in blocks can solve this problem by reducing the number of repetitive units, hence delaying the effect of Central Limit Theorem on rebel behaviour. Discretization of rebels is a better description of the nature of my proposal.
  
  EDIT: I have to admit that whether more randomness can be generated depends on the extent of variability of the trigger point. But one have to note that the current rebel model is statistically deterministic for a large empire, so it is very likely to create more variability if the discretized model is used.

 

[zodium]

I have to dispute this narrative, because it is very different from what I have in mind.
My proposal is not a purely dynamical one; the trigger point of the rebellion is random (within a reasonable range of course). With this single difference, the deterministic/stochastic nature of the proposal is completely reversed. The purpose of my proposal is to discretize the revolts into larger blocks, with a single variable of randomness wielding significant power. Meanwhile, the rebel model of the current game consists of many random chances, but things get even out as empire size grows due to the Central Limit Theorem. While individual revolts are not predictable, the overall behaviour of rebels becomes too reliable to the degree of boring in the late game. Tying up the revolts in blocks can solve this problem by reducing the number of repetitive units, hence delaying the effect of Central Limit Theorem on rebel behaviour. Discretization of rebels is a better description of the nature of my proposal.

EDIT: I have to admit that whether more randomness can be generated depends on the extent of variability of the trigger point. But one have to note that the current rebel model is statistically deterministic for a large empire, so it is very likely to create more variability if the discretized model is used.

Sorry, it seems that I misunderstood your post. I interpreted your proposal more along the lines of having the rebel percentage cause that proportion of the province's manpower and income to go into a pool for some set of provinces defined by cultural or religious affiliation, with rebellion triggering once it reaches some critical mass, which I thought was a far superior proposal.

Now, however, it seems the suggestion is essentially identical to my original proposal. This was that instead of rebels in a large empire popping up on a province-by-province basis, they would start to form coherent large-scale rebellions encompassing cultures/religions and their groups (analogous to a mix of nationalist and pretender rebel mechanics, in terms of current mechanics), in order to provide more of a challenge for a large empire. Could you explain the difference between this solution and your solution a little more?

 

[mcmanusaur]

-There was a mechanic that made secession a real threat dependent on the number of cultures, particularly non-accepted cultures.
-The threat of secession was reduces by government investment in the provinces in question. Ie: buildings.

I would just add that the implementation of this should be less randomized than the current rebels. In the current system, getting rebels is arguably more of a case of t jusbeing unlucky than any real statement about the cohesiveness of your empire.

 

[mcmanusaur]

I think your summary is a fair one, zodium, but I would simply remark that certain elements that you identify can really be generalized outside the late game (even if the examples that prompted the identification of such were from the late game). To over-simplify it for the sake of making a point, there's too much reliance on randomness for difficulty in the late game, but are random events really a strong point in any time period? On the topic of randomness, I think pretty much every random event (not the DHE's but the recycled ones) should at least involve a choice between 2-3 different tradeoffs, which provides some strategic depth, as opposed to "choose your reward/punishment" or "take this reward/punishment".

Are there any problems with simply concluding "Dynamical equations produce superior game dynamics except where computationally unfeasible" and "Continuous trade-offs are preferable to hard limits and dichotomies"?

If by the latter you mean trading arbitrary limits and black-and-white binaries for smooth gradients, then personally I would say that those are both very correct conclusions. I think that the reliance on randomness is a consequence of wanting to introduce factors outside of the player's direct control without simulating them with any rigor.

Unfortunately though, these two conclusions, while each correct IMO, are actually in conflict with one another to some extent. Hard limits and binaries do if nothing else lend themselves to less randomness (since they translate directly to on/off thresholds) even if they are arbitrary instead, and though smooth gradients never produce fully random outcomes, they are often implemented with some weighted randomness (how else would you translate it into an on-off trigger for a rebellion, for example?).

 

[unmerged(798670)]

Reposting this here by request (from http://forum.paradoxplaza.com/forum...s-to-1.1.3&p=16403370&viewfull=1#post16403370 )

The major game mechanics that work together to cause all of the other things to be problems and limit choices, this list is not intended to be a suggestion of what “needs” to change, only things that could be changed to allow for a major increase in strategic options with some examples to point out their importance, and things that are so omnipresent that a great deal of other things would be drastically altered by them.

