Overuse of Nukes

[Yakk]

I disagree. The most powerful nuke is the one not used.

I disagree. Fleet in Being has lots of strategic use, but Fleet in Being is not always the most strategically powerful use of a Fleet.

You will note that the AI tells you if it has a nuke, doesn't it? So it had already said the threat "if you go to war with me, I will nuke you". You went to war with the AI regardless.

And Civ4 lacks a concept of "limited war". I can imagine how to code it... but it would be quite hard.

Limited Wars:
Border Skirmish: Your units may not enter into squares where you lack at least 5% culture.
Conventional War: No nuclear or chemical weapons are used. (WMDs)
Proxy War: Units may not enter into each others cultural territory.

You could see how the above wars would mirror the modern world. A proxy conventional war could allow two sides to fight over the territory of a vassal. A Border Skirmish would let you fight over disputed territory.

...

More simply, the UN resolution that bans building nuclear weapons could also ban using them. If you use nuclear weapons in a war, an automatic "expel and go to war with the user of nuclear weapons" resolution pops up (requires 50% support), as does a "open nuclear season on the nuker" resolution (2/3 support).

 

[Dom Pedro II]

I disagree. Fleet in Being has lots of strategic use, but Fleet in Being is not always the most strategically powerful use of a Fleet.

You will note that the AI tells you if it has a nuke, doesn't it? So it had already said the threat "if you go to war with me, I will nuke you". You went to war with the AI regardless.

True, but in many of these cases, it's in the AI's best interest to end the war as quickly as possible, so it couldn't hurt for them to give an ultimatum before they start shooting. The only reason why I don't think this would work though is because the human player doesn't get the diplo screen until the start of his turn. So unless the it was coded so that the AI's ICBMs fired out-of-turn upon refusal by the human player, the AI would have to wait until the next turn, which would give the human opportunity prep.

And Civ4 lacks a concept of "limited war". I can imagine how to code it... but it would be quite hard.

Limited Wars:
Border Skirmish: Your units may not enter into squares where you lack at least 5% culture.
Conventional War: No nuclear or chemical weapons are used. (WMDs)
Proxy War: Units may not enter into each others cultural territory.

You could see how the above wars would mirror the modern world. A proxy conventional war could allow two sides to fight over the territory of a vassal. A Border Skirmish would let you fight over disputed territory.

Well, these wouldn't merely give the AI a concept of limited war, they would actually bind the players to fighting in certain ways.

More simply, the UN resolution that bans building nuclear weapons could also ban using them. If you use nuclear weapons in a war, an automatic "expel and go to war with the user of nuclear weapons" resolution pops up (requires 50% support), as does a "open nuclear season on the nuker" resolution (2/3 support).

Yeah, that's an idea too.

 

[Dom Pedro II]

Strategic strikes like that are what I would really like to see the AI do.

I agree. I have no problem with the AI starting nuclear war. I have a problem with the AI starting it when there is no real benefit to do so.

 

[Yakk]

Well, these wouldn't merely give the AI a concept of limited war, they would actually bind the players to fighting in certain ways.

There is the states of "peace" and "war" right now, with various rules and in-game consequences to moving between them.

It may be easy and consequence-free to move between the war/limited/proxy/conventional war.

Limited war, for example, could start by the middle ages. Going to limited war with someone doesn't generate the same 'hatred' diplomatic hits that going to real war does.

A UN resolution that enforced going to full war by a turn of anarchy and immobile units might make going to full war less practical. The real problem is the time scales involved (Civ4 turns are a full year).

The states are just an attempt to describe what actions are permitted -- what kind of ways to move between states is not something I'm at all certain of.

But I would love to be able to engage in proxy limited conventional warfare as a pair of superpowers in civ4. (This happened in the real world)

...

The AI was in a no-win situation already. So it tried to hurt the empire that doomed it as much as it could.

 

[MartinHarper]

Since ICBMs aren't destroyed when the cities they are in are nuked, the prisoner's dilemma is not appropriate.

It is still appropriate. Once the invasion has started, the enemy's forces will be on Incan territory. This makes them harder to nuke, and means that the infrastructure damage will be done to the Incans. For their nukes to have maximum effect, the Incans need to use them when they can do maximum damage, which means on the first turn of the war.

We have two possible circumstances.. a moral human player (one who will not nuke first) or an immoral human player (one who WILL nuke first).

How is turning an entire continent into a charred wasteland a moral act?
Regardless, here's a third circumstance: an attacking civ that will not use its nukes, even in response to being nuked. This might be because of losing nuclear deterrence, or:

2) War weariness spikes, 3) Diplomatic penalties with civs that might potentially enter the war, and 4) Global Warming.

Here's a fourth circumstance: a human player that will learn from his mistakes. The human destroys the Incans this game, but he learns that the Incans are willing to use nuclear weapons if someone declares war on them, even at the cost of their own destruction. Thus, the next time the human is in a game with a nuclear-armed Inca, the human does not declare war on them.

