Uncap attrition please!

[currylambchop]

I was playing Zimbabwe with defensive ideas (+1 attrition for enemies from each) and Delhi was throwing 50k stacks on my forts in south India and tanking the attrition like it was nothing. They were literally jungle provinces. Delhi shouldn’t be able to just ignore attrition like that.

 

[Opanashc]

That's easily moddable.

 

[grandadmiralbob]

JC, how big was Delhi?

 

[MatthewP]

That's easily moddable.

+1 to this. I understand the complaint, but I also don't really want to see the AI's manpower melting down to 0 in every war just from bad pathing. If your answer is, "They should improve the AI," then we are in agreement, and let's revisit this when they do.

 

[PDXJon]

I was playing Zimbabwe with defensive ideas (+1 attrition for enemies from each) and Delhi was throwing 50k stacks on my forts in south India and tanking the attrition like it was nothing. They were literally jungle provinces. Delhi shouldn’t be able to just ignore attrition like that.

The problem with an uncapped attrition is the AI’s inability to deal with it properly. I think a better course would be increasing the attrition modifiers for ideas and monuments. +1 attrition really isn’t much.

 

[Word_Smith]

Shiny bells and whistles like ideas and monuments aren't a substitute for good game design. If the AI can't handle something as fundamental to the game as attrition, Paradox should fix the AI.

 

[MatthewP]

Shiny bells and whistles like ideas and monuments aren't a substitute for good game design. If the AI can't handle something as fundamental to the game as attrition, Paradox should fix the AI.

I think the issues are diminishing returns and what marginal improvement really gets you. Sure, if they made it a focus paradox could make the AI somewhat better at handling attrition. But they could probably spend a year on just that problem and still not approach the level of a good player. It’s a deceptively hard problem.

So, if the AI gets slightly better but is still doing really dumb stuff, does it improve the game to have that dumb stuff kill a crippling 20% of its army instead of a manageable 5%? I tend to think not, unless the dumb stuff is greatly reduced.

And truthfully we should be careful what we wish for. An AI that handled attrition - or war in general - like a good player but with unlimited capacity for micro might not be as fun an opponent as it seems.

 

[PDXJon]

Shiny bells and whistles like ideas and monuments aren't a substitute for good game design. If the AI can't handle something as fundamental to the game as attrition, Paradox should fix the AI.

You are making the assumption that they haven’t tried. Sometimes fixing things are harder than saying they should be fixed.

 

[currylambchop]

The problem with an uncapped attrition is the AI’s inability to deal with it properly. I think a better course would be increasing the attrition modifiers for ideas and monuments. +1 attrition really isn’t much.

If the AI can’t handle it then just make AI take less attrition. They already don’t use sailors, 1 extra cheat won’t matter:

 

[Saat98]

I believe that there is an easier fix for all of this. We could only uncap attrition at certain scenarios. For example, I think that it is reasonable to uncap attrition only for static troops (so troops sieging a fort would get uncapped attrition), but cap it when traversing land. This way, attrition doesn't become an unmanageable hell and AI's manpower won't disappear every single war. And I think that it might be easier to teach the AI to not siege a fort with way more the number of troops that are strictly needed.

 

[MatthewP]

I believe that there is an easier fix for all of this. We could only uncap attrition at certain scenarios. For example, I think that it is reasonable to uncap attrition only for static troops (so troops sieging a fort would get uncapped attrition), but cap it when traversing land. This way, attrition doesn't become an unmanageable hell and AI's manpower won't disappear every single war. And I think that it might be easier to teach the AI to not siege a fort with way more the number of troops that are strictly needed.

But who does it help? If you teach the AI not to hit uncapped attrition, then nothing really changes. If you don't, then you have the original problem.

As for multiplayer where players might be affected, that's even worse because the new super annoying meta is to tell all but the minimum siege stack of troops to move somewhere right before the month tick then cancel. Whatever fix you come up with for this is going to hit weird corner cases. It's a rabbit hole with what seems to be pretty low payoff even if you finally get it right.

If the AI can’t handle it then just make AI take less attrition. They already don’t use sailors, 1 extra cheat won’t matter:

Ah, so your OP was about multiplayer? That was not obvious to me, but makes sense. But tbh if we're creating a special rule just for multiplayer it does seem like a mod is the way to go. If the cap isn't moddable that could be a suggestion itself.

