How much does « flavour » affect gameplay?

[InvisibleBison]

No PDX GSG is able to simulate "the principles that run our universe", therefore we need to introduce railroading so that the came can follow the only known instance of history as it happened - our irl history. We are decades away from a simulation that can produce anything even remotely interesting as far as alt-history is concerned, anything natively generated by the Clausewitz engine will be AI-slop.

What you're asking for is flat-out impossible, though. As soon as the player takes actions that are different from what their nation historically did, the whole game is irretrievably thrown off course from real-world history, and railroading to get it back on track is at best incoherent nonsense even worse than the so-called "AI-slop" you so enthusiastically condemn, and at best fully impossible (eg, how are you going to have the Napoleonic Wars if France has been partitioned between Spain and Scotland?).

 

[JacceDON]

What you're asking for is flat-out impossible, though. As soon as the player takes actions that are different from what their nation historically did, the whole game is irretrievably thrown off course from real-world history, and railroading to get it back on track is at best incoherent nonsense even worse than the so-called "AI-slop" you so enthusiastically condemn, and at best fully impossible (eg, how are you going to have the Napoleonic Wars if France has been partitioned between Spain and Scotland?).

Of course a complete reconstruction of our universe is impossible, but the same goes for anything resembling a reconstruction of the general conditions that gave rise to history as well. The best we can do is railroad the game so that we get most of the outcomes despite not actually simulating the reasons why. Nothing interesting can come from the dynamic systems of the Clausewitz engine, and even if the Napoleonic wars are made impossible due to a partition between Spain and Scotland all other things not affected should still happen.

 

[JacceDON]

I fail to see how intelligently giving general conditions for some events or content to happen is "not interesting".

What is even "intelligently giving general conditions for some events or content to happen" supposed to mean?

It can be done partially at first, like when you remove the wheels from a bicycle but don't let go. If the "generic" implementation fails because there are historically adapted events happening everywhere, such as every country having a "war of the roses" like civil war and every country opting for space marines, then you readjust the requirements.

In fantasy land maybe, but in reality it will all just turn into randomized (and thus uninteresting) AI-slop.

 

[InvisibleBison]

The best we can do is railroad the game so that we get most of the outcomes despite not actually simulating the reasons why.

No, that's not "the best we can do", that's creating an acausal game that doesn't respect the player's actions. I'd much rather have a game where there's no specific content (what you would call boring AI-slop) that reacts believably to my actions than one with lots of interesting, bespoke mechanics but where what I do doesn't matter.

Nothing interesting can come from the dynamic systems of the Clausewitz engine,

I disagree. I've had plenty of interesting things arise from dynamic systems in various Clausewitz engine games.

and even if the Napoleonic wars are made impossible due to a partition between Spain and Scotland all other things not affected should still happen.

That's the thing - over the scale of time that EUV operates on, there are no things not affected by any given event. The consequences of even very minor things cascade out over time to change everything in the world.

 

[Kvesir]

If you lack imagination hehe




You aren't calling the principles that run our universe "chatgpt" and "IA slop", are you?


Yeah, but that still isn't good for roleplay.


Yes.


I disagree that they are problems. This "content" is simply the historically determined consequences of events that happened before the start date. And most of them shouldn't even be 100% likely to happen. And it is true, there is so much you can simulate, but doesn't mean they should't try.


I will go even further than that and i will say that historical flavour should be optional through a game rule.


Oh no, please, leave that to eu4.


I would advise against eating two cakes at the same time.


Well, I was going to say what a country, nation, tag, state is. It is a huge organisation and structure determined by its geography, terrain elevations, soil fertility, ressources, climate, diplomacy, geopolitics, cultures, religions, governments, rulers, history, etc. That's where uniqueness and deepness should come from. The uniqueness and deepness that come from historically railroaded flavour are a charming illusion that hides a poor game design.


The only thing I will add to this is that Civ is much more railroaded than apparent. And no, I won't elaborate any further lol.

I will finnally add an example of something that is kind of similar to how I role play:

Massalia Tales - Rome 2 Narrative Let's Play [Complete]

A Rome 2 NLP as Massalia, fighting to remain an independent city state and maybe even become a power to rival the other Mediterranean rulers...

m.youtube.com

For me, the fact the barebones of the game is a sandbox where anyone can come on top, and the way historical situations are implemented such that alternative outcomes can happen, is good enough. My bigger concerns are about the variability of playstyles and the game being fun and not feeling like you are actually always playing the same country (which is what Vicky 3 suffers from).

