AI investing in coal pre-enlightenment

  • We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.

Fliits

Recruit
56 Badges
Oct 13, 2018
9
76
  • Magicka
  • Stellaris - Path to Destruction bundle
  • Hearts of Iron IV: Expansion Pass
  • Imperator: Rome
  • Age of Wonders III
  • Europa Universalis IV: Cradle of Civilization
  • Hearts of Iron IV: Expansion Pass
  • Stellaris: Humanoids Species Pack
  • Stellaris: Apocalypse
  • Europa Universalis IV: Rule Britannia
  • Stellaris: Distant Stars
  • Europa Universalis IV: Dharma
  • Shadowrun: Dragonfall
  • Europa Universalis IV: Golden Century
  • Imperator: Rome Deluxe Edition
  • Hearts of Iron IV: Death or Dishonor
  • Stellaris: Ancient Relics
  • Stellaris: Lithoids
  • Hearts of Iron IV: La Resistance
  • Stellaris: Federations
  • Imperator: Rome - Magna Graecia
  • Crusader Kings III
  • Crusader Kings III: Royal Edition
  • Battle for Bosporus
  • Europa Universalis 4: Emperor
  • Stellaris: Necroids
  • Stellaris: Nemesis
  • Hearts of Iron IV: By Blood Alone
  • Victoria 3 Sign Up
  • Hearts of Iron IV: No Step Back
  • Europa Universalis IV: El Dorado
  • Europa Universalis IV: Pre-order
  • Europa Universalis IV: Common Sense
  • Europa Universalis IV: Cossacks
  • Europa Universalis IV: Mare Nostrum
  • Stellaris
  • Stellaris: Galaxy Edition
  • Stellaris: Galaxy Edition
  • Hearts of Iron IV: Cadet
  • Hearts of Iron IV: Colonel
  • Europa Universalis IV: Rights of Man
  • Stellaris: Digital Anniversary Edition
  • Stellaris: Leviathans Story Pack
  • Hearts of Iron IV: Together for Victory
  • Europa Universalis IV: Mandate of Heaven
  • Stellaris: Synthetic Dawn
  • Europa Universalis IV
  • Europa Universalis IV: Call to arms event
  • Europa Universalis IV: Third Rome
  • Europa Universalis IV: Conquest of Paradise
I think that the AI's habit of mass developing any potential coal developing provinces ASAP is not only unrealistic but also game breaking. Provinces like Auvergne, Asturias and Metz become sprawling metrolpoli by the late 1500s all the while countries like Austria, Bohemia, France and Spain ignore their capitals and trade centres to dump all possible development to future coal mines. Needless to say, this is impossible to play around and makes many countries in Europe needlessly difficult to play, since the AI's obsession with theoretical coal development gives the unnatural levels of development way too early, increasing their economic and military capacity while making conquering these otherwise less than stellar provinces practically impossible if they're within the HRE and an absurd challenge without.

I am a fan of the AI learning to develop it's economy and provinces. I love that the game is more difficult because of the challenge the AI provides and not because of some cheap buffs on them or handicaps on the player. I think it is necessary for the AI to be able to manage its economy before it can learn to manage a military. But this isn't that. This is the opposite of a smart AI. This is the AI knowing that it will eventually benefit from an investment without losing much in the process and abusing that fact to its highest limit. This isn't interesting or fun to play against, it's just frustrating and makes the game a race to capture any potential coal provinces near your border before the AI gets them to +40 development and makes them almost impossible to conquer without creating a coalition.

TL;DR: the AI develops its coal provinces too much, which is impossible to play against without development spam, destroying the balance of the game and making it very unfun to play.
 
  • 8
  • 3Like
  • 2Haha
  • 2
Reactions:
One more reason to be glad I didn't buy the British exceptionalism DLC.
 
  • 2
Reactions:
Developing provinces itself is just broken. The AI is actually being incredibly stupid doing this rather than developing all of their provinces to 10/20 dev like players do in mp. Coal mines don't even give a benefit for being developed high, so there's no sense at all behind doing it (especially when the AI devs the tax/manpower of the province).
 
Last edited:
  • 3
  • 1Like
Reactions:
Developing provinces itself is just broken. The AI is actually being incredibly stupid doing this rather than developing all of their provinces to 10/20 dev like players do in mp. Coal mines don't even give a benefit for being developed high, so there's no sense at all behind doing it (especially when the AI devs the tax/manpower of the province).

100% this. Developing provinces (in single player) is something that should almost never be done, and certainly not in the manner the ai generally does.
 
  • 10
  • 1Like
  • 1Haha
Reactions:
I don't have a problem with AI developing coal provinces especially when it is done to spawn institution but there is big problem how it is developed. I had games where Asturias was 45 dev but it was 20 tax, 5 production and 20 manpower. What's the point of developing future coal province if you don't increase production?
When there are multiple coal provinces AI can choose poorly - PLC will develop one of the provinces in Ukraine when it can develop Lublin with accepted culture, cloth, CoT and better terrain. It's also closer to the rest of higher developed provinces so institution can spread better than to neighbouring 3 dev provinces.