The above two circumstances assume that both the AI and the human will behave in their rational self-interest. However, to maximise the deterrent effect of a nuclear arsenal, a civilisation must occasionally behave irrationally. This is a standard result of game theory:
http://hnn.us/articles/17183.html

CivFanCSS used his entire nuclear arsenal to destroy the Incans. This was massive overkill, and probably caused him significant issues with war weariness and global warming. He was unable to build replacement nukes, and so left himself dramatically weakened against third parties. He probably could have capitulated the Incans after the first dozen nukes and saved the rest for someone else. If we want to improve the AI's game play, it needs to aspire to this level of irrationality.

 

[Dom Pedro II]

It is still appropriate. Once the invasion has started, the enemy's forces will be on Incan territory. This makes them harder to nuke, and means that the infrastructure damage will be done to the Incans. For their nukes to have maximum effect, the Incans need to use them when they can do maximum damage, which means on the first turn of the war.

Well, but you're talking about degrees then. The prisoner's dilemma is an all or nothing circumstance. The comparison is still not appropriate.

How is turning an entire continent into a charred wasteland a moral act?

I think you may have missed my point. By "moral" and "immoral" player, I was simply designating those who would and those who wouldn't use nuclear weapons first. In that sense there cannot be a third or fourth circumstance because its either one or the other.

Here's a fourth circumstance: a human player that will learn from his mistakes. The human destroys the Incans this game, but he learns that the Incans are willing to use nuclear weapons if someone declares war on them, even at the cost of their own destruction. Thus, the next time the human is in a game with a nuclear-armed Inca, the human does not declare war on them.

You seem to be making the false assumption that I think CivFanCSS got a raw deal and that I am arguing against the AI going straight for nuclear first strike for the sake of the player. I am not making this argument. What I am saying is that the logical consequence of using nuclear weapons first when you're up against someone with a vastly larger nuclear arsenal is extermination.

The above two circumstances assume that both the AI and the human will behave in their rational self-interest. However, to maximise the deterrent effect of a nuclear arsenal, a civilisation must occasionally behave irrationally. This is a standard result of game theory:
http://hnn.us/articles/17183.html

Of course, once they've exhausted their nukes, there is no deterrent. Acting like a madman makes people think you'll launch the nukes, and that makes them avoid attacking you. But once the bombs are dropped and there cannot possibly more, you're just going to be left with a really, really pissed off neighbor filled with a terrible resolve. Committing an unspeakable act doesn't deter people. Convincing them that there could be more does. Sure, the player will learn a lesson, but is that what the AI's for? To teach the player a lesson? Or is intended to be another player with its own agenda and goals?

CivFanCSS used his entire nuclear arsenal to destroy the Incans. This was massive overkill, and probably caused him significant issues with war weariness and global warming. He was unable to build replacement nukes, and so left himself dramatically weakened against third parties. He probably could have capitulated the Incans after the first dozen nukes and saved the rest for someone else. If we want to improve the AI's game play, it needs to aspire to this level of irrationality.

It needs to aspire to CivFanCSS's level of irrationality? It seems to me like the AI already does pretty dumb things as it is without having to code into irrational behavior.

 

[MartinHarper]

Once the bombs are dropped and there cannot possibly more, you're just going to be left with a really, really pissed off neighbor filled with a terrible resolve.

Actually launching nukes is, in many cases, irrational. As the link demonstrates, irrationality is a positive attribute in nuclear diplomacy. The AI can't pretend to be irrational, because a human player will know whether it is bluffing based on prior experience. Therefore the AI has to be genuinely irrational.

Consider: Ghandi says "if you declare war on me, I will nuke you". The player thinks this is a bluff and declares war. Ghandi doesn't use his nukes, as that would be irrational. Instead he says "if you invade me, I will nuke you". The player thinks this is a bluff, and invades. Ghandi doesn't use his nukes, as that would be irrational. Instead it says "if you capture a city, I will nuke you". The player thinks this is a bluff, and captures all the Ghandi's cities except one, using conventional forces. Ghandi relocates its nukes to his capital, and doesn't use them, as that would be irrational. Instead he says "if you attack my capital, I will nuke you". The player nukes the capital and takes it in a single turn.

Consider: Montezuma says "if you declare war on me, I will nuke you". The player thinks this is a bluff and declares war. Montezuma uses all his nukes the very next turn to causes maximum damage, even though this is irrational. The player gets cross, nukes the entire civilisation, trashes his diplomatic reputation, and conquers a charred and worthless continent.

Given the choice, and knowing about the different strategies of these two leaders, a player would prefer to attack Ghandi. This means that Montezuma's strategy is better at deterring nuclear attack.

Is that what the AI's for? To teach the player a lesson? Or is intended to be another player with its own agenda and goals?

If the AI is intended to simulate a human player, then "take a final pointless strike at the civilisation that doomed me" is a very human reaction. Revenge is an entirely reasonable goal, even if it is irrational from the perspective of winning the game or achieving a good score.