So our expectations are different, and that's okay, but I think you will ultimately be disappointed at release. The devs chose a certain scope for the depth of the sandbox mechanics in EU5 and flavor that exceeds these mechanics uses more shallow solutions instead. They are not attempting to build a simulation that can create any scenario organically (and let's be realistic here, there are performance limitations to that). An example that annoys me is governors - appointing governors over parts of your lands (like in Imperator Rome) isn't in the base game according to Pavia but certain countries have important flavor that references "governors". We will probably see more such examples when the game comes out.

 

[Kvesir]

No, that's not "the best we can do", that's creating an acausal game that doesn't respect the player's actions. I'd much rather have a game where there's no specific content (what you would call boring AI-slop) that reacts believably to my actions than one with lots of interesting, bespoke mechanics but where what I do doesn't matter.


I disagree. I've had plenty of interesting things arise from dynamic systems in various Clausewitz engine games.


That's the thing - over the scale of time that EUV operates on, there are no things not affected by any given event. The consequences of even very minor things cascade out over time to change everything in the world.

Why do you and so many other people in this thread assume that just because a mechanic represents a historical situation it will always have a single outcome? Of course our choices will matter. EU5 is a game, not an interactive history book.

 

[InvisibleBison]

Why do you and so many other people in this thread assume that just because a mechanic represents a historical situation it will always have a single outcome? Of course our choices will matter. EU5 is a game, not an interactive history book.

I don't think historical mechanics in the game will have a single outcome. I'm arguing against that idea because JacceDON is arguing for it. If they weren't making the arguments they're making, I wouldn't be addressing that idea at all.

 

[SAmaster]

It’s not that there’s an optimal path in these games, we know the history enough to say that some strategies are vastly superior to others, but that the difficulty of executing that strategy is relatively consistent, even when it goes hard against the historical trends. In Vicky 3 countries feel very similar because interest groups all feel the same with only minor variations, and disrupting historical tendencies is far too easy (ending American slavery and industrializing ahistorically being the two big ones, but I’d note how I turned the U.S. into a monarchy under Norton on a whim as exemplary of how maliable countries are, to the point they barely have an identity). Imperator might actually be worse because at least in Vicky 3 the starting positions are about what you’d expect. In Imperator every country plays like a weird mashup of Rome and the Diadochi, with the Diadochi feeling the best, but really nobody coming off as especially well represented. Should it have been possible to adopt Roman style inclusive citizenship? Sure, it was absolutely a winning formula, but for most tags getting it adopted should be extremely destabilizing, and is absolutely shouldn’t be implemented by default with how the assimilation mechanics function.

I’m optimistic EU5 will avoid this though. The HRE shouldn’t play like France, which shouldn’t play like China, which shouldn’t play like Japan, which shouldn’t play like Timur, etc. not maybe I’ll be wrong and all the worst problems of the previews where every tag was a blober where going all in on trade capacity was the winning move will be as as present as ever, never to be rectified, but the devs seem conscientious of these problems and even if they aren’t I can see the tools for moderately to fix these things.

Yeah, I felt that there was pretty much no difference in gameplay for playing tribal nations vs. 'civilized' ones. Playing as Athens SHOULD feel EXTREMELY different than say playing a Caledonian tribe. Flavor was pretty much restricted to a handful of tags thanks to IR's short lifespan.

If you lack imagination hehe




You aren't calling the principles that run our universe "chatgpt" and "IA slop", are you?


Yeah, but that still isn't good for roleplay.


Yes.


I disagree that they are problems. This "content" is simply the historically determined consequences of events that happened before the start date. And most of them shouldn't even be 100% likely to happen. And it is true, there is so much you can simulate, but doesn't mean they should't try.


I will go even further than that and i will say that historical flavour should be optional through a game rule.


Oh no, please, leave that to eu4.


I would advise against eating two cakes at the same time.


Well, I was going to say what a country, nation, tag, state is. It is a huge organisation and structure determined by its geography, terrain elevations, soil fertility, ressources, climate, diplomacy, geopolitics, cultures, religions, governments, rulers, history, etc. That's where uniqueness and deepness should come from. The uniqueness and deepness that come from historically railroaded flavour are a charming illusion that hides a poor game design.


The only thing I will add to this is that Civ is much more railroaded than apparent. And no, I won't elaborate any further lol.

