New Tech system - Human vs AI imbalance

[Violent AI]

The new tech system has a big disparity in the way a human player would play with it, as opposed to AI is playing now, when it comes to regions away from Europe
Some examples that I saw from my various games:

Timurids can embrace Feudalism right at the start, by taking 2-3 loans. A human player would never as Timurids pay 50% more MP for teching, yet an AI would wait 40-50 years for it to spread further before embracing, wasting 3k-4k MP due to lack of Feudalism.

In The east - Ming etc, end up with 240% tech cost in 1590 (I was playing Muscovy and saw their tech cost in Great Power screen). A human player in Asia would never reach that high a tech cost, because he knows that even 50% malus for a single Institution is 300*3 MP extra for each new level of ADM, DIP and MIL, and that 1800-2000 MP (equivalent to 2 tech levels of adm/dip/mil) dumped into development is enough to spawn an institution as well as get more development.
The game outside europe becomes very unbalanced due to this, since the human will always find himself ahead of the AI tech-wise.

In short, the human player knows that if the institution is not going to spread on it's own in his country, he will spawn it himself rather than pay high tech penality for a century or more.
Either the AI needs to be improved at handling this, or the system needs to be reworked.

 

[raikaria]

The new tech system has a big disparity in the way a human player would play with it, as opposed to AI is playing now, when it comes to regions away from Europe
Some examples that I saw from my various games:

Timurids can embrace Feudalism right at the start, by taking 2-3 loans. A human player would never as Timurids pay 50% more MP for teching, yet an AI would wait 40-50 years for it to spread further before embracing, wasting 3k-4k MP due to lack of Feudalism.

In The east - Ming etc, end up with 240% tech cost in 1590 (I was playing Muscovy and saw their tech cost in Great Power screen). A human player in Asia would never reach that high a tech cost, because he knows that even 50% malus for a single Institution is 300*3 MP extra for each new level of ADM, DIP and MIL, and that 1800-2000 MP (equivalent to 2 tech levels of adm/dip/mil) dumped into development is enough to spawn an institution as well as get more development.
The game outside europe becomes very unbalanced due to this, since the human will always find himself ahead of the AI tech-wise.

In short, the human player knows that if the institution is not going to spread on it's own in his country, he will spawn it himself rather than pay high tech penality for a century or more.
Either the AI needs to be improved at handling this, or the system needs to be reworked.

Reminder that humans embraceing institutions creates a ripple effect. Especially if there are alliances [making Friendly Adjacent Province modifiers as well as adjacent].

It dosen't take too long for Korea to cause the entire of East Asia to catch up; for example. Ming keeping up basically keeps the entire world up.

 

[Chaingun]

This is a problem every time you design a new complicated system that the AI needs to use.

I expected Gnivom to continue improving the AI for the new tech system (he wrote this in the first place), but AI probably will always need to play more conservatively than the player in comparison, and not go for exotic teching strategies. That said, the art of making challenging AI is also about keeping the game free enough of exceptional cases so the AI is able to play by as simple rules as possible.

 

[DanubianCossak]

I will be the devil's advocate here and say 2 unpopular things.

1.) EU is not civilization. Im not even sure it would be good if AI's developed optimally to get institutions. It would effectively make everyone equal.

2.) EU is called EUROPA universalis. Not NATIVIA or ROTWia universalis. If everyone is the same tech wise, then it is very difficult or impossible to recreate some of the defining points in history that really drive this game.

 

[Agenor]

Institutions do give the human player a very real advantage. I save up my MPs and wait until I acquire the next institution before teching up, and so save a lot of MPs. Also, the lag can be some fifteen years until a neighbour acquires the institution, which can give you a non-trivial advantage. At least the old system was a level playing field. Now it is skewed in the human player's favour for a decade or two every fifty years.

 

[SignedName]

I will be the devil's advocate here and say 2 unpopular things.

1.) EU is not civilization. Im not even sure it would be good if AI's developed optimally to get institutions. It would effectively make everyone equal.

2.) EU is called EUROPA universalis. Not NATIVIA or ROTWia universalis. If everyone is the same tech wise, then it is very difficult or impossible to recreate some of the defining points in history that really drive this game.

