Human player goes heavily in debt, wins the war and then makes his enemies pay his debt.If a human player considers this as a factor, why shouldn't the AI?
Maybe I'm wrong on this, but I think it's a small change for the better.
- 1
Human player goes heavily in debt, wins the war and then makes his enemies pay his debt.If a human player considers this as a factor, why shouldn't the AI?
Maybe I'm wrong on this, but I think it's a small change for the better.
Well they took loans probably...Just checked If can call Ottomans to war, nope they are over 40k in debt. I cant imagine how ai can get this much debt? Especially stronk empire like Otto. Any ideas?
A player loosing a war might try to peace out as early as possible, if he is willing to give up on what the AI demands. Player assumes agency and largely determines when and to which conditions the peace deal will take place. That also means a player winning a war will not let the AI out until total victory and maximim warscore is reached. While player assumes all agency, AI is heavily confined and will always be at the mercy of the player to accept peace deals.Pretty easy to rack up that debt. The Ottomans have a large income so each individual loan will be bigger, if they have taken economic they will get the national bank event which could allow them to take loans at 24x monthly income so assuming a standard Ottoblob income of 150 ducats makes for 3600 loans. If they are then losing wars and shelling out on mercs with large chunks of their country occupied 11 loans doesn't seem like that much.
Now of course a player rather than go that heavily into debt would try and get out of a war quickly for minimal concessions if it was clear they were going to lose. However, the AI probably can't see when a war is essentially lost and will go balls to the wall in a futile effort. So really what we are talking about here is getting the AI to clear its debts more efficiently whilst at the same time trying to avoid getting declared on. Humans do this by firing advisors, keeping a large standing army but at 0 maintenance, mothballing their heavy fleet (if they have one), mothballing their forts and making alliances for protection, I can't say what the AI does but its not very good at clearing debt.
True but an Ottoman 40k debt wont be the result of one lost war with the player, this would be the result of some pretty heavy dogpiling initiated either by the player or the Commonwealth.A player loosing a war might try to peace out as early as possible, if he is willing to give up on what the AI demands. Player assumes agency and largely determines when and to which conditions the peace deal will take place. That also means a player winning a war will not let the AI out until total victory and maximim warscore is reached. While player assumes all agency, AI is heavily confined and will always be at the mercy of the player to accept peace deals.
Humans do this by firing advisors, keeping a large standing army but at 0 maintenance, mothballing their heavy fleet (if they have one), mothballing their forts and making alliances for protection, I can't say what the AI does but its not very good at clearing debt.
Sorry I didn't mean to be overly critical I was mostly interested in what measures the AI takes to get out of a debt spiral. One pertinent one with the Ottomans, in my current game I saw them at 15k of debt, would be are they firing advisors? Their monthly income would certainly allow them to hire lvl 2-3 but this is the easiest measure to take to clear a deficit. Of course an issue here would be an AI falling behind in MP gain particularly in mil which would probably leave them more exposed. I can certainly sympathise trying to come up with a set of logical rules to follow with so many variables and a heavy dose of intuition requiredThe primary problem is the AI can't sufficiently assess how far in the chicken race of not keeping an army/navy/forts it could go. A human can assume his alliance will reliably protect him, that he will grow enough to pay a greater debt than he could repay at current income, or that he will simply be lucky to avoid a war (should he be incorrect, he can probably restart the game). Another fairly big problem is that the AI absolutely will keep spending troops (potentially going over force limit to fight a losing war, again it boils down to incredibly complex evaluations that humans learn by experience but that can't be hand coded into computer logic in any reasonable amount of time.
I worked specifically on its repayment logic and it's way better than in 1.14 (in my subjective opinion) where lots of AI countries would be completely disabled by debt. You can always dig up *some* case during a campaign where an AI country dies due to debt, but that's not a bad thing (in fact the opposite for flavor reasons) as long as it's sufficiently rare. Where I should still attempt to tweak further is how AI spends its money during wars and how much to avoid racking up debt for poor reasons, as there is better use of time than repaying debt for lost wars. However, it's extremely easy to get it wrong and there are already frequent enough "AI gave up too easily/without a fight" complaints.
The primary problem is the AI can't sufficiently assess how far in the chicken race of not keeping an army/navy/forts it could go. A human can assume his alliance will reliably protect him, that he will grow enough to pay a greater debt than he could repay at current income, or that he will simply be lucky to avoid a war (should he be incorrect, he can probably restart the game). Another fairly big problem is that the AI absolutely will keep spending troops (potentially going over force limit to fight a losing war, again it boils down to incredibly complex evaluations that humans learn by experience but that can't be hand coded into computer logic in any reasonable amount of time.