I will finnally add an example of something that is kind of similar to how I role play:

Massalia Tales - Rome 2 Narrative Let's Play [Complete]

A Rome 2 NLP as Massalia, fighting to remain an independent city state and maybe even become a power to rival the other Mediterranean rulers...

m.youtube.com

You're being awfully reductionist if you think all it takes to simulate history is the right numbers on a spreadsheet and a good AI. It's also an extremely impersonal approach to the subject, and pretty must destroys any role of historical narratives.

I fail to see how intelligently giving general conditions for some events or content to happen is "not interesting". It can be done partially at first, like when you remove the wheels from a bicycle but don't let go. If the "generic" implementation fails because there are historically adapted events happening everywhere, such as every country having a "war of the roses" like civil war and every country opting for space marines, then you readjust the requirements.

I don't like the 'lets fuck up first, and try to correct afterwards' mentality.

 

[dbruser]

For me, the fact the barebones of the game is a sandbox where anyone can come on top, and the way historical situations are implemented such that alternative outcomes can happen, is good enough. My bigger concerns are about the variability of playstyles and the game being fun and not feeling like you are actually always playing the same country (which is what Vicky 3 suffers from).

So our expectations are different, and that's okay, but I think you will ultimately be disappointed at release. The devs chose a certain scope for the depth of the sandbox mechanics in EU5 and flavor that exceeds these mechanics uses more shallow solutions instead. They are not attempting to build a simulation that can create any scenario organically (and let's be realistic here, there are performance limitations to that). An example that annoys me is governors - appointing governors over parts of your lands (like in Imperator Rome) isn't in the base game according to Pavia but certain countries have important flavor that references "governors". We will probably see more such examples when the game comes out.

The thing is, a barebones sandbox is one of the biggest criticisms of the Paradox games that fail.
Lacking unique flavour/content kills a lot of the replicability of the game and playing nations end up feeling very samey.

Imperator is a great example. The biggest complaint (especially after they fixed mana) and one of the biggest downfalls that cause Imperator to be dropped was 90% of the map had very minimal flavour. It also was one of the biggest detractors from Vic3 at launch (along with it's anemic warfare system).

The biggest thing is railroading events - especially early game - tend to make for more interesting and varied setups, and definitely more challenge (as without strong modifiers, AI fails to put up any meaningful challenge). A good example is AI Ottomans/France. If they didn't have good country bonuses, they would end up being pushovers, the Ottomans would rarely meaningfully expand. If AI Spain/Portugal didn't have colonial bonuses, frankly they would be terrible at colonizing and super weak ever single game.

Past games have proven that flavour and content drives engagement and raises playerbase. People are WAY more likely to play countries with lots of unique things. Games lacking in flavour (Vicky/Imperator) fall off and are never able to recover from the bad launch.

 

[SAmaster]

What you're asking for is flat-out impossible, though. As soon as the player takes actions that are different from what their nation historically did, the whole game is irretrievably thrown off course from real-world history, and railroading to get it back on track is at best incoherent nonsense even worse than the so-called "AI-slop" you so enthusiastically condemn, and at best fully impossible (eg, how are you going to have the Napoleonic Wars if France has been partitioned between Spain and Scotland?).

So you don't believe in the 'in spite of a nail' approach? The whole 'butterfly effect' gets really overblown by alt-historians. Real history happens for several compounding effects piling on top of one another.

As an example- while Hitler was certainly the prime driver of WW2, WW1 was pretty much destined to happen from the outset as pretty much every nation involved wanted a war, and failed to really appreciate the kind of war they were gonna get. If Franz Ferdinand had survived his assassination, some other flashpoint likely would have kicked it off, or in another alt-history scenario some other conflict or flashpoint could have kicked WW1 off earlier.

A lot of historical events aren't going to shift so drastically because of player intervention. As an example- if say I as Scotland manage to conquer all of the British Isles by 1350, should that really impact the history of Meso-America? Should every AI in Meso-America be set to fully random in the name of 'freedom' and 'alt-history'?

Why do you and so many other people in this thread assume that just because a mechanic represents a historical situation it will always have a single outcome? Of course our choices will matter. EU5 is a game, not an interactive history book.

It's a weird strawman. No-one on the pro-missions side is arguing for 100% historical accuracy or railroading. Maybe 75% at most. But they seem to be only satisfied with 100% sandbox. I've brought this up a few times in this and other threads, but nobody seems to want to engage on this point- what middle point if any should be aimed for? I think it's because that's the more moderate position, but they aren't arguing for a moderate solution, they're arguing for the extreme 100% sandbox approach. Which I find ridiculous, because a game that puts so much effort into an authentic 1337 start date will never be 100% sandbox. I think they confuse 'simulation' with 'sandbox'- EU5 will certainly be more simulationist than abstracted than EU4, but while 'railroading' is definitely an abstraction, that's not inherently a bad thing.

