AI Surrender Conditions?

    This site uses cookies. By continuing to browse this site, you are agreeing to our Cookie Policy.

    • AI Surrender Conditions?

      My apologies if I have managed to not find the existing thread & answer with the Search tool. I really did try but all I found were threads of folks trash talking each other into surrendering and the like.

      What I am trying to figure out is what needs to happen for the AI of a given country to surrender.

      The range of options (in decreasing order of rationality and historical realism) seem to be:

      1) Capture and hold the enemy capital for x number of days.
      2) Capture and hold some number to all of the enemy cities (core provinces) for x number of days.
      3) Capture and hold some number to all of the enemy core provinces for x number of days.
      4) Capture and hold all provinces the enemy was holding for x number of days.
      5) Hunt down and destroy every last living, breathing unit and drive a stake through their heart.

      Inquiring minds would like to know as I put in motion my plan to rule the world!

      Thank you in advance for your time and patience in dealing with what is certainly a noob question that has an answer obviously posted where I should have looked.
    • I'm going to follow up on this as I learn the answers the hard way.


      Capturing a Capital has minimal effect on the enemy. You gain some portion of loot but they still fight on. This means you essentially have to eradicate the enemy forces where ever you find them.

      From reading elsewhere on the forums, it has become apparent that the Naval and Air forces continue to do battle also with magic supplies delivered to them like manna from heaven.

      Unless and until they change this game mechanic, be prepared to wage a war of total destruction. Not very realistic, IMHO but it is what it is.
    • The enemy AI forces must be totally eradicated, realistic? no. Easy to code? yes.
      Even worse with live players: a totally eradicated country (with no remaining forces) can be 'gifted' a province and sufficient resources to build more forces, and suddenly become 'alive' and attack from a completely different location.
    • Perhaps, with time, the programmers will add code that allow for the various historical results ranging from the Alsace - Lorraine swapping between France and Germany, the Vichy French situation, the partition of Poland and finally to the unconditional war on Germany and Japan.

      Those four examples have fairly cleanly defined thresholds and results.

      Each of those cases would provide a different end result than our current situation.

      As threshold conditions are met, the AI can offer surrender terms. If accepted, peace breaks out with the new borders and arrangements. If not accepted, some period of time would have to pass or a new set of conditions met if order for surrender terms to be offered again.

      If the developers need someone to toss together a flow chart for this, I am certain the community here can come up with some good choices with minimal complexity.
    • F. Marion wrote:

      Perhaps, with time, the programmers will add code that allow for the various historical results ranging from the Alsace - Lorraine swapping between France and Germany, the Vichy French situation, the partition of Poland and finally to the unconditional war on Germany and Japan.

      Those four examples have fairly cleanly defined thresholds and results.

      Each of those cases would provide a different end result than our current situation.

      As threshold conditions are met, the AI can offer surrender terms. If accepted, peace breaks out with the new borders and arrangements. If not accepted, some period of time would have to pass or a new set of conditions met if order for surrender terms to be offered again.

      If the developers need someone to toss together a flow chart for this, I am certain the community here can come up with some good choices with minimal complexity.
      I like the idea as it adds dynamics to the game,but I can only imagine the amount of coding this would take as you would have to set all kinds of parameters to make the AI offer conditions or allow you to offer conditions.

      "Cry 'Havoc!', and let slip the dogs of war"




      "The best weapon against an enemy is another enemy."Friedrich Nietzsche
    • purplepizza117 wrote:

      F. Marion wrote:

      the Alsace - Lorraine swapping between France and Germany
      As far as I know you can achieve this by trading provinces between players, but I assume you mean for war consequences?

      Yes. As you know, the Alsace - Lorraine swap(s) were a result of winning / losing wars. For as far back as european history goes, it has always been about the winner getting land from the loser. If we attempt to replicate that in the game, it would entail the AI having a mechanism to agree to a land transfer in return for peace.

      The real challenge to this whole idea is making sure the AI has a metric for whether or not it is winning or losing the war.

      I am guessing that there is _something_ in the AI code but I have not seen enough of the AI actually doing anything to guess what that something might be.


      As a simple metric, I would have it look at things I know it is already calculating. Or at least are being calculated for players...

      Trend in number of provinces, number of VPs, and positive / negative flows for goods are all good proxy indicators of victory or defeat. Certainly all those things could be in the red right before a grand victory or in the green just prior to a crushing loss but in general they make good proxies. We can worry about the edge cases later, after the first order code is done.

