Dollar auction and total war: AI overcommits to minor wars

[Careful Plum]

I'm in favor of AI considering opportunity cost, but it is very complicated, AI really need to understand how the situation would develop if both enemies get drained. Who has better reserves, whose realm can handle larger stress, who is more lukely to get backstabbed.

No, I'm not in favor of formulas that establish how much AI should lose before giving up on war. And, I'm not in favor of AI giving up on wars in which it is objectivly stronger just because AI lost a few battles. Eve just making AI recognise wars where it only wants to defend and would peace out on white peace, and woyldn't waste troops on sieges would be a good start.

Well, uhm, it still sounds like we agree with each other here. Maybe it's really me who misunderstood the OP. Of course such a hypothetical algorithm would "need" to account for all costs, opportunity costs, strategic costs, whether the province in question belongs to a vassal or its demesne, not just those costs directly expressed in a number in game... And this would mean that, sometimes the AI would fight a war it knows it will lose if draining the enemy is worth it, of course. But the right now, the usual result is the AI wrecking itself completely and being pounced on by everybody who can get a CB on them (slight hyperbole but I think this phenomenon isn't as pronounced as it could be simply because of the relative scarcity of CBs. And because every AI wrecks itself...).

I absolutely didn't mean to say I'm in favor of having a formula to make the AI give up after a certain amount of losses. In fact, this is pretty much the current system! Instead, in an ideal world, the AI would calculate how much it will likely lose (or rather, still in an ideal world, it would calculate how likely each possible result would be), and then decide whether to fight or give up. Naturally, this would probably make a little more hidden information necessary, or there wouldn't be any wars .

And of course this algorithm is almost impossible to get "right" (or even as good as the one the player effectively uses).

What we maybe disagree on is whether an actual, realistically implementable algorithm could be better than the current "always total war always" system.

 

[StarSword]

If someone considerably bigger than me declares war on me for a single county, I just look at his size, look at my levies, look at his size again and then instant surrender. I don't fight wars that are almost certainly lost although sometimes I just say "fuck it" and go for it.

I once started as some Norse duchy in Ukraine and needed a port, so I declared a war on Byzantium for the county they have in Crimea, I got it and raised massive amounts of tribal armies to keep it. Byzantium kept marching its death stacks onto me and also invading from the sea, but Odin must have been watching over Crimea as I won battles against 4 times bigger stacks and finally invaded Byzantium to get the warscore to 100% for a single county.

The point is. Byzantium completely depleted its army to fight for a measly county in Crimea and the consequences of doing that meant civil war and invasions from its neighbors. Sometimes giving away the county is the best option.

Yeah, I had a similar thing happen during my conquest of Iceland. The war shouldn't have worked out for the PC alliance blob: it was me and a couple Anglo petty kings attacking from overseas with a total of about 4500 troops versus thousands of warriors from all over Scandinavia. It worked mainly because of AI priorities: occupy or recapture war target territory first, then go after other territories. I already controlled Vestisland and we occupied Austisland before they could hit any of my territory in Britain. That means they came after us in the frozen north instead of countering in the south like I would, and despite the levy numbers and their ability to pay for troops with prestige, my armies were man-for-man much better (higher tech level and a high percentage of heavy infantry and cavalry) and I had strong defensive terrain into which the Norse had to attack from the sea.

Anyway I don't think you can really fault the AI for pressing wars that mathematically it ought to be able to win. We've seen that enough in real life often enough (Vietnam, Iraq, Afghanistan three times).

 

[alanschu]

Anyway I don't think you can really fault the AI for pressing wars that mathematically it ought to be able to win.

One caveat to this is that i'd like to see the AI take into consideration potential joiners in Holy Wars. I'm a Fraticelli heresy and now control the Kingdom of Gujarat (thanks Pope... lol) and I handed it out to some vassal duke viceroys. They are some nearby duchies and counties that are ripe for the picking, but not when 3-4 nearby allies come to help out. My Moral Authority isn't doing so hot with -30%+ due to lost holy wars (it's a big issue in Hungary as well).

 

[Achab]

Player: Dear AI blob, this is a formal DoW for that worthless holding up there.
AI: Okay, raising my full 20k doomstack, sending it your way, die scum.
Player: But you would rise that amount of troops to defend an invasion of your whole realm, you have to scale a bit ....
AI: Oh, yeas, so I have 300 holdings .. but you war me for only one ... so 20k / 300 .... 66.6 troops on the way ... die scum!