 

[Palfouri]

Why do you and so many other people in this thread assume that just because a mechanic represents a historical situation it will always have a single outcome? Of course our choices will matter. EU5 is a game, not an interactive history book.

Most of the people arguing for them in these threads seem to be doing so on the basis of being able to railroad history with them. If the goal is to force a specific outcome then that definitely seems to me like arguing that choices ultimately shouldn't matter.

 

[JacceDON]

Most of the people arguing for them in these threads seem to be doing so on the basis of being able to railroad history with them. If the goal is to force a specific outcome then that definitely seems to me like arguing that choices ultimately shouldn't matter.

I think that it's mostly me who is arguing for that, don't take my (arguably quite extreme) stance and make it out to be the standard for anyone who isn't in favour of maximum sandbox.

 

[Kvesir]

Most of the people arguing for them in these threads seem to be doing so on the basis of being able to railroad history with them. If the goal is to force a specific outcome then that definitely seems to me like arguing that choices ultimately shouldn't matter.

Idk what most people think but my argument is that generic flavor is less immersive than historically based flavor, and that having deep systems that simulate so much that flavor isn't necessary is unrealistic. People who expect the latter will be disappointed, and I think ultimately a big selling point of EU5 will be the amount and quality of unique flavor that it will have.

What I can't understand is why people think historical flavor = historical outcomes.
Like, we have the mechanic of the 30 years war and it's probably railroaded to happen once enough countries convert to Protestantism in the HRE. It seems like a really fun mechanic and it definitely improves the quality of the game, and idk how a generic mechanic could have replaced it unless we got a deep bespoke system about religious schisms (which we don't). But the outcome of the 30 years war is not railroaded.

 

[Palfouri]

What I can't understand is why people think historical flavor = historical outcomes.
Like, we have the mechanic of the 30 years war and it's probably railroaded to happen once enough countries convert to Protestantism in the HRE. It seems like a really fun mechanic and it definitely improves the quality of the game, and idk how a generic mechanic could have replaced it unless we got a deep bespoke system about religious schisms (which we don't). But the outcome of the 30 years war is not railroaded.

Because so much of this historical flavor is basically built on the presumption that historical outcomes in previous events have occurred and that your game world in 1500 or 1600 is extremely similar to IRL. One of the most frequent questions I saw in the thread about the Reformation/30Years' War was people asking if they could avert it or even the Reformation from even happening at all and the devs pretty much said no. It's very flavorful and unique content but its existence impacts how much you can actually affect history in-game and I think a lot of people don't like that, or the fact that similar religious friction in other religions can't be represented because the devs haven't made specific content for it. This ties into the end of your posts, because I think people WANT Paradox to put effort into designing those "genetic" bespoke systems for religious schisms that can be used in a variety of situations instead of just arbitrarily locking them behind tags and regions and specific events.

 

[ATSCMS]

No PDX GSG is able to simulate "the principles that run our universe", therefore we need to introduce railroading so that the came can follow the only known instance of history as it happened - our irl history.

Not with that attitude.

We are decades away from a simulation that can produce anything even remotely interesting as far as alt-history is concerned, anything natively generated by the Clausewitz engine will be AI-slop.

In fantasy land maybe, but in reality it will all just turn into randomized (and thus uninteresting) AI-slop.

That analogy with AI slop is so over the top it doesn't even deserve to be taken seriously.

So our expectations are different, and that's okay, but I think you will ultimately be disappointed at release. The devs chose a certain scope for the depth of the sandbox mechanics in EU5 and flavor that exceeds these mechanics uses more shallow solutions instead. They are not attempting to build a simulation that can create any scenario organically (and let's be realistic here, there are performance limitations to that). An example that annoys me is governors - appointing governors over parts of your lands (like in Imperator Rome) isn't in the base game according to Pavia but certain countries have important flavor that references "governors". We will probably see more such examples when the game comes out.

We will see. They can always change it (a bit) before release or patch it later on so that both playstyles can exist.

You're being awfully reductionist if you think all it takes to simulate history is the right numbers on a spreadsheet and a good AI. It's also an extremely impersonal approach to the subject, and pretty must destroys any role of historical narratives.

What do you mean by "awfully reductionist"? I mean, yeah, I want to awfully reduce the railroading lol. And by "impersonal approach"? And "historical narratives"?