 

[PieceOfMind]

Actually launching nukes is, in many cases, irrational. As the link demonstrates, irrationality is a positive attribute in nuclear diplomacy. The AI can't pretend to be irrational, because a human player will know whether it is bluffing based on prior experience. Therefore the AI has to be genuinely irrational.

Consider: Ghandi says "if you declare war on me, I will nuke you". The player thinks this is a bluff and declares war. Ghandi doesn't use his nukes, as that would be irrational. Instead he says "if you invade me, I will nuke you". The player thinks this is a bluff, and invades. Ghandi doesn't use his nukes, as that would be irrational. Instead it says "if you capture a city, I will nuke you". The player thinks this is a bluff, and captures all the Ghandi's cities except one, using conventional forces. Ghandi relocates its nukes to his capital, and doesn't use them, as that would be irrational. Instead he says "if you attack my capital, I will nuke you". The player nukes the capital and takes it in a single turn.

Consider: Montezuma says "if you declare war on me, I will nuke you". The player thinks this is a bluff and declares war. Montezuma uses all his nukes the very next turn to causes maximum damage, even though this is irrational. The player gets cross, nukes the entire civilisation, trashes his diplomatic reputation, and conquers a charred and worthless continent.

Given the choice, and knowing about the different strategies of these two leaders, a player would prefer to attack Ghandi. This means that Montezuma's strategy is better at deterring nuclear attack.



If the AI is intended to simulate a human player, then "take a final pointless strike at the civilisation that doomed me" is a very human reaction. Revenge is an entirely reasonable goal, even if it is irrational from the perspective of winning the game or achieving a good score.

You seem to ignore the fact that to deliver an ultimatum and not follow up with the consequences, that is in fact the most irrational. In your example, Gandhi is the preferred target by the human player but not because he knows Gandhi won't use his nukes. It's because he knows Gandhi's threats mean nothing. If Gandhi were to make a threat like that at the start of the war only some of the time, he'd still be the slightly preferred target but at least now his actions would be more rational. If a human player wishes to take advantage of Gandhi's preference for peace, then so be it - that's part of the roleplaying in the game.

If an AI is going to make an ultimatum, it has to follow through. The thing you then have to examine (if you so wish) to be irrational or not is the decision to make that ultimatum or not. If Gandhi does not intend to use his nukes then he should not have threatened to use them.

As far as I can tell, your argument has nothing to do with the "madman theory" discussed in the linked article.

Anyway, I think it's all getting a bit convoluted and over most of our heads. I really don't think the way the AI has to treat nukes is anywhere near as complicated as some are making it sound here.

Building a bit of irrationality into the AI regarding nukes will almost inevitably mean linking it to the RNG in some way. Otherwise, it is going to be predictable by a human player and hence exploitable. Perhaps an AI should have a very small chance of initiating nuclear war, especially if it is in a defensive war. Still, the nukes should only be used against strategically important targets because it is going to hurt them diplomatically. Every AI should IMO have a very high chance of using nukes in a retaliatory manner - tit for tat.

 

[r_rolo1]

If the AI is intended to simulate a human player, then "take a final pointless strike at the civilisation that doomed me" is a very human reaction. Revenge is an entirely reasonable goal, even if it is irrational from the perspective of winning the game or achieving a good score.

And again we are in this point.....

It is desirable to have a inteligent oponent in a game, but most of the times inteligent != human behaviour or even inteligent != rational ( rationality is very time and resource consuming, especially in complex enviroments. Most of the times it is better a bad decision now than a rationally good decitsion too late in RL situations ). But I digress....

As I already said more than once, the objective of this mod is to make a AI that plays this game better ( hence the name of the mod ), and not make a AI that plays like a human ( and even then we would have to discern what human are we talking about ( the classic DaveMcW vs obsolete showdown in Strat and Tips forum ( need to find the links ) is a clear example that two humans can aproach and win the game using completely diferent strategical aproaches to the same map ), in what map, against who ( humans play completely diferently in SP and MP, and even in MP they tend to behave diferently in FFA than in Pitboss or PBEM ), the roleplay factor ... ). So, IMHO making human = inteligent is a bad idea to start with, atleast for this mod...

Now on topic:

@Martinharper

The comparison you are making is not how ti goes, because we aren't talking of a zero-sum game. Monty will not win anything by being destroyed ( well, neither will gandhi ), so equalizing player dislike for a AI strategy = good strategy for the AI is clearly not appliable here ( and you forgot that not using the nukes at any point of your gandhi story might be a irrational move compared with the alternative ( in the end it is always a matter of comparison ) ). It also assumes that the human will go berserk and nuke with all it's might always... you know, there are people with a enough cold head to not use more than apropriate force ( especially in a strategy game ). Just because a person has nukes, it doesn't mean that it has to enter in MAD mode automatically ( MAD only is useful when the 2 parties in question have roughly the same punch power )

@ DP II ( btw, a question I've wanted to do for a very long time : any Brazilian connection? )