I worked specifically on its repayment logic and it's way better than in 1.14 (in my subjective opinion) where lots of AI countries would be completely disabled by debt. You can always dig up *some* case during a campaign where an AI country dies due to debt, but that's not a bad thing (in fact the opposite for flavor reasons) as long as it's sufficiently rare. Where I should still attempt to tweak further is how AI spends its money during wars and how much to avoid racking up debt for poor reasons, as there is better use of time than repaying debt for lost wars. However, it's extremely easy to get it wrong and there are already frequent enough "AI gave up too easily/without a fight" complaints.
The biggest issue I have is not the utterly-crippling mega-debts, but the medium-size debts that AI goes into pretty much 3/4 of the time. Such as 2k debt for France, or 0.5k debt for a HRE minor.The primary problem is the AI can't sufficiently assess how far in the chicken race of not keeping an army/navy/forts it could go. A human can assume his alliance will reliably protect him, that he will grow enough to pay a greater debt than he could repay at current income, or that he will simply be lucky to avoid a war (should he be incorrect, he can probably restart the game). Another fairly big problem is that the AI absolutely will keep spending troops (potentially going over force limit to fight a losing war, again it boils down to incredibly complex evaluations that humans learn by experience but that can't be hand coded into computer logic in any reasonable amount of time.
I worked specifically on its repayment logic and it's way better than in 1.14 (in my subjective opinion) where lots of AI countries would be completely disabled by debt. You can always dig up *some* case during a campaign where an AI country dies due to debt, but that's not a bad thing (in fact the opposite for flavor reasons) as long as it's sufficiently rare. Where I should still attempt to tweak further is how AI spends its money during wars and how much to avoid racking up debt for poor reasons, as there is better use of time than repaying debt for lost wars. However, it's extremely easy to get it wrong and there are already frequent enough "AI gave up too easily/without a fight" complaints.
I clear debt by going to the ledger, seeing who has the most money, and going to war.
Would be nice for AI to send as much of it's forces as proportional to what it wants from the war and make favors paid for aid in wars scale with the amount of aid you ask for.
That works for a human but again, AI doesn't automatically reason about such relationships. If the programmer didn't hand code it in, the AI won't see it. Before you say "add it", please think about the possible responses I might make.
The AI does not have much of an intent with wars. It's more like "how much do I want this war; is it above some threshold, and is it the most wanted war?" It does not aim to achieve a specific result when starting it (with few exceptions).
Would it be possible to allow the AI to limit it's involvement though?The AI does not have much of an intent with wars. It's more like "how much do I want this war; is it above some threshold, and is it the most wanted war?" It does not aim to achieve a specific result when starting it (with few exceptions). It categorizes wars into two types: "are we comfortably winning?" or "we need to put everything into winning!", the latter is probably what players are complaining about.
You would struggle to find any post where I advocate improving the AI because...skynet.
Would it be possible to allow the AI to limit it's involvement though?
The AI goes balls to the wall every war because it has no concept of a limited war or a war where there is little scope for territorial gain. As a consequence the AI has to be programmed to go into total war mode in every war to ensure that it can win a war full stop.Well doesn't the AI tend to overcommit with Mercenaries and such becuase it assumes any war it fights must be a total war? I noticed France and the Ottomans love to spam Mercenaries like crazy if they happened to lose a battle or two at start, and the proceed to sit on said Mercenaries for years even if the tide turns. I suppose its like that becuase of Multiplayer concerns but its a bit silly.
1. If they managed to defeat one or more of your allies, that should really give you extra war exhaustion / premature "call for peace" (PDX should add this).I played as the Mamluks yesterday. I managed to conquer minor nations in anatolia before ottomans could do it. Ottomans won defensive war against Venice and their allies. According to ledger ottomans had 31k troops and around 8k manpower. I had 35k troops and max 31k manpower so I decided its right time to strike against ottomans together with poland,lithuania and moldavia. War went quite good...I defeated ottomans in anatolia and drained all their manpower.After this ottomans build ridiculous ammount of mercenaries and they managed to defeat poland and sign white peace with them so I stayed against ottomans alone. Their mercenari army killed 30k of my troops and I had to hire mercenaries too. Huge revolts started in balkan and their loans started to mount. Ottomans ignored this and they send more mercenaries against me. When I destroyed their last 25k stack they finally agreed with my peace offer.
Ottomans are effectively out of game now. All their manpower is drained, they have huuuge loans, ankara and some provinces around lost to mamluks, rebels liberate bulgaria, lot of rebels in greece. I dont understand why AI keep fighting in lost wars and fuck up whole country forever.
Edit: ottomans were 10k over limit all the time