Late Game Conquest

[Stalker0]

Thal, could this implement your boldness idea?

Give the AI a great general unit with unlimited range that gives all units some X bonus. After so many Great general points that bonus increases, or they get a second unit that somehow stack with the first bonus.

The effect is that AIs with more GG points (aka boldness, aka the war ones) would get an increasing military bonus with time.

Possible?

 

[Stalker0]

Thal, I had posted this idea in the core forum, but thought if you could code it it might really help the conquest issue:

The secret to war weariness would be to apply it only in enemy territory.

For example, every attack I make in nonfriendly territory applies X amount of weariness. Attack in friendly territory does not apply weariness. Considering the system already can apply bonuses and penalties for attacking in friendly/non-friendly territory, this seems like an easy thing to implement.

So aggressors are penalized while defenders gain an increasing advantage since they are not affected by weariness.


What I'm thinking is that you add a new column (like GP points). Attacks in nonfriendly territory adds X points to that column. Every Y points = 1 unhappiness or -1 science or gold or something. The points decay Z amount per year.

This means defenders are not affected, while attackers slowly decay with time. A human doing extensive conquest has to deal with increasing penalties to maintain the conquest.

 

[MortalD]

Could you be more specific on what 'Added a few changes to help AIs conquer other AIs (no changes to human-vs-AI warfare)' means in the patch notes?

Dunno what changed but in my new game on 132 Egypt got conquered by Spain almost at the start of the game

Now thats interesting!

Update : France declared war on Arabia and took one city. Arabia had ally with City-State and that CS took and puppeted French city of Orleans...

Interesting indeed.

 

[Thalassicus]

I'd like to leave the recent AI-vs-AI changes behind the scenes so it has a minimal influence in our Human-vs-AI decisions. It's a dozen small tweaks in various places across the mod.


Link to video.

Give the AI a great general unit with unlimited range that gives all units some X bonus. After so many Great general points that bonus increases, or they get a second unit that somehow stack with the first bonus.

The effect is that AIs with more GG points (aka boldness, aka the war ones) would get an increasing military bonus with time.

I think this would also affect Human-vs-AI combat? I want to focus solely on AI-vs-AI.

For example, every attack I make in nonfriendly territory applies X amount of weariness. Attack in friendly territory does not apply weariness. Considering the system already can apply bonuses and penalties for attacking in friendly/non-friendly territory, this seems like an easy thing to implement.

I like this idea. It could be a science or gold penalty so warfare slows down the economy like in Civ 4. I think that would be feasible.

 

[Txurce]

I like this idea. It could be a science or gold penalty so warfare slows down the economy like in Civ 4. I think that would be feasible.

War weariness is potentially a powerful and wide-ranging tool to balance the militaristic civs - and possibly make them competitive, if you alter the Domination VC to something the AI can achieve. (Think how huge that would be.) But there has been discussion of WW before, and it's probably a big enough topic to consider broadly, if you want to tackle it.

 

[Jaybe]

I like this idea. It could be a science or gold penalty so warfare slows down the economy like in Civ 4. I think that would be feasible.

Could be a flat or % penalty of income or stockpile (with checks to be sure there IS a stockpile of gold or progress on a tech). Better yet, make it RANDOM for each event so effect is NOT predictable: income/stockpile, applied to science, gold or culture.

For each game (or war within a game), we don't know which would be more critical or appropriate. So spread the effects around unless you apply them to ALL (sci/gold/culture).

If it was a strictly gold penalty it would greatly encourage pillaging to compensate.

--
Speaking of pillaging, there have been vast periods in human history where pillage spoils went to the troops (occasionally so they wouldn't starve), not the governor. Would there be a way to encourage pillaging yet give no gold for it without a particular SP?

 

[Seek]

Speaking of pillaging, there have been vast periods in human history where pillage spoils went to the troops (occasionally so they wouldn't starve), not the governor. Would there be a way to encourage pillaging yet give no gold for it without a particular SP?

That's a good idea, Jaybe! The gold received from pillaging is pretty paltry, I could see changing it to give some healing to the pillaging unit or something similar.

 

[albie_123]

Thal, after finishing my first game with 132, I can safely say that whatever was going on behind that curtain, it was well done

I have never seen an AI do so well in conquest.

The only issue is, as soon as Bismarck (The aforementioned AI) conquered his entire continent, he could not get any further - despite having a tech lead, an absolutely gigantic army, and, after a while, nukes.

He did manage to realise after another hundred turns that he had to declare war on us Civs on the other side of the world, but once he did, he didn't reach us, despite there being 150 turns left in the game.

I really feel strongly about AI being able to win conquest victories. It's simply unfair that this AI, who has worked so damn hard, has less chance of winning than someone who's been turtling the entire game and spends some money on the last turn for a Diplomatic Victory... which is exactly what I did.

I don't know how to fix it without changing the actual domination victory itself. A certain percentage of land and of population, like Civ IV, and maybe 50% of all Capitals, to keep it similar.

EDIT: I should just reiterate that despite my complaints, the aggressive AIs did so, so much better than they normally do, and I have to give absolutely mad props to you, for whatever it was you did.