Benefit is always a very relative thing. You said, and with a lot of reason, that a AI should not start a nuke war without any benefit, but i strongly suspect that our definitions of benefit are somewhat diferent In the way I see it, and inside the discussion we are making ( AI player inferior both in nuke and conventional means to a attacker with unknown nuclear strategy in a world where you can't build more ), in this situation survivavility is the more important variable (there is no good in following a strategy that will make me die earlier, no matter how "rational" it might look at first sight ), and in that aspect gaining some extra turns as a charred wasteland ruler trumps easily letting your empire being eaten by your enemy conventional troops ( suposing you can do a nuke strike that wipes out a lot of the enemy troops OFC, as it looks to be the case in here ), barring the possibility of surrender ( in there OFC not provoking the enemy is a good strategy, but if you intend to surrender, why not do it without firing a shot, conventional or not, in the first place? )

My point is ( speaking like if I was in the AI place ) : "Someone declared war at me. I'm inferior both in nuke terms ( neither me or the enemy can do more ) as in standing army terms, so , if they put things at full speed even if only in conventional terms I'm doomed no matter what I do ( and I don't have any reason to believe they won't put things at full speed either in nuke or non nuke terms ). So what I'm left is to survive the longest possible and wait for something good comes from a exogenous source ( enemy gets too much protesters at home, UN or the Pope gets in the way and forces peace, enemy gets bored, aliens invade the planet, whatever... ), a thing that I can't do if I'm dead. Now I have a possibility of using nukes to strike a blow to the enemy that can delay my demise in some turns ( this is a important part ... IMHO it applies to the game we are talking, since they blew with a massive ( and, by the description, only in the area ) naval fleet, that would take a while to get back ), but I risk a massive nuke blowout if I do that... So, should I nuke or not?

Well, the fact is, like I said, I am doomed no matter what, barring some almost divine intervention. OFC there the possibility of being nuked no matter what I do , or my enemy can only nnuke me in response, or might even not nuke me reagrdless what I do ,to keep his nuke power intact ( as he has more conventional troops he might decide that they are enough ). In every case , if there is a possibility of dealing a good punch to the enemy that can give me some more turns I gain more time and I do prefer to be the living ruler of a charred land ( i really don't care about the pop of the empire as long as they work ) than the dead ruler of a florescent land that was captured by the enemy ( Apres moi le deluge ) it is clear than striking first is a good option "

@ all

I guess that, at this point, it is clear that the AI management of nukes needs a clear revamp.

First of all the AI needs to be aware of all the consequences that launching nukes can have to them ( yes including global warming... nuke wars are not necessarilly Ragnarok scenarios ( atleast in game ) and it is quite common that both sides survive the war, so they should consider that nuking the enemy might awaken the evil sand rains fairy upon their lands ) . This appies especially in cases where another player gives nukes to them during a war... as it is today, that is a sure way of making a international pariah out of any player in game, no matter how he was loved before ( a thing that can be used for diplo wins: bribe the guy you are competing with for the UN into a war, give him nukes, see his diplo go to the reds, profit )

Second, ( this is probably the easiest to apply ) the AI should nuke in pairs when possible ( 2 sucessful nukes warrants complete cleaning of units in tile ) and should not nuke just because there is a city or a unit in range ( I mean, given the special nature of nukes, with it's teoretical infinite kill ability but only one strike in the useful life, the AI should be heavily discouraged to use them against unrewarding targets ).

Third, the AI should consider the presence of SDI when considering building and using nukes. SDI as it exists in game ( with it's completely over-bloated intercept ratio ) skewes completely the balance of making nukes ( and of what nukes to make ) compared with the conventional army. Same, in far lesser degree, to the nuke shelters.

Fourth, the AI, in a enviroment where they can't build more nukes, should be somewhat weary of using them. Well, this applies to all units, but here it definitely shows a lot.

P.S Massive x-post with Piece of Mind

 

[PieceOfMind]

rolo, I agree with a lot of what you say but I disagree strongly on one point in particular, and that is the AI taking global warming into considering.

Firstly, as I've said before, global warming hurts everyone indicriminately and proportional to their empire size. (And might I remind the reader, if all players are weakened in a manner proportional to their current power, then afterwards the ratios of all powers are still equal.) If I could destroy every military unit of my enemy but at the same time every 2nd tile in the world was turned to desert, the consequence is irrelevant because my position as a player would be much stronger.

Secondly, you say global warming should be taken into consideration but how on Earth would you do it? Sounds like something that is going to gobble up resources in a mod that already has to be very careful about what extra bloaty routines might get added. And for what gain? The only real reason for including GW considerations for the AI is purely roleplay reasons (maybe running Enviro civic etc.). After all, it is you who is arguing that Better AI is about making the AI better at winning or at its game. What actual gain are you going to get with GW concerns?

By the way, SDI should make the AI prefer to build tactical nukes. Generally speaking tactical nukes are better anyway, but particularly so after SDI is built. ICBMs should be more likely to be built in less threatened cities (i.e. not border cities) too.