The problems from which all others derive:

The Pacing/Time Limit- the number of provinces and time limit of the game puts you in a position where you need to maintain a pace of a bit less 3 provinces per year from around the world, this is obviously overlysimple but for every mechanic that allows growth at a faster pace than that, the slower options become more viable in limited circumstances. If large countries were able to expand more rapidly, and extremely large countries were capable of seizing the world you could easily have a WC viable strategy with a much slower start than we have now, and a slower/more dangerous ramping up in conquest where every person at every stage of growth would always “feel” they and their country were doing better while facing deadlier challenges because artificial barriers would be gone without necessarily making the game any easier and potentially making it far more deadly.

Coalitions- Coalitions are an unbreakable brick wall that once formed must be militarily ignored almost entirely or will continue to grow infinitely. If there was a way to destroy the coalition, permanently weaken countries at war, or a way to lose AE actively such as overly generous peace terms many of the other problems would be mitigated because coalitions would be a mechanically acceptable event.

Coalition WS-Without scaling to size and scope of the battle coalitions are almost never worth fighting, and victory generally makes your rewards for further victory shrink. There have been at least dozens of threads about this but obviously if a player is so strong they can defeat a coalition, and he actually could take an appropriate reward coalitions would merely be an extremely dangerous and deadly foe capable of destroying you as opposed to an effective game over screen.

AE Multipliers/Caps/Catchall/decay- The case could be made that AE and not coalitions are the problem and this can basically be put into three large cases, The first being that AE is currently a catchall negative which makes many of the expansion methods just a straight comparison to tell which ones are objectively better than others as opposed to a good idea to use all of them in limited circumstances.
-The second being the lack of a cap and the “fixed” tiny decay- Most people may have noticed this attitude factor is capable of being an order of magnitude or two larger than every other modifier. At the end of the day there is only so much hatred you can have for expansionist behavior, and there are so many generations to keep complete visceral hatred that completely ignores diplomacy without any actual aggression. If there was a sane cap (say something like -200), a method of lowering AE actively or a scaling reduction with amount of AE the possibility to intentionally take a massive landgrab and sit until the world chill-axes which would allow unviable strategies to be useful in very limited circumstances.
-The third being the AE multipliers, particularly as they ramp up with size and in europe, even taking single provinces with the “best” CB’s gives so much overall AE as to not be viable even when taking a single core of a reasonably big country to return cores.

Annexation AE- The ludicrous AE for annexing which apparently gets worse at an uncapped rate with respect to size is so large that you literally can’t annex Europe completely diplomatically without a massive coalition forming. Everyone seems to hate on the idea that people should HATE the idea of being annexed, and perhaps I’m mistaken on what it’s supposed to represent the current government being kept in place while starting to enforce your laws and rely on a more centralized military. The leaders are all still wealthy powerful respectable people, which is infinitely better than landless, beheaded footnotes in history. In any event if a “0%” multiplier to AE is too large to be viable, it goes without saying that no other strategy is practical, not to mention core feeding, or releasing vassals (releasing and reannexing vassals generating more AE is especially comical).

Personal unions- I don’t remember why I put this in this section of the outline, I believe it was because it would be a viable alternative to Europe conquering if you forced the HRE to form and helped a couple powers blob and had a non-random way of claiming thrones. My personal recommendation is to make it so you can only claim a throne in a war if you do not have a PU or a restoration CB, giving you at most 7 taken by force and any inherited by random chance. As is there is simply no reliable way to take a large country even with a large time investment, some other way might be more practical, but the current method was very reasonable, it was just too easy to chain abuse.

I planned on mostly answering stuff and expanding in my thread if I had questions on it, but whatever is preferred.

 

[bleakie]

Sorry, it seems that I misunderstood your post. I interpreted your proposal more along the lines of having the rebel percentage cause that proportion of the province's manpower and income to go into a pool for some set of provinces defined by cultural or religious affiliation, with rebellion triggering once it reaches some critical mass, which I thought was a far superior proposal.

Now, however, it seems the suggestion is essentially identical to my original proposal. This was that instead of rebels in a large empire popping up on a province-by-province basis, they would start to form coherent large-scale rebellions encompassing cultures/religions and their groups (analogous to a mix of nationalist and pretender rebel mechanics, in terms of current mechanics), in order to provide more of a challenge for a large empire. Could you explain the difference between this solution and your solution a little more?