I don't like the 'lets fuck up first, and try to correct afterwards' mentality.

It is still better to try new things than to be stuck in the rusty old ways. There are so many railroaded history games it isn't even funny.

For me, the fact the barebones of the game is a sandbox where anyone can come on top, and the way historical situations are implemented such that alternative outcomes can happen, is good enough. My bigger concerns are about the variability of playstyles and the game being fun and not feeling like you are actually always playing the same country (which is what Vicky 3 suffers from).

The thing is, a barebones sandbox is one of the biggest criticisms of the Paradox games that fail.
Lacking unique flavour/content kills a lot of the replicability of the game and playing nations end up feeling very samey.

Imperator is a great example. The biggest complaint (especially after they fixed mana) and one of the biggest downfalls that cause Imperator to be dropped was 90% of the map had very minimal flavour. It also was one of the biggest detractors from Vic3 at launch (along with it's anemic warfare system).

It doesn't have to be barebones.

The biggest thing is railroading events - especially early game - tend to make for more interesting and varied setups, and definitely more challenge (as without strong modifiers, AI fails to put up any meaningful challenge). A good example is AI Ottomans/France. If they didn't have good country bonuses, they would end up being pushovers, the Ottomans would rarely meaningfully expand. If AI Spain/Portugal didn't have colonial bonuses, frankly they would be terrible at colonizing and super weak ever single game.

Billions must complain again about the Ottomans being overpowered.

Past games have proven that flavour and content drives engagement and raises playerbase. People are WAY more likely to play countries with lots of unique things. Games lacking in flavour (Vicky/Imperator) fall off and are never able to recover from the bad launch.

Unfortunately, profits take primacy over artistic direction.

So you don't believe in the 'in spite of a nail' approach? The whole 'butterfly effect' gets really overblown by alt-historians. Real history happens for several compounding effects piling on top of one another.

As an example- while Hitler was certainly the prime driver of WW2, WW1 was pretty much destined to happen from the outset as pretty much every nation involved wanted a war, and failed to really appreciate the kind of war they were gonna get. If Franz Ferdinand had survived his assassination, some other flashpoint likely would have kicked it off, or in another alt-history scenario some other conflict or flashpoint could have kicked WW1 off earlier.

Your examples are all about short term consequences. Of course those events would have a high probability of happening. High but not complete.

A lot of historical events aren't going to shift so drastically because of player intervention. As an example- if say I as Scotland manage to conquer all of the British Isles by 1350, should that really impact the history of Meso-America? Should every AI in Meso-America be set to fully random in the name of 'freedom' and 'alt-history'?

This example is odd, there is no reason for two mostly unrelated places to affect each other most of the times. Regardless of what we are discussing here.

It's a weird strawman. No-one on the pro-missions side is arguing for 100% historical accuracy or railroading. Maybe 75% at most. But they seem to be only satisfied with 100% sandbox. I've brought this up a few times in this and other threads, but nobody seems to want to engage on this point- what middle point if any should be aimed for? I think it's because that's the more moderate position, but they aren't arguing for a moderate solution, they're arguing for the extreme 100% sandbox approach. Which I find ridiculous, because a game that puts so much effort into an authentic 1337 start date will never be 100% sandbox. I think they confuse 'simulation' with 'sandbox'- EU5 will certainly be more simulationist than abstracted than EU4, but while 'railroading' is definitely an abstraction, that's not inherently a bad thing.

I litterally suggested to have a game rule so that one can simply toggle the playstyle they don't like out.

 

[Friedrich von Deutschland]

Playing as Athens SHOULD feel EXTREMELY different than say playing a Caledonian tribe.

The Poleis-Kingdom distinction is understated too. Greek Poleis weren’t part of the Italian Thunderdome and should play accordingly. Historically extracting military power from them was basically impossible, but in game unifying or conquering them is really easy. Macedon annexing Poleis should t just be a political nightmare as they all unify against you and call in some other great power like Rome or the Selucids, it should also be a let loss as you waste resources occupying them for little return.

Circling back to EU5 though, it’s not just that playing different tags should feel different, how you interact with them should be different as well. As Europe expanding into Mesoamerica, North America, the Andes, India, China, etc should all feel different and require different strategies, even if you are playing the same tag in all cases.

 

[SAmaster]

Idk what most people think but my argument is that generic flavor is less immersive than historically based flavor, and that having deep systems that simulate so much that flavor isn't necessary is unrealistic. People who expect the latter will be disappointed, and I think ultimately a big selling point of EU5 will be the amount and quality of unique flavor that it will have.