 

[r_rolo1]

Well, true, GW is a global variable, but it hurts locally I couldn't care less what happens to other people ( in game ), but I do care if the evil sand rain fairy comes to my backyard. That is what I'm trying to say: I would like it to be a factor, not THE factor. I'm not saying "I will not launch the nukes because I don't want GW", I'm saying " It is worth to launch this nuke, considering that I might be hurting myself more via GW than I will gain? " , a thing that is completely diferent ( and that doesn't require that much of calcs: get GW count ( btw IMHO this number should be acessible to humans, but this is not the place to discuss it ) , get number of tiles you have, the number max of turns that it take for the game to finish, calculate the average damage you will get from GW until the max end of game, divide by the number of nuke blasts needed to activate GW, plot vs expected gain by nuke usage ( as I refered in the previous post of mine ), make a decision based on this. As a routine called only to nuke units while in war, I can't see this being bloaty ).

IMHO your view is being skewed ( take no offense in this ) by the way you see nuke wars in game. Ok, nukes can be used in a decisive way to win a game ( like I said before in this thread , I already won a game by conquest with the exclusive use of nukes ), but they can also not win a war or the game. The nuke usage in Civ IV is not a game decider by it self ( people do not surrender just because I have nukes ... thank the heavens, that would be extremely stupid, but I already seen people seriously suggesting it ), and i've seen enough of late AI-AI warfare with nukes flying to both sides and none of them acheiving complete victory ( and even human-human late warfare can end like that if both sides use nukes and obliterate most of the enemy land troops with it ). So, if there is a strong possiblity of a long post-nuke game period ( like I said above ) GW should be a consideration to have, because the damage done until the end of the game to myself via GW might not be worth the use of the nukes. OFC if the game is pratically ending ( what happens most of the times in standart settings, but not being necessarily true in mods ( like Next War derivatives ) or in Modern ( or even Industrial ) age starts ( especially with Advanced starts ) ) GW ishould be a minimal concern. And, OFC, if GW is already there, what the hell with it

To end, also remember that it would not be the first time that a mechanic virtually or completely useless of standart settings of unmodded games was inserted in BBAI just for possible mod usage ( we both know that a LOT of mods have BBAI inside(TM) )

PS: Enviromentalism does 0 effect over GW Building removing/ giving buildings does 0 effect on GW . I strongly recommend people to read this before discussing GW ( just because no one can't discuss in useful terms a thing that he/she doesn't know how it works No offense intended ) EDIT: added link ( )

 

[Dom Pedro II]

Actually launching nukes is, in many cases, irrational. As the link demonstrates, irrationality is a positive attribute in nuclear diplomacy.

The article you supplied certainly demonstrated that there was some use to having a big pompadour, owl-sized glasses and high-heeled shoes. And it was all about nuclear diplomacy not nuclear war. Not even conventional war with a nuclear option! Several nuclear powers have been attacked by conventional forces since the end of WWII, and one was facing complete destruction, and I don't think anybody would say that the best way to increase their chances would've been to just start dropping nukes.

Nuclear weapons are one of those funny things where being committed to follow through on your threats increases your chances of survival prior to the point where you have to do it, but dramatically decreases your chance of survival once you're forced to commit.

Consider: Ghandi says "if you declare war on me, I will nuke you". The player thinks this is a bluff and declares war. Ghandi doesn't use his nukes, as that would be irrational. Instead he says "if you invade me, I will nuke you". The player thinks this is a bluff, and invades. Ghandi doesn't use his nukes, as that would be irrational. Instead it says "if you capture a city, I will nuke you". The player thinks this is a bluff, and captures all the Ghandi's cities except one, using conventional forces. Ghandi relocates its nukes to his capital, and doesn't use them, as that would be irrational. Instead he says "if you attack my capital, I will nuke you". The player nukes the capital and takes it in a single turn.

Consider: Montezuma says "if you declare war on me, I will nuke you". The player thinks this is a bluff and declares war. Montezuma uses all his nukes the very next turn to causes maximum damage, even though this is irrational. The player gets cross, nukes the entire civilisation, trashes his diplomatic reputation, and conquers a charred and worthless continent.

Given the choice, and knowing about the different strategies of these two leaders, a player would prefer to attack Ghandi. This means that Montezuma's strategy is better at deterring nuclear attack.

I agree with you completely. But you're also countering an argument that no one has made: That the AI should never fire its nukes first under any circumstances.