 

[Legionary Guard]

Personally, I think that the AI should be willing to surrender at lower warscores than just 100 percent. Perhaps 10 percent or 15 for some random barony out in bumf*ck nowhere, somewhere around 40 percent for a duchy, and 70 or 80 percent for a kingdom. They should also be willing to surrender more easily if the war target is an exclave. Exceptions to this would be A) Exclaves that contain either de jure land or which contain a holy site (For example, Byzantium takes down to Jerusalem before a war or wonky inheritance of whatever cuts Armenia Minor from the Empire. Even though the Levant is now an exclave, it should still be willing to fight a total war to defend Jerusalem and Antioch), B) Jihads/Crusades, especially those which (for empires) target de jure kingdoms (Same scenario: Byzantium takes Jerusalem from the Fatimids before the Seljuks declare a Jihad for Armenia. Byzantium should be able and willing to fight to the death to keep Armenia out of the hands of the Jihad), and C) land that the top liege directly controls. Obviously, it'd need to be scaled to realm size - a 400-holding empire should give up a random barony in bumf*ck nowhere after it loses the first skirmish almost (With perhaps an exception if the primary attacker and primary defender are rivals), but a count should fight to 100 percent warscore to keep that same barony.

Some way around would be to change the way vassal levies works (in defensive wars)
1) you dont raise your vassal levies, you call your vassals to war (similar to tribal mechanic)
2) vassals called that way are "minor participants", they will mobilise small percent of their levies (I think 10% is OK) and give them to you. Generally speaking vassals will not care about that war - they already give you troops, so their obligations are fulfilled.
3) you can call only vassals that are near the contested counties. I would suggest creating geographical area bigger than de iure duchy, something like de iure kingdom, but without drift (lets call it "territory"). You can call vassal only if his capital and any of contested counties is in the same or neighbourning territories.
4) vassals can offer to join war as normal participant. AI will do that only if their land is contested, their ally/friend land is contested, they are your ally/friend, they are aggressor rivals or in exchange for a favor.

This way HRE cannot mobilise all it's forces for some distant county. On the other hand, it still can mobilise when HRE title is at stake (because formally in war for main title every county in realm is contested).
Of course the way you mobilise forces for aggressive wars also needs to be changed. Personally I'd like to see something between Stellaris war declarations (when you are in alliance_ and CKII-style fractions. But I digress.

I also essentially agree with this, with the added caveat that such restrictions be removed for the above-stated scenarios.

 

[A-D.]

I would actually welcome a system like EU has had, where the warscore directly influences just what kind of deal you can make, except its purely about land. If you have a claim on the HRE as per the OP example, then you should be required to get to 100% warscore and then maybe some more like destroying the majority of their army or holding X amount of provinces of the de jure territory (say 25%) to actually press the claim on the HRE. But having lower warscore you can broker a peace for less than that, say the duchy of Saxony, provided your warscore is high enough.

The AI should arguably accept also a lesser trade, for itself at least, because it doesnt have to spend too long at war, which will make the vassals unhappy and deplete gold in the long run, plus wasting manpower leaving it extremely weak to any other power nearby. Additionally it would also encourage a more gradual play without having to literally blob up all the time. Currently the only way to win against an Empire, or an Uber-Empire (2 titles) is to either have one of your own or to cheat (or have insane amounts of money for mercs).

So yeah i'd actually love the "peace brokering" mechanic from EU4 in CK2 as well. Or just generally apply the mechanic from Invasions, where you only take those lands you actually sieged fully if you try to get a peace earlier than at 100%

 

[BeyondExpectation]

I would actually welcome a system like EU has had, where the warscore directly influences just what kind of deal you can make, except its purely about land. If you have a claim on the HRE as per the OP example, then you should be required to get to 100% warscore and then maybe some more like destroying the majority of their army or holding X amount of provinces of the de jure territory (say 25%) to actually press the claim on the HRE. But having lower warscore you can broker a peace for less than that, say the duchy of Saxony, provided your warscore is high enough.

The AI should arguably accept also a lesser trade, for itself at least, because it doesnt have to spend too long at war, which will make the vassals unhappy and deplete gold in the long run, plus wasting manpower leaving it extremely weak to any other power nearby. Additionally it would also encourage a more gradual play without having to literally blob up all the time. Currently the only way to win against an Empire, or an Uber-Empire (2 titles) is to either have one of your own or to cheat (or have insane amounts of money for mercs).

So yeah i'd actually love the "peace brokering" mechanic from EU4 in CK2 as well. Or just generally apply the mechanic from Invasions, where you only take those lands you actually sieged fully if you try to get a peace earlier than at 100%

What I'm really hoping for in CK3 is something like that, but where you can only take territory you had a CB for. E.g. Edward III attacks France with a claim to its throne then gets it to 40% warscore, so negotiates peace in which he get Aquitaine and independence for himself from France (as he was its subject as Duke of Gascony).