What I can't understand is why people think historical flavor = historical outcomes.
Like, we have the mechanic of the 30 years war and it's probably railroaded to happen once enough countries convert to Protestantism in the HRE. It seems like a really fun mechanic and it definitely improves the quality of the game, and idk how a generic mechanic could have replaced it unless we got a deep bespoke system about religious schisms (which we don't). But the outcome of the 30 years war is not railroaded.

It feels like they're not arguing about the game that exists, but a game that exists in their head.

Because so much of this historical flavor is basically built on the presumption that historical outcomes in previous events have occurred and that your game world in 1500 or 1600 is extremely similar to IRL. One of the most frequent questions I saw in the thread about the Reformation/30Years' War was people asking if they could avert it or even the Reformation from even happening at all and the devs pretty much said no. It's very flavorful and unique content but its existence impacts how much you can actually affect history in-game and I think a lot of people don't like that, or the fact that similar religious friction in other religions can't be represented because the devs haven't made specific content for it. This ties into the end of your posts, because I think people WANT Paradox to put effort into designing those "genetic" bespoke systems for religious schisms that can be used in a variety of situations instead of just arbitrarily locking them behind tags and regions and specific events.

It's unrealistic to make the Catholics not want to be corrupt. It's kind of endemic to Catholicism (by which I mean- one giant centralized religious political organization that spans europe).

 

[SAmaster]

The Poleis-Kingdom distinction is understated too. Greek Poleis weren’t part of the Italian Thunderdome and should play accordingly. Historically extracting military power from them was basically impossible, but in game unifying or conquering them is really easy. Macedon annexing Poleis should t just be a political nightmare as they all unify against you and call in some other great power like Rome or the Selucids, it should also be a let loss as you waste resources occupying them for little return.

Circling back to EU5 though, it’s not just that playing different tags should feel different, how you interact with them should be different as well. As Europe expanding into Mesoamerica, North America, the Andes, India, China, etc should all feel different and require different strategies, even if you are playing the same tag in all cases.

Yeah, I didn't feel any sort of democratic weight whatsoever in the game. Monarchies just felt mechanically simpler.

 

[ATSCMS]

It feels like they're not arguing about the game that exists, but a game that exists in their head.

I'm arguing about a game I desire, a game I want to exist.

 

[Friedrich von Deutschland]

Yeah, I didn't feel any sort of democratic weight whatsoever in the game. Monarchies just felt mechanically simpler.

I was more talking about how the Greek concept of eleutheria made the Poleis basically ungovernable, since it was pretty much impossible to do a Rome and make one polity expressly superior to the others. Rome could govern multiple Italian cities directly while also having numerous “allies” which were de facto subjects of Rome despite being self governing. Ignoring that the Socii are represented terribly in Imperator, reproducing this in Greece shouldn’t be possible, since eleutheria requires every Poleis be fully independent, unlike the Italian libertas which could exist and even prosper under Roman dominion. Rome kinda duped the Greeks on this by accident, but then they also couldn’t really exploit the Poleis until the Empire for that reason: if they tried the fiction of eleutheria would’ve been shattered.

Back to EU5: this sort of situations where certain populations are highly resistant to annexation should be commonplace. SoPs should be very hard to stamp out and thus require constant military presence until you can expel or assimilate them. Steppe nomads should be similar, very hard for secondary societies to govern as as such potentially not being worth the effort to conquer (Russia and the Qing are the only two states that really pulled this off, we shouldn’t assume just any state could’ve done it, but that these two has institutions which made them uniquely capable at this task of subjugating the steppe. Not that anyone should be locked out, but investment should be required). For more urban polities Mesoamerica is a good example. The tributary empire the Aztec built isn’t comparable to the unitary or quasi-federal states seen in, say, Iberia. This isn’t one tag, or even three tags, ruling directly over an empire but a network of tributary states, each their own tag, which the Aztec triple alliance subjugated. Turning this into a unified tag should be possible, but it should be a timely affair which requires a lot of investment and can’t generally occur prior tolerances European contact. A repeat of Imperator’s Rome where the Socii are basically an odd quirk of your starting position, as opposed to the source of your strength, should be avoided. There an institution which was expanded and only went away towards the end of the timeframe is abolished almost immediately and without consequence. I don’t want that for Mesoamerica. The tributary arrangement should be the path of least resistance and even if the long term payoff is worth it, perusing alternative strategies should be an uphill battle.