@ DP II ( btw, a question I've wanted to do for a very long time : any Brazilian connection? )

Sim, sou brasileiro. Moro nos estados unidos

Benefit is always a very relative thing. You said, and with a lot of reason, that a AI should not start a nuke war without any benefit, but i strongly suspect that our definitions of benefit are somewhat diferent In the way I see it, and inside the discussion we are making ( AI player inferior both in nuke and conventional means to a attacker with unknown nuclear strategy in a world where you can't build more ), in this situation survivavility is the more important variable (there is no good in following a strategy that will make me die earlier, no matter how "rational" it might look at first sight ), and in that aspect gaining some extra turns as a charred wasteland ruler trumps easily letting your empire being eaten by your enemy conventional troops ( suposing you can do a nuke strike that wipes out a lot of the enemy troops OFC, as it looks to be the case in here ), barring the possibility of surrender ( in there OFC not provoking the enemy is a good strategy, but if you intend to surrender, why not do it without firing a shot, conventional or not, in the first place? )

My point is ( speaking like if I was in the AI place ) : "Someone declared war at me. I'm inferior both in nuke terms ( neither me or the enemy can do more ) as in standing army terms, so , if they put things at full speed even if only in conventional terms I'm doomed no matter what I do ( and I don't have any reason to believe they won't put things at full speed either in nuke or non nuke terms ). So what I'm left is to survive the longest possible and wait for something good comes from a exogenous source ( enemy gets too much protesters at home, UN or the Pope gets in the way and forces peace, enemy gets bored, aliens invade the planet, whatever... ), a thing that I can't do if I'm dead. Now I have a possibility of using nukes to strike a blow to the enemy that can delay my demise in some turns ( this is a important part ... IMHO it applies to the game we are talking, since they blew with a massive ( and, by the description, only in the area ) naval fleet, that would take a while to get back ), but I risk a massive nuke blowout if I do that... So, should I nuke or not?

Well, the fact is, like I said, I am doomed no matter what, barring some almost divine intervention. OFC there the possibility of being nuked no matter what I do , or my enemy can only nnuke me in response, or might even not nuke me reagrdless what I do ,to keep his nuke power intact ( as he has more conventional troops he might decide that they are enough ). In every case , if there is a possibility of dealing a good punch to the enemy that can give me some more turns I gain more time and I do prefer to be the living ruler of a charred land ( i really don't care about the pop of the empire as long as they work ) than the dead ruler of a florescent land that was captured by the enemy ( Apres moi le deluge ) it is clear than striking first is a good option "

Well, I think that's where we disagree then. I agree with your statement, but I disagree that simply fighting the conventional war will result in the AI being destroyed earlier than if they drop a couple of nukes. I mean, obviously, if the AI hammers the human with nukes, it's going to cripple the human's ability to attack immediately. But we're talking about a player that had only 3 nukes. A three-nuke arsenal is neither a deterrence nor an effective defense. Maybe in the real world where millions of lives are at stake, but not in Civ.

 

[Dom Pedro II]

Alternatively, maybe the game should have some kind of doomsday clock... instead of just global warming, dropping X number of nukes in X amount of time could cause nuclear winter


I remember playing a game of Rise of Nations with my friends at a cyber cafe. We were new to the game, and we were playing against the AI on a harder setting. Well, the AI was kicking our butts. RoN had such a doomsday clock. The AI was hammering us with nukes, and my friends focused on building a conventional army while I just kept building nukes and firing them off as soon as I got them. Long story short, we didn't lose the game. We didn't win either. We simply ended they game by detonating too many nukes

I was very satisfied

 

[Afforess]

Alternatively, maybe the game should have some kind of doomsday clock... instead of just global warming, dropping X number of nukes in X amount of time could cause nuclear winter


I remember playing a game of Rise of Nations with my friends at a cyber cafe. We were new to the game, and we were playing against the AI on a harder setting. Well, the AI was kicking our butts. RoN had such a doomsday clock. The AI was hammering us with nukes, and my friends focused on building a conventional army while I just kept building nukes and firing them off as soon as I got them. Long story short, we didn't lose the game. We didn't win either. We simply ended they game by detonating too many nukes

I was very satisfied

Yeah, I never understood why Nukes cause GW. Really, shouldn't they counter GW?

I don't like the idea of a doomsday clock. I want to be able to nuke as many times as I want...

 

[r_rolo1]

Sim, sou brasileiro. Moro nos estados unidos

Good to know ... Também com um nickname desses só podia ser ( tirando os lusófonos ninguém sabe quem foi D. Pedro II do Brasil e dos lusófonos decididamente seria um brasileiro a pessoa mais inclinada a usa-lo )

Well, I think that's where we disagree then. I agree with your statement, but I disagree that simply fighting the conventional war will result in the AI being destroyed earlier than if they drop a couple of nukes. I mean, obviously, if the AI hammers the human with nukes, it's going to cripple the human's ability to attack immediately. But we're talking about a player that had only 3 nukes. A three-nuke arsenal is neither a deterrence nor an effective defense. Maybe in the real world where millions of lives are at stake, but not in Civ.

Well, like I said, I was refering to the previous example. In there IMHO there was not much of chances of the Incan AI doing more than die or surrender ( and if it is to surrender, better do it fast ), because it already had their *ss served by a 4 city civ that had some military help of the human, so it shouldn't be expectable that it would withstand a direct conventional confrontation. Also, given the described geography ( human in one continent, shaka + Inca in other ) the human would have to ferry or airdrop the forces to give a fight on their land , so dealing with the human fleet would be a primary concern and if that happened it would take quite a while for the human being able to set feet in his continent, barring vassaling shaka ( and use his cities for airdrop ). So, in my book nuking his city was a tactically sound strategy, even if it unleashed the wrath of the nuke-filled human

Like I said in the post you quoted, this is only advisable if the less nuke-filled civ has a opening to make a devastating blow. If not, it is clearly better to wait.