When checking AIs development I also noticed that 1.34 added coal to Hainaut.
 
  • 8
Reactions:
Why should the AI (at least on hard/ very hard) not use a feature of the game in the same way as a player?
I find it a problematic argument that "it makes the AI too strong and too difficult to conquer".

What I find much more compelling is, that apparently the logic is flawed. The AI should be smarter in developing, either focusing on a province for institution spawning (Castile will normally not need to force-spawn institutions, and even if, Asturias is definetly not their best province for doing so) or spreading it to min-max the effect. Also it should not focus on coal until it appears, but more valuable trade goods (Castile got a Gold mine!). It should also be avoided that they are wasting MP for dev that they would need for ideas/tech.
 
  • 4
  • 1Haha
Reactions:
Why should the AI (at least on hard/ very hard) not use a feature of the game in the same way as a player?
I find it a problematic argument that "it makes the AI too strong and too difficult to conquer".
90% of the playerbase would rage quit the game if they had to deal with an AI that properly developed. The other 10% would still bitch because the AE/OE/coring costs for conquest would go up by like +100-200% on average (meaning, you need 50-75% admin efficiency just to get back to parity). And lots of people on both sides would additionally bitch about the ahistoricity of every province in the HRE having 90 dev by 1600.

This is a sign that the development mechanic is flawed and broken. Granted I understand you said "hard/very hard", so it'd technically be optional, but the difference would be just absurdly massive way beyond any other difficulty increase that the game has had before.
 
Last edited:
  • 4
  • 1
Reactions:
The problem for this specific case is that the development points for paper/grain/whool and so on is carried over to the development points for coal, which doesn't make any sense because you need completely differnet infrastructure to support these resources. There are two different solutions for this:
1. completely reset the diplo point development or
2. Have two separate diploe dev points for the old resource and for coal

90% of the playerbase would rage quit the game if they had to deal with an AI that properly developed. The other 10% would still bitch because the AE/OE/coring costs for conquest would go up by like +100-200% on average (meaning, you need 50-75% admin efficiency just to get back to parity).
90% of the playerbase would rage quit the game if the AI would abuse any game mechanic as the player would, just look at any of the dev clashs or other multiplayer events, the AI is a stepover, otherwise there would be no way that a Byzantium player could survive
 
  • 2Like
Reactions:
Thanks for mentioning this. I had an Aragon game not too long ago, and I was shocked to discover I'd conquered a 50 dev Auvergne. I couldn't figure out why the AI had done that, but now the mystery is solved. Thanks again.
 
Its not a problem that AI might develop specific provinces or use fore-knowledge of upcoming events, since player does exactly the same.

What does seem to be silly is that the AI apparently has no "upper limit" on development of coal provinces. A player will stop developing provinces at something like 30 development, since by that point there is almost certainly something better to be spending MP on. But the AI just keeps on going to the moon, leading to 60+ dev coal provinces.

It just needs another sense-check in its prioritisation logic. Something that says "is it really worth spending 300 MP pushing this province from 50 to 51 dev, or should we maybe dev up the province next door from 9 to 10??"
 
  • 3
Reactions:
Its not a problem that AI might develop specific provinces or use fore-knowledge of upcoming events, since player does exactly the same.
I agree but the player using future knowledge knows he will build the university and get development efficiency at tech 17, maybe some other discounts to the dev cost from ideas, missions, etc. so he wouldn't develop future coal province in 1500 when it has livestock or wool now. Why can't AI wait until about 1650 before it starts developing coal?
 
  • 1
Reactions:
90% of the playerbase would rage quit the game if they had to deal with an AI that properly developed. The other 10% would still bitch because the AE/OE/coring costs for conquest would go up by like +100-200% on average (meaning, you need 50-75% admin efficiency just to get back to parity). And lots of people on both sides would additionally bitch about the ahistoricity of every province in the HRE having 90 dev by 1600.

This is a sign that the development mechanic is flawed and broken. Granted I understand you said "hard/very hard", so it'd technically be optional, but the difference would be just absurdly massive way beyond any other difficulty increase that the game has had before.

I see it as a sign the playerbase is flawed and broken.

Could it be better? Yes. Most things have room for improvement.
 
  • 2Haha
Reactions:
I find it very problematic to say that the AI becomes too powerful because it develops the asturias. The player never hestitates to develop their provinces whenever they have a mana surplus, while the AI in older patches used to be at 999 mana and take tech 9 years ahead of time, wasting all their points, putting the player in the lead. This obviously gave an advantage. I agree that there is an AI problem when they only dev the coal provinces, however the consequence would actually be for the AI to dev all their provinces to around 10-20 at least, since thats the optimal distribution. This would also feel ahistorical for all regions to turn out equal, but thats how dev cost actually works for the player. A solution might be for the AI to dev different geographies differently, so it feels more historical to have a high dev farmland province. However this will in fact make the game much harder which OP opposed, and there might be programming issues for why paradox havent made the AI develop more.