Tver´s vodka is of very stronk.

[Wizzington]

We've discussed adding provincial caps but feel it would be too restricting on players who want to play really small nations. We're going to limit the AI instead, so it doesn't develop backwater provinecs to a ridiculous degree.

 

[Wizzington]

If it's developing a province when it isn't smart, sure, fix it, but please don't make it play intentionally bad.

While I'm normally against having the AI play badly, I think in this case the immersion breaking effect of 40 development Shetlands is enough to justify it. It'll only really impact tiny AI countries, anyhow.

 

[Wizzington]

Honestly, it sounds like appeasing some vocal minority here on the forums. There are many cases in history where backwater provinces were being developed for no apparent reason [incidentally that's how they stopped being backwater provinces].

This change has nothing to do with appeasing anyone, we've been discussing this issue since Common Sense came out and it's the middle ground we eventually settled on to both maintain immersion and 'tall OPM' playstyles.

 

[Wizzington]

It feels odd to say it's immersion breaking for the AI to do it, but then have it benefit the player if they do it themselves. If something is considered immersion breaking, there should be an incentive not to do it at all.

This is basic player psychology. Player breaking the fourth wall and being silly is expected, AI doing the same is not.

For example, if you're playing an RPG and are constantly jumping at all times it won't break immersion, but imagine if you walk into a town and every NPC there is bunnyhopping.

 

[Wizzington]

Another analogy would be making a snake to westernize - the AI could certainly be programmed to do that, and it offers gameplay benefits, but it's a terrible idea from an immersion standpoint. In fact, the AI is heavily restricted in what it conquers purely for immersion/pretty borders reason, so that's what you should really be railing about if you want AI to play optimally at all times.

 

[Wizzington]

Hey, don't knock bunny-hopping off till you tried it

Unrelated, but does this mean you are back to fully work on EU4 and future expansions [or were you just teasing with that cat/tiger in a helmet]?
Or will you go back to helping Stellaris/HoI4 after the patch again?

I'm back on EU4.

 

[Wizzington]

Yes? I've said several times that the current tech system isn't great. It works fine for the purpose of how the game develops, but tech groups are a blunt tool and westernization is an unfun chore. That doesn't mean we're just going to replace it with any random idea though, you don't throw out a working system until you have a much better idea you believe in (and time to implement it).

That has nothing to do with the current discussion, however. Do you think the AI should make westernization snakes and be willing to conquer any territory it would gain gameplay benefits from, no matter how messy the map becomes?

 

[Wizzington]

Does this work the same regardless of difficulty settings? I would be willing to rail about it if that would result in a more interesting "hard" setting. Throw immersion out of the window and free the AI from its shackles.

It wouldn't necessarily lead to a more difficult game, because AIs are ultimately going to be conquering other AIs the vast majority of the time. Much like OPMs not developing as much won't either, because it means less easy grab high development provinces.

 

[Wizzington]

I think a major difference is that one is a legacy mechanic and the other is a fairly new system. I certainly understand that AI can't be perfect and concessions have to be made. I'm wondering if this particular instance is showing a flaw in the development system in general, and if that flaw is better handled directly than having the AI turn a blind eye to it as it does with other things.

I'm perfectly happy with the AI not doing annoying exploitative stuff, but I'd like if players have incentive to not do that as well, in cases where it makes sense.

I think you're stretching your arguments pretty hard here. The reasoning is the same in both cases: We don't want to block the player from making silly conquests like say, East Frisian Italy, but the AI should never go for such a goal no matter how much gameplay sense it might make. Similarly, we don't want to block the player from building 100 development Hamburg but the AI really shouldn't. How new the mechanic is doesn't matter at all - it's the same reasoning behind AI behaviour, and it's fine to be for one and against the other because that's how you would prefer it, just recognize that you don't really have a very good argument for your preference.

 

[Wizzington]

When you do this, could you consider limiting based on an algorithm that considers nearby development? For example, a 30-development Venice is less out of place next to a 26-development Treviso, 24-development Brescia so raising development in Venice from 29 to 30 is more likely than if it were surrounded by 15-development provinces.

Hmm, perhaps.

 

[Wizzington]

Will these limits causes the AI to waste monarch points by going over their cap? It's one thing to say that the AI will prioritize other things than development in certain circumstances, it's another to say they will lose out in general because some hidden cap blocks them from spending.

They'll generally spend the points on policies and maintaining ahead of time bonuses instead.