About 3 nukes vs 40 being a effective detterent: they aren't, definitely. But nukes have a interesting caractheristic in Civ IV: in spite of no one fearing them a lot, 2 of them in theory can kill all the enemy army In certain tactical scenarios, those 3 nukes are worth more than 40 nukes of the enemy because of that ( SoD vs highly spread out army ). Not that I'm expecting a scenario like that to happen a lot

Alternatively, maybe the game should have some kind of doomsday clock... instead of just global warming, dropping X number of nukes in X amount of time could cause nuclear winter


I remember playing a game of Rise of Nations with my friends at a cyber cafe. We were new to the game, and we were playing against the AI on a harder setting. Well, the AI was kicking our butts. RoN had such a doomsday clock. The AI was hammering us with nukes, and my friends focused on building a conventional army while I just kept building nukes and firing them off as soon as I got them. Long story short, we didn't lose the game. We didn't win either. We simply ended they game by detonating too many nukes

I was very satisfied

Oh, RoN.... I just hated that doomsday clock (in the same way i hated Next war mod doomsday clock until they putted it in a more adequate way )

But i definitely agree that GW and nukes should be basically unrelated, especially since the day that nukes aren't the only causer of GW ( before you could rationalize and say that GW was the name they got for a variable that punished nukers )....

 

[Jawa'sRevenge]

Alternatively, maybe the game should have some kind of doomsday clock... instead of just global warming, dropping X number of nukes in X amount of time could cause nuclear winter

I was very satisfied
Yeah, I never understood why Nukes cause GW. Really, shouldn't they counter GW?

I don't like the idea of a doomsday clock. I want to be able to nuke as many times as I want...

Yes, no doomsday clock, Nukes should have that much affect, plenty have been detonated. As for global warming

 

[Dom Pedro II]

I don't like the idea of a doomsday clock. I want to be able to nuke as many times as I want...

Please, let us know if you are ever running for high political office...

Good to know ... Também com um nickname desses só podia ser ( tirando os lusófonos ninguém sabe quem foi D. Pedro II do Brasil e dos lusófonos decididamente seria um brasileiro a pessoa mais inclinada a usa-lo )

Or Santos Dumont

Well, like I said, I was refering to the previous example. In there IMHO there was not much of chances of the Incan AI doing more than die or surrender ( and if it is to surrender, better do it fast ), because it already had their *ss served by a 4 city civ that had some military help of the human, so it shouldn't be expectable that it would withstand a direct conventional confrontation. Also, given the described geography ( human in one continent, shaka + Inca in other ) the human would have to ferry or airdrop the forces to give a fight on their land , so dealing with the human fleet would be a primary concern and if that happened it would take quite a while for the human being able to set feet in his continent, barring vassaling shaka ( and use his cities for airdrop ). So, in my book nuking his city was a tactically sound strategy, even if it unleashed the wrath of the nuke-filled human

Like I said in the post you quoted, this is only advisable if the less nuke-filled civ has a opening to make a devastating blow. If not, it is clearly better to wait.

About 3 nukes vs 40 being a effective detterent: they aren't, definitely. But nukes have a interesting caractheristic in Civ IV: in spite of no one fearing them a lot, 2 of them in theory can kill all the enemy army In certain tactical scenarios, those 3 nukes are worth more than 40 nukes of the enemy because of that ( SoD vs highly spread out army ). Not that I'm expecting a scenario like that to happen a lot

Well, I think aside from some small details, we've found common ground. Couldn't ask for better from a good debate Don't really have much else to say.

Oh, RoN.... I just hated that doomsday clock (in the same way i hated Next war mod doomsday clock until they putted it in a more adequate way )

Well, any possible doomsday clock should definitely have high decay rate to it. So if you have two players that have 50 nukes each, and they both fire them all in one or two turns, the map will become very bad. But if they fire 100 nukes over 20 turns, it won't really do much (aside from the destruction to the targets of course).

 

[Afforess]

Please, let us know if you are ever running for high political office...

Hey now, Reagan acted like a madman, and it worked against the USSR.

 

[PieceOfMind]

Well, true, GW is a global variable, but it hurts locally I couldn't care less what happens to other people ( in game ), but I do care if the evil sand rain fairy comes to my backyard. That is what I'm trying to say: I would like it to be a factor, not THE factor. I'm not saying "I will not launch the nukes because I don't want GW", I'm saying " It is worth to launch this nuke, considering that I might be hurting myself more via GW than I will gain? " , a thing that is completely diferent ( and that doesn't require that much of calcs: get GW count ( btw IMHO this number should be acessible to humans, but this is not the place to discuss it ) , get number of tiles you have, the number max of turns that it take for the game to finish, calculate the average damage you will get from GW until the max end of game, divide by the number of nuke blasts needed to activate GW, plot vs expected gain by nuke usage ( as I refered in the previous post of mine ), make a decision based on this. As a routine called only to nuke units while in war, I can't see this being bloaty ).