Note however, that in EU4, there is basically the same all or nothing problem. In the game, a 1600s war between France and Spain ends when two thirds of Spain is occupied by France and the latter has no money or manpower left, when in RL such a war would be fought almost solely near the forts around the border and in the colonies.

 

[chrnno]

Yesterday I was playing and got one province revolts that I could easily deal with. All my allies immediately raised their full levies and even the closest one couldn't arrive before it was over.

Yes the AI really should be smarter, would be too easy to exploit when they are the target but with the stupid automatic CTA I keep calling all my allies, being called by allies or seeing the AI between themselves where realms all across the world raise their full armies for something that won't last long enough for them to even arrive nearby.

They should also not commit their full might to wars on the other side of Europe that in no way impacts them because either you abuse it and let them to everything or things are over before they get there and thus all that cost in gold, time and opinion is pointless.

 

[keynes2.0]

If the AI were to respond proportionately, you could take large empires one county at a time,

Or the game could be balanced around this new equilibrium... If the player raises every levy in the empire to fight for a county, the player has to spend far more then the county if worth handing out titles and bribes to get those troops. So yeah you could take a county by massively over committing, you just will have weakened yourself.

Empires did actually lose land historically.

What would you prefer in its place, people chipping individuals counties off the HRE, it's only one county what's the point in defending it?

Yes. Exactly that. The power of the HRE was extremely limited. They couldn't just summon an army on a whim to pacify an Italian rebellion. As a result they had to buy the Italians off or lose their control of Italy. Likewise they lost the Hanseatic league, the low countries, southern Denmark...

 

[Aries666]

Or the game could be balanced around this new equilibrium... If the player raises every levy in the empire to fight for a county, the player has to spend far more then the county if worth handing out titles and bribes to get those troops. So yeah you could take a county by massively over committing, you just will have weakened yourself.

Empires did actually lose land historically.



Yes. Exactly that. The power of the HRE was extremely limited. They couldn't just summon an army on a whim to pacify an Italian rebellion. As a result they had to buy the Italians off or lose their control of Italy. Likewise they lost the Hanseatic league, the low countries, southern Denmark...

What about other large empires should they just allow all comers to chip little pieces off them at will because it's not worth defending? This would essentially result in a game where every nation eventually became exactly the same size as large nations wouldn't defend themselves and small nations could grow with impunity.

 

[Gnorrosch]

This calls for crime theory.

Very well then. You only have a certain amount of resources to fight all kind of crime - from jaywalking to murder and from false parking at the local carnival to smuggling schemes from organized crime and terrorism. The more dangerous crimes also require more training, you can easily train someone in three weeks to warn off jaywalkers and tow cars, but you will need dedicated specialists to fight organized crime and terrorism. Now, taken your broken window theory, would you commit 100 % of the police to warning jaywalkers and other low level crimes, completely ignoring any more high level crimes going on? (Because that would be identical to sending 100 % of your imperial army to defend against Count Meagre and his claim on the one holding county of Forsakia).

What about other large empires should they just allow all comers to chip little pieces off them at will because it's not worth defending?

Large empires would still defend their territory - they would just not send their whole army to teach Count Measly and his meagre band a lesson. Instead, they would tell the local duke to take care of the problem. (In reality, they would do that because sending your whole army to one specific area is not very prudent, just as a good commander would always keep some troops in reserve unless he is really desperate.)

 

[keynes2.0]

What about other large empires should they just allow all comers to chip little pieces off them at will because it's not worth defending? This would essentially result in a game where every nation eventually became exactly the same size as large nations wouldn't defend themselves and small nations could grow with impunity.

Where did I say they shouldn't defend themselves? Seriously, where. Go back and point me to it.

They should fight. It's just they shouldn't be able to raise massive armies for free just because they are big. Raising an army from all over an empire is an expensive proposition. That is why empires wouldn't raise huge armies to defend worthless parcels of land. The worthless local parcel of land should only be defended by local forces because it doesn't justify paying to bring in outside forces. If those local forces dont have a hope of defending... the land isn't worth defending!

 

[Aries666]

Very well then. You only have a certain amount of resources to fight all kind of crime - from jaywalking to murder and from false parking at the local carnival to smuggling schemes from organized crime and terrorism. The more dangerous crimes also require more training, you can easily train someone in three weeks to warn off jaywalkers and tow cars, but you will need dedicated specialists to fight organized crime and terrorism. Now, taken your broken window theory, would you commit 100 % of the police to warning jaywalkers and other low level crimes, completely ignoring any more high level crimes going on? (Because that would be identical to sending 100 % of your imperial army to defend against Count Meagre and his claim on the one holding county of Forsakia).