Please stop using 2.) as an argument. CRUSADER Kings doesn't just let you play as Crusaders, and lets you play as more than just kings, as well. VICTORIA doesn't just focus on Great Britain. It's just a name and balancing a game on a Euro-centric power fantasy is just silly, given the main draw is that you can play as ANY country. That said, institutions as they are now are unlockable only through gamey mechanics or waiting, which is not challenging, just boring. If there were actually ways of gaining institutions diplomatically, I think it would go a long way towards making the game much more fun, especially in peace time which everyone complains as a time where you have nothing to do.

 

[grommile]

That said, institutions as they are now are unlockable only through gamey mechanics or waiting, which is not challenging, just boring.

And yet they're still better than Westernization, which if you wanted to do it early enough to be really effective involved ridiculously "gamey" strategies (no-CB war to force-vassalize Shirvan? rushing to range-extend until you can discover the Maghreb so you can get a border with Portugal?) on top of what was honestly a pretty "gamey" mechanic. (And the validity of singling out particular mechanical aspects of EU4 as "gamey" is pretty debatable anyway.)

If there were actually ways of gaining institutions diplomatically,

There kind of is a way to do that, but it's rather limited. Persuading people who share a coastal zone with you to like you unlocks the "nearby friendly province" entry in the institution propagation rules.

 

[DanubianCossak]

Please stop using 2.) as an argument. CRUSADER Kings doesn't just let you play as Crusaders, and lets you play as more than just kings, as well. VICTORIA doesn't just focus on Great Britain. It's just a name and balancing a game on a Euro-centric power fantasy is just silly, given the main draw is that you can play as ANY country. That said, institutions as they are now are unlockable only through gamey mechanics or waiting, which is not challenging, just boring. If there were actually ways of gaining institutions diplomatically, I think it would go a long way towards making the game much more fun, especially in peace time which everyone complains as a time where you have nothing to do.

Have you played EU4 or EU series, like at all?

Euro-Centrism has been like the core concept for years.

EU has never been about creating fantasy scenarios in which India becomes the technological center of the world and it becomes the place of birth of colonialism, imperialism and all of that. If guided by player you can become the leader of the world in many areas, but most of the core game mechanics are scripted precisely, exactly, specifically in such a way that Europe is the main driving force that pushes forward the global narratives.

So no, 2.) is still perfectly valid argument and i stand by my claim. EU IS a game primarily about Europe with everything else being just an added bonus.

Its not a simulator that aims to faithfully recreate history, but it is a simulator of a kind, that aims to recreate a spirit of its era in plausible ways, where Johan(bbhn) is the arbiter who decides what can pass for plausible. For the time being, at least, Europe seems to be the defining force when it comes to plausibility (although i will definitely agree that with the new system the player can achieve a lot more in ROTW than they used to be able, i mean, hell, i was the tech leader with Mali and Ayuttahaya...).

 

[TheMeInTeam]

The game outside europe becomes very unbalanced due to this, since the human will always find himself ahead of the AI tech-wise.

It's not new to this patch, you could say this about any previous patch also...especially ones where the less intuitive methods of securing a western border existed.

At least the AI will sometimes develop for institutions this patch. It's too rare arguably, but I've seen pre-1500 renaissance in India and Korea a few times out of the AI. Compared to 1.17.1 westernization where ROTW AI would regularly lag in tech until well after 1700, it at least has a small chance now.

I will be the devil's advocate here and say 2 unpopular things.

Playing devil's advocate requires participating in the discussion by addressing and refuting points that were made. #1 is obviously untrue (even with optimal human play there is significant variance by region, which is fine) and #2 is tangential to OP's discussion.

At least the old system was a level playing field.

That's not consistent with my experience in the game.

Institutions could use a little work, but it's an improvement overall...assuming you have Cossacks. That's a pretty low blow to rely on prior DLC, but that's the reality.

Its not a simulator that aims to faithfully recreate history, but it is a simulator of a kind, that aims to recreate a spirit of its era in plausible ways, where Johan(bbhn) is the arbiter who decides what can pass for plausible.

Self inconsistency is never plausible, regardless of who commits it. Implementations that are not self-consistent necessarily have at least one aspect that doesn't fit the design.

The notion that institutions should be changed because they are unrealistic from a historical perspective is *absurd* if you're willing to accept the implementation of mechanics like coalitions or war score. Attempting a run at the "simulator" path of logic, you necessarily contradict yourself when arguing this stuff without calling for a different game.

Self-inconsistent arguments are not viable and should not be treated as such with regards to discussion of what makes a good implementation. 60 is not more than 60, and the game is right to fix such inconsistencies (props to QA/devs for fixing that particular tooltip), not propagate them.