 

[Thalassicus]

We have limited control over pillaging with our current modding tools. The pillage costs are set per-improvement before the game starts, and I don't think we can change them in an active game.

 

[Txurce]

Thal, after finishing my first game with 132, I can safely say that whatever was going on behind that curtain, it was well done

I have never seen an AI do so well in conquest.

The only issue is, as soon as Bismarck (The aforementioned AI) conquered his entire continent, he could not get any further - despite having a tech lead, an absolutely gigantic army, and, after a while, nukes.

He did manage to realise after another hundred turns that he had to declare war on us Civs on the other side of the world, but once he did, he didn't reach us, despite there being 150 turns left in the game.

I really feel strongly about AI being able to win conquest victories. It's simply unfair that this AI, who has worked so damn hard, has less chance of winning than someone who's been turtling the entire game and spends some money on the last turn for a Diplomatic Victory... which is exactly what I did.

I don't know how to fix it without changing the actual domination victory itself. A certain percentage of land and of population, like Civ IV, and maybe 50% of all Capitals, to keep it similar.

EDIT: I should just reiterate that despite my complaints, the aggressive AIs did so, so much better than they normally do, and I have to give absolutely mad props to you, for whatever it was you did.

+1 on everything, especially on redefining the Domination VC so that in albie's case Germany would probably have won the game.

 

[Jaybe]

It's simply unfair that this AI, who has worked so damn hard, has less chance of winning than someone who's been turtling the entire game and spends some money on the last turn for a Diplomatic Victory... which is exactly what I did.

Once we get access to the core, the timing of who has their turn last before the vote is taken would be RANDOMIZED. So you ...

are notified it's UN Vote turn,
you do your vote manipulating (bribing),
you vote,
and instead of getting a victory screen, the turn may continue, allowing some (possibly all) of the other civs to make their arrangements ...
before the vote is actually taken.

If another civ has the opportunity to bribe more votes (AND they are programmed to take advantage of it), you could lose the game.

 

[albie_123]

I'm confused as to what that has to do with Bismarck losing unfairly when he's crushed half the continent and in control of five Capitals

 

[mitsho]

The problem with for Bismarck in albie_123's game seems to have been naval invasions, right? I guess that's a whole other behemoth to tackle, let's focus on the combat side and the other new things introduced first. But glad to hear good results so far

 

[Thalassicus]

Hopefully the AI will do better naval invasions once it can stack its ships with embarked units in G&K.

 

[orangecape]

I completely agree that redefining domination would be good from a game perspective (not necessarily a RP perspective... not sure on that either way). For example, if there are 8 starting civs and I crush 5 of them I feel like I should win a domination victory. Here is the difficulty though if we allow you to win by only controlling 4 of 8 capitals:

Imagine a world with two large continents, 4 civs per continent. On your continent you go wide and do some building, some conquering, and after killing off 1 civ and crippling the other two you learn Astronomy, send out a caravel.... and get the 'you lose' screen because some AI steamrolled his continent and won on turn 150. Crap.

I think letting people win by killing half the capitals is going to be a real mess and pretty much force a domination strategy on continents maps to avoid losing before you even meet half the players! However, if we made you have to own 5 of 8 capitals then the AI on its 4 civ continent will never win. Either way is a mess.

I don't know how to solve the problem with our current tools. If the domination victory was owning at least 60% of the capitals on pangea then it would be all good. If that victory condition was true on continents though it wouldn't help the AI much (or at all) but would be fantastic for the human.

Until the AI is able to wage even a slightly competent (enough to leverage a 12 teach lead into beating people on other continents, say) then I think the non pangea maps are just going to present insurmountable problems to making actually collecting the domination victory simpler. Even if you replace capitals with land area, population, etc. the AI has to either be allowed to win by owning its continent or not and either way is very problematic.

 

[albie_123]

Hopefully the AI will do better naval invasions once it can stack its ships with embarked units in G&K.

They'll have to actually change the AI - they don't even seem to send units over.

 

[Txurce]

I completely agree that redefining domination would be good from a game perspective (not necessarily a RP perspective... not sure on that either way)...

Even if you replace capitals with land area, population, etc. the AI has to either be allowed to win by owning its continent or not and either way is very problematic.

I would redefine a Domination victory by making it x% of the total land mass; plus x% of total population; plus 3 capitals. As such, it's about building a big enough empire on either a big enough continent, or a continent plus some island expansion (which the AI will do after t200 or so). That eliminates cheap victories for the AI. It also reduces their likelihood of achieving it, but any % is a lot better than zero.

If there is a downside in such a shift, for me it would be that Conquest becomes to easyfor the human. But there's no downside to trying it out.

 

[Thalassicus]

I think letting people win by killing half the capitals is going to be a real mess

I believe the domination requirement in Civ 4 was 66% of population and land area, not 50%... but it was years ago and my memory is hazy. I do remember it depended on map size.

 

[albie_123]

Yeah, it certainly wasn't 50%. 50% would be a real mess, since that's usually just one continent. I don't think 50% of Capitals as an extra is that bad though. It means that simply expanding peacefully in a big way doesn't result in a domination victory.

 

[wobuffet]

I like 51% of Capitals (must retain own capital & conquer half of the others') and 66% of land mass.