      I don't deny this is a significant coding effort but it is certainly in the realm of possible. Even better, it does not muck with existing code. Pulling data from variables that already exist in the slate is a pretty trivial thing. The output is a a fairly straight forward range of responses.

      Winning = SurrenderOffer is false
      Tie = SurrenderOffer is 1 return to status quo
      Losing = SurrenderOffer is 2 land/goods for peace
      Disaster = SurrenderOffer is 3 or Vichy France style of puppet government

      Implementing that output is not trivial. I don't know how complex or simple the AI is. Getting the military forces of the NonPlayerCountry to behave correctly for this will be a challenge. Especially since, at the moment, there does not even seem to be a way to tell them they have lost the war and should surrender.

      Anyway, I'm sure the programmers would love to tackle this if there were enough time/money in the budget. It is just too much fun to not want to code this. As we can tell from the fact that I am writing about it instead of writing code for my employer. Gotta go...
    • Well, AI part of this game is almost completely unsignificant. Only time you really have to worry about an AI is first few days if you want to kill some small neighbour and try to take its production center and for that, you will just go for total destruction. In mid and late game, what you are worried about are human players and wars against them. Yea, there could be some standard or inactive human AI present, but if you want to take it, just do a few plane sweeps, take what you need, make peace or finish it and continue dealing with human enemies.

      So, implementing AI surrender, or any other kind of advanced AI behaviour makes little sense since since AI is already unimportant part of the game. This is a multiplayer game, main focus is PVP. AI is there just to fill the void.
    • I can see that in a more active game (more active players) the AI is not as important.

      I am only in one game and it is my first one. As such it seems to have spots for 45+ players. About 15 were occupied by the end of the beginning. I had one shared border with another player. Two shared borders with AI.

      Today, after the inactive player time outs, we have 4 players left. None of us are adjacent. Essentially one in North America, one in Africa, one in Central Europe and one in Scandinavia. We won't be able to even attack each other for another 4 days. That is just travel time. Not counting the number of AI combatants between us.

      Dunno what we will do. Effectively the game should likely be abandoned but that is rather disheartening. At present, I will consolidate as many of the AI countries around me as I can. It is good practice for the mechanics of making things work. Perhaps the next game will be with a more active (serious) group of players.
    • F. Marion wrote:

      The real challenge to this whole idea is making sure the AI has a metric for whether or not it is winning or losing the war.

      I am guessing that there is _something_ in the AI code but I have not seen enough of the AI actually doing anything to guess what that something might be.
      Actually, the "Elite" AI is a little smarter in dealing with human players. I don't think it has the kind of considerations you are suggesting. And you suggested elsewhere the difficulties of the AI code could be heavy. Frankly, I'd like to get ahold of the AI code myself and totally re-work it's strategic factoring. But I don't work for Bytro.

      F. Marion wrote:

      Anyway, I'm sure the programmers would love to tackle this if there were enough time/money in the budget. It is just too much fun to not want to code this. As we can tell from the fact that I am writing about it instead of writing code for my employer. Gotta go...
      Curious...what kind of coding do you do?

      F. Marion wrote:

      I can see that in a more active game (more active players) the AI is not as important.

      I am only in one game and it is my first one. As such it seems to have spots for 45+ players. About 15 were occupied by the end of the beginning. I had one shared border with another player. Two shared borders with AI.

      Today, after the inactive player time outs, we have 4 players left. None of us are adjacent. Essentially one in North America, one in Africa, one in Central Europe and one in Scandinavia. We won't be able to even attack each other for another 4 days. That is just travel time. Not counting the number of AI combatants between us.

      Dunno what we will do. Effectively the game should likely be abandoned but that is rather disheartening. At present, I will consolidate as many of the AI countries around me as I can. It is good practice for the mechanics of making things work. Perhaps the next game will be with a more active (serious) group of players.

      A sad reality is that many players abandon after the first day or two. If you really like the game and desire to see people staying in action, you might consider joining an active alliance or even the player's League contest Player's League - Sign up for October

      Another thing you can do...if you have time and to alleviate the boredom of waiting for things to happen, is to join more matches at the same time. A word of advice, though, you will want to have only a few very active matches at most because it can snowball on you when your nations get to become huge empires...assuming you make it that far.
      It seemed like such a waste to destroy an entire battle station just to eliminate one man. But Charlie knew that it was the only way to ensure the absolute and total destruction of Quasi-duck, once and for all.

      The saying, "beating them into submission until payday", is just golden...pun intended.

      R.I.P. Snickers <3