 

[DanubianCossak]

#1 is obviously untrue (even with optimal human play there is significant variance by region, which is fine)

What are you talking about? What variance?

If you develop for institutions there is no place in the world that cant keep up with techs.

In other words if AI developed for institutions, vast majority of the world would be same tech (the actual difference would be in the number of unlocked ideas). Difference would come down to some countries not having good provinces to develop (terrain making it too expensive for example) or... having already highly developed provinces, which would make it more expensive?

What am i missing?

and #2 is tangential to OP's discussion.

Nope. Things are very much connected. Unless Europeans get some sort of technological advantage, they will not be able to unlock idea groups and fill them up. Without this there is no colonization in a timely fashion. Without this other parts of the world get a chance to pick exploration and colonization ideas and beat Europeans to it. With this you get a world thats heavily diverging from history (which doesnt seem to be the overall design philosophy).

If other parts of the world get higher techs faster, they start unlocking ideas and technologies that boost their incomes. They start snowballing and becoming more powerful and able to project their power overseas (which doesnt seem to be the overall design philosophy).

With techs they also unlock new governments, open up ways to do other stuff and get new CBs.

All of this is fine if done by player, but is not optimal when done by many AIs.

 

[TheMeInTeam]

What are you talking about? What variance?

If you develop for institutions there is no place in the world that cant keep up with techs.

Some areas of the world start with less than 3 tech. Some areas of the world don't have provinces that cost less than 20-25% more monarch points to develop. Some areas of the world are much, much stronger in trade, which has implications for institution spread and affordability.

There's nothing wrong with those things per se', but they make the "everyone would be equal" untrue. There is no way to make something like Central African AI perform "equally" to something like Northern India or Italy. Assuming equal skill in both regions, Central Africa start will always be behind if relying on development for institutions.

And even in such a scenario, Europe would have plenty of advantage with which to steamroll, so I don't see the issue. European AI doesn't fail to conquer India because England or France are lacking in troops or troop quality in the game. European AI fails to conquer India because even now, its logistics on overseas invasions is leagues behind its ability to fight on its own borders. If Great Britain were that bad at handling troops in reality, they'd have failed to even impress Bengal, let alone conquer most of India .

 

[grommile]

If Great Britain were that bad at handling troops in reality, they'd have failed to even impress Bengal, let alone conquer most of India .

To be fair, the BEIC did need the assistance of a disgruntled local nobleman to take Bengal.

 

[TheMeInTeam]

To be fair, the BEIC did need the assistance of a disgruntled local nobleman to take Bengal.

I'm well aware, but I doubt said nobleman would have been inclined to participate when a smattering of leader-less grunts disembarked on a nearby nation and got shredded with no reinforcement or viable diplomacy conducted in the area. Usually such folk try to sell out to a winning side...

 

[DanubianCossak]

Self inconsistency is never plausible, regardless of who commits it. Implementations that are not self-consistent necessarily have at least one aspect that doesn't fit the design.

The notion that institutions are unrealistic from a historical perspective is *absurd* if you're willing to accept the implementation of mechanics like coalitions or war score. Attempting a run at the "simulator" path of logic, you necessarily contradict yourself when arguing this stuff without calling for a different game.

Self-inconsistent arguments are not viable and should not be treated as such with regards to discussion of what makes a good implementation. 60 is not more than 60, and the game is right to fix such inconsistencies (props to QA/devs for fixing that particular tooltip), not propagate them.

Im not sure what you are talking about here, could you elaborate please.

European AI fails to conquer India because even now, its logistics on overseas invasions is leagues behind its ability to fight on its own borders. If Great Britain were that bad at handling troops in reality, they'd have failed to even impress Bengal, let alone conquer most of India .

Right now i think they fail to conquer India because naval invasions dont seem to be working optimally. I have CNs in western Africa in my mod, and i often see Spain plop a CN there only to have it destroyed by Mali because Spain is unable to land troops and defeat Mali for some reason.

 

[TheMeInTeam]

Im not sure what you are talking about here, could you elaborate please.

So no, 2.) is still perfectly valid argument and i stand by my claim. EU IS a game primarily about Europe with everything else being just an added bonus.

Its not a simulator that aims to faithfully recreate history, but it is a simulator of a kind, that aims to recreate a spirit of its era in plausible ways, where Johan(bbhn) is the arbiter who decides what can pass for plausible.