First of all, I must apologize for not organizing and expressing my ideas in detail. I shall give a more complete proposal.


Your conceptualization of my proposal is mostly what I am thinking of, except for one thing: when the rebels will rise. If there is only a fixed trigger point, then rebels would become too predictable, and players can time their moves according to the rebel "clock". So there must be some form of randomization.

My idea is that once rebel strength reaches 50% of maximum (the maximum determined by the support base of the rebel group) or a revolt group has positive rebel strength for 20 years, then there is a chance for a revolt to fire right away, with a lower strength. The chance increases as the rebel strength increases further, until the rebel strength reaches the maximum, when a revolt will automatically trigger.

This design is to make it impossible for players to rely on the "timetable" of the rebels to their benefit (such as timing large-scale revolts to be between wars).
The very nature of rebels (being people that you cannot control) requires them to be not completely predictable. But at the same time, placing a rebel mechanic that a player has zero control of is not the way to make a game fun. What I hope to achieve with my proposal is a transparent mechanic with the ability to generate large-scale revolts and just enough unpredictability to prevent players from exploiting it.

And the differences between our proposals:

1. The conditions for a large-scale revolt to occur. Your stance in the OP is that a large and culturally diverse empire immediately equals an event chain for a large-scale revolt, while my version is the accumulation of rebel strength to a threshold value.

2. The variability of my proposal lies in the scale and the exact time of the revolt. Whether the revolt will happen or not is not random: a revolt is guaranteed to happen if a player mismanages the global revolt risk. On the contrary, relying on a random event as you proposed in the OP is binary in nature. Wielding another mtth event with game-changing power is IMO not a good way for improving the game.


I am sorry to have reacted poorly in my second post. I felt that the description is problematic, but I could not pinpoint the exact point that I want to make. My point is actually that my proposal does not really reduce variability of revolts even though it changes from a stochastic model to a dynamical model, because it is necessary to introduce a continuous variable to the dynamical model to avoid the mechanic being exploited, while the long-term behaviour of the current stochastic model (mtth for revolts in individual provinces) quickly converges to a deterministic outcome for a large empire due to the Central Limit Theorem. Meanwhile, I oppose using stochastic model with singular game-changing mtth events, because it will significantly increase variability/randomness for a large empire.

 

[rbl0010]

I think that culture shift should not *always* be possible, like converting provinces is not always possible either. About cores, well there needs to be a really threatening way to break up giga inherited empires. For example Austria that inherits both Poland and Burgundy (PU and event), it should not be so easy 1) to maintain the PU 2) to keep Burgundy.

They changed the rebel system to make them "real" wars in CK2. I don't know how feasible that would be, but maybe it could work as well, adding to CBs (rebel CBs) and new peace elements specific to these rebels. Example the only green option for Burgundian rebels following the succession would be for Austria to give up Burgundy, but maybe there could be an option to get Burgundy released as a PU or a vassal, and then they would get independance CB or something... which they would use, hopefully.

But certainly there is also some space for improvements on the culture system. It seems to contribute today to the stability of the late game AI empires, as I believe players mostly spend diplo points on other things than culture converting, but AI really loves it and then it just kills the core. This is something France did IRL to a great extent, without being able to successfully make everything disappear from people's minds, especially for the greater assimilated culture groups (Breton, Basque, Catalan/Occitan, Alsacian, to name a few). On the other hand, it's something Austria-Hungary did not do or at least did differently: all the slavic countries, a separate Hungary, Romania, Moldovia, etc. I dont mean to imply a causality here, but simply what seems to be a difference in empire management, if that is something. I just mean with all this that culture converting is too straightforward, too "easy" (just spending diplo pts) and leads to results of extreme stability for very much ahistorical beings. Once well into the 1700's, the configuration in Europe hardly changes in the game, and is also hard to change for a player, either by being too small, or by having to face incredible amounts of AE.

 

[Colombo]

rbl0010: At the final stage of Habsburg empire, they did it. Mainly after split into Austria-Hungaria, where one part (the "Austrian") tried to convert everyone into "german" culture and the other part (hungarian one) tried to convert everyone into hungarian culture. Czechs and Slovaks are proof, that they weren't sucesfull. Even at the beggining of the WW1, our representatives did strife to fight for multicultural A-H empire rather than split into petty states. But strongers means of reppresion from Austrains finallized decision to split.

