A "Surrender" mechanic

[Vanillamarine]

In normal wars you are totally right, but coalition and religious wars are different. They are not something you have no choice of, the AI decides actively to participate and a player aswell. The criteria can be much more unforgiving here in my opinion.

Afterall one can simply NOT join such a war.

EDIT: Yes you are right, once the war leader has surrendered you can only take his shit. If you want stuff from England you better make THEM give it to you, instead of benin.

 

[TheMeInTeam]

In normal wars you are totally right, but coalition and religious wars are different. They are not something you have no choice of, the AI decides actively to participate and a player aswell. The criteria can be much more unforgiving here in my opinion.

Afterall one can simply NOT join such a war.

Oh I see now. Hmm, maybe that would work. Amusingly, right now while coalition wars get to ignore the 5 year occupied rule for some stupid reason, religious league war doesn't. In 1.11 I won the league wars twice by suicide assaulting the WL down after stalling a few years, pretty hilarious outcomes ensued on occasion given what was on both sides. Assaults needed a nerfing but man, they're basically out of the game now and that sucks.

 

[Vanillamarine]

Well, its good that we have figured it out quite quick how to do this stuff. We should do a DLC someday.

It still amazes me that in a game about wars they have never thought about a diplomatic party saying "I give up! What do you want?" to the winning side of a war. Haha.

 

[yerm]

I'm not convinced the AI really even needs a surrender condition, and if it does, it should be either imminent bankruptcy or defending multiple wars.

What it needs is modifiers to its ability to accept peace agreements that are more reasonable. As I've said repeatedly (and TMIT mentioned) it actually does know that it's screwed; I'm quite certain AI Austria can do the same calculations that AI Augsburg is doing that makes Augsburg decide to declare unilateral war into the emperor. They both know Austria is screwed, but Austria can't make peace because the modifiers aren't there. They just need those modifiers - in debt, no manpower, multiple wars, under siege just not occupied yet, etc - to factor into the numeric peace screen.

Hypothetical example - I attack Brunswick. No allies involved. I siege down Hoya and Hannover (I think they're called), and lay siege to Brunswick (the capital with the fort too). I've stackwiped them, obviously. The warscore at this point is likely to be under 50, and their willingness to accept peace is likewise quite low. In fact, since I don't own the only nearby fort, they're going to be extremely stubborn! Fast forward half a year, the fort is breached, the siege is at 49%, the warscore is still under 50 and their acceptance has likely barely gone up. One day later, siege finishes, 100% warscore, accept any demands, no stipulations and no penalties. The problem here is not and will not be that Brunswick needs unconditional surrender, its that it needs to realize it's screwed before the siege finishes, and not go from 40% to 100% overnight when the reality of the situation hasn't actually changed for them, they were just as screwed before. Extend this to things like Austria, stuck in 3 wars, 0 manpower, 15 loans in debt, it has to know it's screwed just as well as the other AI do (they can all do the same AIing after all) but in this case, they refuse to make peace deals while Salzburg and Bavaria have no numerical calcs preventing them from becoming the 4th and 5th to declare wars on Austria.

Basically, just give the AI reasonable ability to make peace on their actual situation, based on what the AI clearly already knows is actually up and less on silly all or nothing counts, with things like a +/- for how the war's going, for manpower&debt, for its interest in the demand outcome, etc. I don't think the AI really should even need unconditional surrender in most normal circumstances, in fact I don't think a human should even need it in SP outside the very extreme (eg silly coalitions).

 

[Vanillamarine]

While many of your points are truly valid yerm there is one thing you are missing and that is the factor of how far the AI should push a war, you completely correctly describe that the AI many times fails to peace out in situations where its over but there is a counter side to this point. What if your AI austria was willing to make peace but the attacking AI is pushing it further? "Making gains" so to say.
Why would the attacking AI side make a quick peace? When should two AI warleaders in a war accept a white peace where zero gains take place? Those modifiers are not as trivial to play with.

EDIT: The obvious benefit of an unconditional surrender would exactly jump in where the modifiers fail. It breaks down the decision to end a war to one simple thing, are the surrender conditions true or false.

EDIT2: Furthermore, the unconditional surrender mechanic brings in a very strong incentive to make peace RIGHT NOW. In form of penalties for the attacker and penalties for the defender if they do not come to terms. That is not the case currently. Why would attacking AI france make a quick peace when it is making gains?