I refuted #2 as a valid argument and assert that self-inconsistent arguments/implementations are necessarily irrational, and thus invalid regardless of who makes them. I also refute "argument from authority" as something worth considering from an argumentative perspective.

And yes, naval AI is the driving limitation on overseas conquest. I can understand it struggling against a decent 40+ regiment nation that's lagging in tech, but not egregiously behind yet. However, I've seen nations like Portugal lose an offensive war to Pacajes and give up territory while the latter is still primitive, and that's downright perplexing. It's like it isn't sending troops at all sometimes.

 

[DanubianCossak]

I refuted #2 as a valid argument and assert that self-inconsistent arguments/implementations are necessarily irrational, and thus invalid regardless of who makes them.

There is no law or rule that states that they have to be. A game consists out of different modules (or sets of features). There is no law or rule that states that there has to be one (or multiple) unifying or consistent logic that has to apply to all of them. If well made a game can totally function despite that. Reducing arguments to irrational due to this is not justified.

I also refute "argument from authority" as something worth considering from an argumentative perspective.

You can refute whatever you want, thats anyone's right and a completely subjective thing. However what i described is the objective reality.

If you (not you in person but anyone) have your own personal and subjective expectations about a game, and this game does not meet them, the problem that arises from that is not game's fault. Game is under no obligation to meet your expectations.

And yes, naval AI is the driving limitation on overseas conquest. I can understand it struggling against a decent 40+ regiment nation that's lagging in tech, but not egregiously behind yet. However, I've seen nations like Portugal lose an offensive war to Pacajes and give up territory while the latter is still primitive, and that's downright perplexing. It's like it isn't sending troops at all sometimes.

Agreed, ive seen this as well, and its something that cant possibly be WAD.

 

[TheMeInTeam]

There is no law or rule that states that they have to be. A game consists out of different modules (or sets of features). There is no law or rule that states that there has to be one (or multiple) unifying or consistent logic that has to apply to all of them. If well made a game can totally function despite that. Reducing arguments to irrational due to this is not justified.

If the rationale for doing something is X, doing other things that undermine and contradict X is irrational. There is no "law" or "rule" stating one has to be rational, but that doesn't change the reality that self-contradicting is necessarily irrational.

I am of course operating under the assumption that the stated rationale is honest, which is a fair assumption absent evidence to the contrary. There are other environments (prisoner's dilemma type scenarios or certain kinds of games like town of Salem) where apparent self-contradiction can be rational, because one or several agents are lying about parts of it. There's no incentive for that here though.

You can refute whatever you want, thats completely subjective. However what i described is the objective reality.

You have it back to front. Arguments from authority (your Johan line) are logical fallacies, and the reality is that the game is not a historical simulator in any consistent sense.

If you (not you in person but anyone) have your own personal and subjective expectations about a game, and this game does not meet them, the problem that arises from that is not game's fault. Game is under no obligation to meet your expectations.

I'm not arguing from my "subjective expectations", and it is both non-sequitur and intellectually rude to attempt that line given the discussion to this point. I'm arguing from what is actually in the game, plus recollection of stated intention regarding mechanics.

 

[raikaria]

Institutions could use a little work, but it's an improvement overall...assuming you have Cossacks. That's a pretty low blow to rely on prior DLC, but that's the reality.

Common Sense; not Cossacks.

Cossacks is one of the Big 3 with Estates; but it's Common Sense that development is tied to. [And development should REALLY have been made a free feature with this patch and something added to Common Sense owners in return; because Development is now integral to a core mechanic]

 

[bbqftw]

Just give the AI a development cost bonus on normal and increase the VH bonuses commensurately if that is getting to be a serious problem.

It does seem to help - I did see both Bahmanis and Ming independently dev-forcing their institutions on VH (which already has AI dev bonus)

 

[Canute VII]

This is a problem every time you design a new complicated system that the AI needs to use.

I expected Gnivom to continue improving the AI for the new tech system (he wrote this in the first place), but AI probably will always need to play more conservatively than the player in comparison, and not go for exotic teching strategies. That said, the art of making challenging AI is also about keeping the game free enough of exceptional cases so the AI is able to play by as simple rules as possible.

Yep, making it that a 50% tech malus evaporates by clicking a button opens the gates for gamey strategies only human players would and should go for. Reducing the tech malus to say 25% but not having it evaporate after embracing an institution but rather decline over time, that would probably give a better system for both humans and AI.