This could be maybe modeled with "tension" mechanics. So that cultural conversion would rise tension etc. that would rise revolt risk and make cost/time/possibility to convert culture harder.
But as was mentioned, maybe requirement of diplomat to culture conversion would be enough to slow it down.
Late game emergence as "revolutionary states" and kind of late game reformation could also be a nice solution.

 

[sinkingmist]

This is bordering on thread-necro, but I think a thread like this needs to be kept alive so that we don't risk rehashing the same ideas, and instead we can build towards something.

Offending mechanic
The key offender in my opinion, is the point mentioned in the first post: Empires are too stable.
All the countries in EU4 are incredibly centralised monolithic powers, where the ruler (player) rules with absolute authority, bar some random events here and there.
This absolute authority is retained as you gain territory, which means as you grow larger, you grow proportionately more powerful, leading to a snowballing effect where large late-game empires become unstoppable under "normal" game mechanics. Presumably, it's in response to this that we see mechanics that artificially limit expansion.
However, in terms of "historic plausibility", if you really did have a centralised monolithic power, where the same ruler with extensive knowledge (and time-travel-capabilities), could rule with absolute authority, and could maintain this absolute centralised power as they took over the world, it seems plausible that world conquest can happen within a few hundred years (or less).
Now, we can't change the fact that a human player will have extensive knowledge (and time-travel-capabilities), thus solutions really should be in tackling this absolute authority.
If this was resolved satisfactorily, we shouldn't need overly restrictive mechanics like coalitions (N.B. It's not the underlying idea of coalitions that I'm against, merely its current, highly restrictive implementation in the game).

Proposed solutions

• Rebels has already been discussed, though there are some gameplay-issues related to it, like how cultural acceptance and conversion is currently handled. I think that culture-conversion should be removed, and instead there should be player-controlled ways of creating and retaining accepted cultures, which I would say is more historically plausible anyway - this would also have the benefit that all cultures remain, and hence the possibility of culture-related rebels, generation of culture-appropriate tags for independence, etc.
• War Exhaustion: If we consider why we don't usually see lasting globe-spanning empires in history, I think we have 2 key points - 1) Lack of support of the population to war. 2) Resistance of the conquered population, and a desire for independence. (there's also 3, Logistics, but let's ignore that for now) Number 2 can be represented with rebels. Number 1 is presumably represented by War Exhaustion, but currently WE is very easy to manage and generally has fairly minor effects on the whole. Part of the whole centralised absolute authority problem is that pretty much your entire populace is highly supportive of any war effort, contributing their full economic and manpower. I would like to see a mechanic where constant wars, or wars that your population is unlikely to be interested in, leading to a lack of support, and hence a lack of resources for war. You could still try to fight with your reduced resources, and if you're large enough, the scraps you have might be enough (but not so much that you can easily steamroll). Alternatively, you can go after targets you can still take with reduced resources (e.g. primitives overseas).

Keeping the varied cultures under your rule happy (possibly by giving more autonomy, meaning penalties on the provinces, or possibly even the release of [diplo-upkeep-free] vassals), and better modelling of the lack of popular support of wars (meaning more penalties!), could restrict resources that can be mobilised by the player, to the extent that even large late-game empires don't wield huge amounts of resources for warfare, hence hopefully making the late-game more difficult.

 

[zodium]

Holy crap, there it is! I really need to start bookmarking threads, I can never find them again with the search function.

 

[Smokahontuss]

I'm very new to EUIV, and to Paradox games in general, but after reading what you said about Alsace and what someone said above I had an idea. Suppose culture conversion was changed and cultures remained the same for the most part, but to keep the empire stable you had to work to make conquered provinces accept your ruling through showing that these provinces are more than just land to you. You could perhaps increase relations with certain culture groups by diplomatic means, building infrastructure, taking decisions that shows tolerence, etc. and eventually the culture would accept you as their ruler, but remain their culture. This would be similar to how Strasbourg is a city in France, but it has a heavily German influenced culture. However, if you had a powerful leader or you just wanted to you could culturally convert provinces over time similar to a westernizing thing (obviously less severe) similar to what Frederick the Great did to Poland. This would give the player choices of picking a ruling style while increasing activity during peace time. This would also make huge empires less stable because you have to either keep up relations or fight hard to change the culture. Once again, I'm new to Paradox games and apologize if this suggestion is clearly ridiculous to the more experienced players.