Large empires would still defend their territory - they would just not send their whole army to teach Count Measly and his meagre band a lesson. Instead, they would tell the local duke to take care of the problem. (In reality, they would do that because sending your whole army to one specific area is not very prudent, just as a good commander would always keep some troops in reserve unless he is really desperate.)

Where did I say they shouldn't defend themselves? Seriously, where. Go back and point me to it.

They should fight. It's just they shouldn't be able to raise massive armies for free just because they are big. Raising an army from all over an empire is an expensive proposition. That is why empires wouldn't raise huge armies to defend worthless parcels of land. The worthless local parcel of land should only be defended by local forces because it doesn't justify paying to bring in outside forces. If those local forces dont have a hope of defending... the land isn't worth defending!

In principle I agree it isn't great that the AI gives everything every time. However, we come back to if it doesn't do this how do you determine the limits for the AI's commitment level without inadvertently creating a whole bunch of extra arbitrary cut off that are just as easily exploited by the player. Ultimately 100% commitment is probably the least worst solution to the problem in my opinion.

 

[keynes2.0]

Why arbitrary? Why should there not be penalties for bankrupting your vassals fighting over a barony on the other side of Europe?

 

[chrnno]

In principle I agree it isn't great that the AI gives everything every time. However, we come back to if it doesn't do this how do you determine the limits for the AI's commitment level without inadvertently creating a whole bunch of extra arbitrary cut off that are just as easily exploited by the player. Ultimately 100% commitment is probably the least worst solution to the problem in my opinion.

Start with the simple then. AI shouldn't commit everything unless they are under threat or the resulting peace would put them under threat and there exists the possibility of it actually happening. No more country to fight other's wars on the other side of the world by itself and no more raising all levies to deal with a minor revolt against an ally.

 

[Aries666]

When I say arbitrary what I am getting at is this: you declare war for a county does the AI respond with 40%, 60%, 80% of its power or maybe 120% of your power? You declare war for a duchy what is the commitment in this case 75% and so on through all tiers of CB. In addition as people have already pointed out county A doesn't necessarily have the same value in levy and tax as county B should the AI increase its or decrease its commitment and by how much? All of these arbitrary points create ever increasing points the player can exploit for example if I know the HRE (with 10k troops) will only make a 40% commitment to defend a county I know that I now only need 5-6k to take that county meaning that even if I am a smaller nation I can realistically take chunks off my larger neighbour. As I said 100% commitment isn't perfect it's just better than the alternative.

 

[psychoak]

Now, taken your broken window theory, would you commit 100 % of the police to warning jaywalkers and other low level crimes, completely ignoring any more high level crimes going on? (Because that would be identical to sending 100 % of your imperial army to defend against Count Meagre and his claim on the one holding county of Forsakia).

If jaywalkers shot the police when they were confronted about jaywalking, yes, I would. Others have pointed out how it will be easier to exploit an AI that doesn't go all in. You can reasonably lose a war going in with 2-1 forces because of fail tactics, terrain, etc. Having an AI that responds equally to preserve their forces would result in the AI depleting their forces. Even if they responded with 2-1, large empires would still frequently lose wars that were pin pricks. If the same logic were used in aggression as well, the PC, as the only intelligent entity in the game, would have an unparalleled advantage in expansion and defense. It would be a more logical progression in the AI, but it would be a hilariously easy game.

 

[keynes2.0]

All of these arbitrary points

We already have an arbitrary point! 100% The most arbitrary point of all. The AI will fight 100% for an empire and 100% for a barony.

Yes a system that assigned value to wargoals wouldn't be perfect but it would be an improvement.

 

[StarSword]

We already have an arbitrary point! 100% The most arbitrary point of all. The AI will fight 100% for an empire and 100% for a barony.

Yes a system that assigned value to wargoals wouldn't be perfect but it would be an improvement.

Actually, the AI will sometimes offer peace terms before 100% warscore. Not always, but I've seen it several times.

 

[Aries666]

We already have an arbitrary point! 100% The most arbitrary point of all. The AI will fight 100% for an empire and 100% for a barony.

Yes a system that assigned value to wargoals wouldn't be perfect but it would be an improvement.

Yes one arbitrary point not many and one that I didn't even mention at that. I would be perfectly happy having the sue for peace system that is used in EU4, however, this would be hard to reconcile with the CB concept in CK2 which is that you are fighting for a specific title and only that specific title. Also I don't know why people keep saying that you always need 100% WS because it just isn't true.