 

[PDXJon]

If the AI can’t handle it then just make AI take less attrition. They already don’t use sailors, 1 extra cheat won’t matter:

I thought you were talking about the AI.

 

[Ekcrbe]

Ah, so your OP was about multiplayer? That was not obvious to me, but makes sense. But tbh if we're creating a special rule just for multiplayer it does seem like a mod is the way to go. If the cap isn't moddable that could be a suggestion itself.

The cap is definitely moddable. The group I'm in uses a 15% cap, and I think that's a good number for an MP environment. It allows you to focus very heavily on playing tall in harsh terrain and reap some rewards. I think the best we saw someone get was an 8% base rate before accounting for sieging, supply limit, and winter effects.

In general, modding your multiplayer experience is strongly recommended. Every group has different tastes. Don't be tied down by the least-common-denominator base game behavior.

 

[Bouchart]

Attrition was uncapped in the past but the AI couldn't handle it. In this patch I have seen 100k+ stacks siege level 2 forts.

 

[currylambchop]

But who does it help? If you teach the AI not to hit uncapped attrition, then nothing really changes. If you don't, then you have the original problem.

As for multiplayer where players might be affected, that's even worse because the new super annoying meta is to tell all but the minimum siege stack of troops to move somewhere right before the month tick then cancel. Whatever fix you come up with for this is going to hit weird corner cases. It's a rabbit hole with what seems to be pretty low payoff even if you finally get it right.

Ah, so your OP was about multiplayer? That was not obvious to me, but makes sense. But tbh if we're creating a special rule just for multiplayer it does seem like a mod is the way to go. If the cap isn't moddable that could be a suggestion itself.

actually I am still talking about the AI, cuz even if you half AI attrition it will still be more if it’s uncapped

 

[TolHydra]

Raising the cap a few percentages really can't hurt. Manpower is a lot easier to come by these days compared to when the cap was first introduced. As is, attrition is a joke. You can completely ignore it throughout mid to late game and somehow be fine.

 

[Temudhun Khan]

If the AI can’t handle it then just make AI take less attrition. They already don’t use sailors, 1 extra cheat won’t matter:

Wait, what? Why are my allies constantly using favors to demand sailors from me?

The motherfuckers lied to me!

 

[EarlKonrad]

I was playing Zimbabwe with defensive ideas (+1 attrition for enemies from each) and Delhi was throwing 50k stacks on my forts in south India and tanking the attrition like it was nothing. They were literally jungle provinces. Delhi shouldn’t be able to just ignore attrition like that.

This is literally a videogame, not real life.

Hyperbole aside, there are reasons for attrition to be uncapped. That was the case in the past and lead to terrible results where AI would bleed all of its manpower against a Human player all the while players were able to stop this from happening to them with ease. While not a good solution, capping attrition does help the AI and it sure as hell needs all the help it can get.

Shiny bells and whistles like ideas and monuments aren't a substitute for good game design. If the AI can't handle something as fundamental to the game as attrition, Paradox should fix the AI.

At this point the game is too far gone for us to hope for any meaningful improvement to the AI. Think about it for a moment. EU 4 was built upon EU 3 so the team that worked on 3 was, for the most part, carried over to 4. EU 3 has about 15 years and EU 4 9. There hasn't been anyone at the team that truly knows how all this stuff works for a long time and the game has had so many changes over the years that trying to steer the boat into another direction only leads towards disaster (see everything that happened to the AI since Emperor).

Having said all of that, the last public patch does give me hope for the future but I still think that any meaningful changes to the AI will only truly he possible in EU 5 when they get a dedicated AI guy to write a new AI from the ground up.

 

[bebrst]

+1
I would love to finally have some actual punishment to the Ottomans marching through half of Europe to siege an OPM, to see that hiding in Siberia has actual consequences, or the ability to fight a real winter or mountain attrition war and to have defensive ideas competing with currently stronger military idea groups - if yes if AI could only handle this

 

[jamal bakr]

Not sure if its a case of "AI can't handle it" (they already cheese the AI's manpower in a lot of different ways) but its another case of wanting to balance the game around multiplayer.