 

[yerm]

AI make peace when the defender is willing to accept everything the attacker expects. If the defender will part with N warscore but the attacker wants N+3, the war will drag on until 3 more warscore acceptance is accrued. If the AI defender calculates that it will willingly part with 137 points worth of demands, the AI attacker will make peace immediately. The only stop here would be another number, again, blocking peace, eg demands mean full annexation or no occupied forts in region giving a big minus to the defender number. Fix the number calculations all around, and the AI attacker will make peace with the AI defender the moment the AI defender's adjusted number falls in range.

 

[Vanillamarine]

AI make peace when the defender is willing to accept everything the attacker expects.

Here lies the inherent problem. No numbers or modifiers are able to tell you what to "expect". At least everything paradox has come up with up to now show that it simply doesn't work.

What do you mean with expect? When and where is the "N+3"? How is the N calculated? Can the AI change the N during the war? How does it change its N?
The defender can calculate these things easily, but how is the winning AI calculating them?

 

[yerm]

Here lies the inherent problem. No numbers or modifiers are able to tell you what to "expect". At least everything paradox has come up with up to now show that it simply doesn't work.

What do you mean with expect? When and where is the "N+3"? How is the N calculated? Can the AI change the N during the war? How does it change its N?
The defender can calculate these things easily, but how is the winning AI calculating them?

The defender, calculating the length of the war, allied strength, its ownership of its capital, etc, determines that it is willing to part with 44 total warscore of demands as long as they do not include a request for full annexation. In this case, N = 44.

The attacker calculates that it wants this province, that province, and the other one, plus reparations of course and their money. This amounts to 57 total warscore of requests. In this case, the attacker wants N+13 and as such will not make peace with the defender until 13 more warscore worth of junk is acceptable, or the attacker's expectations lower 13, or a combination of the two.

If the defender suddenly decides that it will part with 80 warscore now, the attacker will immediately make peace. This will only not happen if another number blocks it - demands exceed warscore, full annexation, or no occupied forts in the area. As long as these other numbers are not being tacked on, the AI will immediately make a peace deal.

When does the AI need to unconditional surrender? If it determines that it is totally beaten and should be willing to accept any demands (exactly what unconditional surrender means) this should not require a separate mechanic; the AI should simply accept these demands in the basic peace mechanics by relenting with a full 100 warscore in demands. If the AI is willing to accept 100 warscore in demands but refuses because of a secondary modifier, such as no occupied forts in the region, then the problem is the AI's numerical calculation that the attacker needs to own those forts needs to be adjusted, not some outside surrender mechanic, such that the AI who would be willing to relent with those no-nearby-fort provinces (unconditional surrender) does so anyway.

So, basically, any time the AI would be programmed to issue an unconditional surrender, it should instead just be programmed to offer 100% no questions asked, which is effectively the same thing. It then would only force the issue using a surrender mechanic if the attacker will not make peace. I am arguing that AI vs AI I do not think this should happen. Human vs AI, I suspect it would mean only when the human has them ground into the dirt but isn't making peace, likely in order to fight allies in the war, and in this circumstance call to peace should just trigger anyway.

So, yeah, unconditional surrender as I see it should not be a crutch for bad gameplay or a tool for the AI, it should be a human-used out for unending wars due to human grief tactics or coalition modifiers, likely no more.

 

[beckermt]

A possible solution: incur immediate "Call to Peace" malus upon reaching 100% warscore. Maybe the warscore needed to achieve it can be tweaked depending on the casus belli. That way you no longer can wage wars with sitting on 100% for months/years and breaking an enemy in that manner. If the normal CtP uptick is too slow, it could also start off rather high.

This would quickly make it damn-near impractical to just keep it going, since you would very quickly get your DP drained out of existence.

Besides, I think it makes sense. I doubt the majority of a country's populace would be happy with keeping a war going that is only sucking up national resources against an enemy who is already defeated and willing to give you everything they can.

Why doesn't Call to Peace happen immediately at 100% warscore?

Maybe additionally, Call to Peace should happen as soon as a VALID offer for peace is sent. It'd be less dramatic than the stabhit, but still a penalty. It may be hard to exploit, because of those lovely Length of War modifiers.