IMHO your view is being skewed ( take no offense in this ) by the way you see nuke wars in game. Ok, nukes can be used in a decisive way to win a game ( like I said before in this thread , I already won a game by conquest with the exclusive use of nukes ), but they can also not win a war or the game. The nuke usage in Civ IV is not a game decider by it self ( people do not surrender just because I have nukes ... thank the heavens, that would be extremely stupid, but I already seen people seriously suggesting it ), and i've seen enough of late AI-AI warfare with nukes flying to both sides and none of them acheiving complete victory ( and even human-human late warfare can end like that if both sides use nukes and obliterate most of the enemy land troops with it ). So, if there is a strong possiblity of a long post-nuke game period ( like I said above ) GW should be a consideration to have, because the damage done until the end of the game to myself via GW might not be worth the use of the nukes. OFC if the game is pratically ending ( what happens most of the times in standart settings, but not being necessarily true in mods ( like Next War derivatives ) or in Modern ( or even Industrial ) age starts ( especially with Advanced starts ) ) GW ishould be a minimal concern. And, OFC, if GW is already there, what the hell with it

To end, also remember that it would not be the first time that a mechanic virtually or completely useless of standart settings of unmodded games was inserted in BBAI just for possible mod usage ( we both know that a LOT of mods have BBAI inside(TM) )

PS: Enviromentalism does 0 effect over GW Building removing/ giving buildings does 0 effect on GW . I strongly recommend people to read this before discussing GW ( just because no one can't discuss in useful terms a thing that he/she doesn't know how it works No offense intended ) EDIT: added link ( )

I mentioned Environmentalism in the context of a roleplay argument. I didn't claim it had anything to do with the GW mechanics.

You're still kinda dodging my point. At least in the standard game, what are you actually going to achieve by making AIs think about GW? How are you going to code it. I'm sure there's lots of wonderful ideas you can teach the AI but there has to be a strong reason if it's going to take up some programmer's time. If a modder wants to change GW mechanics then good for him - he can teach the AI about the new mechanics. It would be impossible for a Better AI programmer to predict what other mods would do anyway. Better to just work with the base game and perhaps help it with changes like those in RevDCM (which seems to happen ).

Can you show me a single game where any player, human or AI, had to give much consideration to GW in regards to nukes and not doing so seriously damaged their chances of winning the game?

My whole point is that it's likely to hurt everyone roughly equally. If you want to go into the nitty gritty, then half the time it's actually going to be beneficial to induce GW because it will hurt someone else more. So far you seem to imply that GW is bad for every nation regardless. Would you be comfortable with this proposed considering of GW including wanting GW to happen and so deliberately trying to cause it? If not, it would seem your wanting them to worry about GW is only because of your dislike of the GW mechanics and seeing tiles in your own empire getting damaged. It's not based on any rational reason in the interests of the AI making the decision.

 

[r_rolo1]

My argument has little to do with my like or dislike of the GW mechanics ( note that I haven't mentioned explicity any dislike, except for the name I gave it , that is irrelevant for the point ). Besides being somewhat baffled with the negative response , in a forum for a mod that wants to make a better AI , that a proposal for making the AI aware of a important and possibly game-breaking variable is having , my argument lies in the fact that the main concern of a civ is their own backyard, and that was the line of interpretation I used for my proposal above ( that, I reckon, needs heavy refinenement to include net negative values for the GW damage value ). And obviously you are right to say that in average most of the people will be hurt the same, but i wonder if you can really say that GW hitting you in one tile will be counterbalanced by your two times bigger neighbour being hit in two tiles ( or the other way around ) ...

And yes, I would be receptive to a mechanism that made the AI wanting GW in some situations, especially if it has the means to have food and hammers without having to resort to working land tiles ( read corps and/or Dutch with lots of water tiles ) . I've seen enough of humans doing that

About the "standart" settings. what is that? For me it is Immortal/Standart map size/Normal speed/ 6 AI/ Fractal ( while not playing SGs , that have their own special caveats ) . I can assure you that other persons will give you other awnsers. The AI has to be able to top with all of them in a minimally educated fashion ( think on what it would be said if you proposed to not teach the AI that it can airdrop paratroopers because, in "standart" settings they have not much of use and the code would only bloat the AI decision making .... ) . I already said and you quoted 2 non-modded settings ( there are obviously more ) where GW will be surely be a major factor in game, and in those keeping the AI in the dark about it is a assurance of underachievement for the part of the AI. Anyway, as you asked, I have a nice little story of a game in "standart" settings where GW went berserk ( or " Remember that saying: "Never trust a Aztec with nukes" from Civ II? It is well and alive in Civ IV" )