Ingrate!

[Atwork]

It's clear that the AI is programmed to defeat the human player at all costs. It only seems logical that they'd all be in it together

I would support such behavior IF the AI civs actually did coordinate strategically to accomplish self-interested goals. In other words, if both AI-A and AI-B have interests, or even mutual interests, that are achievable only by coordinated activity, or would best be furthered by coordinated activity, then by all means, they should act in concert. An opportunistic AI that coordinates strategically with others advantageously is a welcomed change.

However, AI-A should not cooperate with AI-B in a coordinated attack against the human unless (1) AI-A has a significant chance of gaining something as a result & (2) the goal is achievable.

If AI-A stands to gain nothing, or worse yet stands only to lose, significantly, then the AI programming should lead to AI-A avoid coordinating with AI-B in an open attack. Perhaps there is another way that AI-A can advance its interests by working with AI-B, but in a way that won't lead to AI-A's demise. Examples: financial support, gifting military, open borders, resources, helping AI-B diplomatically by denouncing AI-B's denouncers, etc.

In the end analysis, the programming simply should not push AI-A to coordinate with AI-B if AI-A stands only to lose and therefore be weakened relative to the human and all other civs. On the other hand, if coordinated activity is advantageous to AI-A, then AI-A should coordinate its activity with AI-B.

Now, in my game's case, there was no issue of coordination. The Aztecs did not coordinate its activity, but rather isolated itself from everyone (including its sponsor state -- me) and therefore stood no chance of winning, instead guaranteeing its failure. That's horrible programming!

The Aztecs would have had more to gain by coordinating its activity with me against the 3rd-party civ that it was already at war with. I would have been willing had the Aztecs asked. Instead, they DOWed me and ended up surrounded. After losing nearly its entire army, another civ declared war on the Aztecs. Aztecs begged for peace with me -- I accepted and then ended up funneling them a military so they could survive.

Sorry the post is so long!!

 

[Fluffball]

An example i recently posted on the official forum:

England just declared war on me.

• We have been friends since the start of the game (nearly 500 turns on a marathon game) with many trades including a current research agreement and 2 current trade deals.
• Their empire is about 1/3 of the [huge] map away.
• There are two major civs inbetween us.
• They have around 2/3 of my military strength.
• They have less than 100 gold but with a modest GPT so it's not like they have been forced to declare war as they can't afford to pay the GPT they owe me.
• Their in the last 1/3 of the score board.
• They have a small number of cities with a small amount of production (info addict) compared to my large empire with a large production base.
• I am not at war with anyone so no chance of a sneak attack while i am busy and in fact have crushed multiple civs that declared on me in the past including my last war that finished two turns ago.
• They aren't going to gain any friends by DoW me as i am not at war with anyone currently.
• They admitted they have no chance in the DoW.

• So i don't appear weak in any sense and in fact they know they will lose if it actually came to fighting.
• They are no where near me and if they are using some logic that they should attack a top civ there is a bigger civ (Mongolia) on their border so why not attack them if they want glory or death or to try to knock down one of the big boys.
• They can easily afford to pay me the GPT they owe me and have lost a large amount of gold from goods they have already paid for and the RA that has now been cancelled.
• If another civs has bought them to DoW then why would they ask such a pointless opponent instead of one of the many stronger civs actuall on my borders who could at least attack me?

Comparitively they have lost a lot already by simply declaring war and potentially could lose it all if i could be bothered to got fight them and have no potential to gain anything.


6 turns later.
England offered peace, giving me 395 gold, 49GPT, dyes and open borders...everything they had to give apart from their cities.
The 'war' consisted of a lowly pikeman wondering past my border seemingly trying to get home as it was on the other side to where England is.
I almost felt sorry accepting what they offered because it was so stupid.

And a conclusion i came to later in the thread:
The crux of the issue is, and especially because the AI is meant to play to win in this version, what potential benefit could they possibly see in DoWing me to cause them to do so in this situation and when they admit they have no chance at all when declaring war?
How can you even use the term 'Artificial Intelligence' when refering to someone that will declare war in a situation they have no hope, no logical reason and even admit they will lose? Artificial Stupidity would be a more accurate terminology to use and i think we should all change from reffering to the AI to refering to it as the AS.

 

[Smokeybear]

QFT. Absolutely. I don't think it even qualifies as a backstab if the odds are so long against them. More of a bottom pinch.

Yup. Backstabbing should be a rarity, not the strategy of overused choice.

 

[Smokeybear]

And a conclusion i came to later in the thread:
The crux of the issue is, and especially because the AI is meant to play to win in this version, what potential benefit could they possibly see in DoWing me to cause them to do so in this situation and when they admit they have no chance at all when declaring war?
How can you even use the term 'Artificial Intelligence' when refering to someone that will declare war in a situation they have no hope, no logical reason and even admit they will lose? Artificial Stupidity would be a more accurate terminology to use and i think we should all change from reffering to the AI to refering to it as the AS.

I've believed from the start that they simply bent all aspects of the game, diplomacy and all related AI behaviors, decisions and activities- to get wars going. Because it is a wargame, first and foremost, and that is what the devs set out to enable as much as possible... wars. They done good. They don't have to make any sense, just so long as the player is consistently engaged in this activity which was their overarching purpose in creating this particular game in the first place. It's a turn-based combat strategy game- not a civ builder.

 

[Atwork]

It's a turn-based combat strategy game- not a civ builder.

Hmmmm....I think many would disagree with you, in part or in whole.

Personally, I play the game balanced between the two -- a builder unless attacked or unless the opportunity arises for a quick and decisive war. I certainly do not view the game (or the series) merely as "combat strategy".

 

[AxelRhodd]

If we entertain that the AI civs are just a big collective controlled by one AI process then it kinda explains all those stupid/crazy/suicidal AI actions. Basically the AI is working towards presenting the human player with ONE powerful opponent. Perhaps the AI need to suicide all other civs to allow for one nominated civ to become the strongest.

 

[Fluffball]

It's a turn-based combat strategy game- not a civ builder.

If that is the case then it only reinforces the point that the developers were/are out of touch with what the players wanted. Yes some play for war but others play to build while others play for a mix of both.

Wars can keep the game interesting but there should be balance as constant war is just as boring as constant peace without fear of war.
I am only playing on immortal in my current game and after the first 70 turns or so i have been pretty much at constant war, all of which have been declared on me. As soon as i finish one war a couple of turns later (maximum of 5) another seemingly random civ DoW me.

 

[CYZ]

I think it could be true that the AI's are linked. Sometimes their behaviour seems to correline and effect each other in very suspicious ways.

Personally I think coding a collective AI pretending to be several different AI's would not be better than actually coding seperate AIs. But what do I know?

 

[HappyBuilder]

England offered peace, giving me 395 gold, 49GPT, dyes and open borders...everything they had to give apart from their cities.
The 'war' consisted of a lowly pikeman wondering past my border seemingly trying to get home as it was on the other side to where England is.
I almost felt sorry accepting what they offered because it was so stupid.

This had become strategy in some games I'd played at lower difficulty levels. Agree to join in a war against a faraway civ with no intention of fighting it and just wait for them to offer their entire treasury to you to 'stop'.

At higher difficulties, I haven't found the AI so willing to compromise, even when it's getting throttled in a shooting war.

HB

 

[headcase]

How can you even use the term 'Artificial Intelligence' when refering to someone that will declare war in a situation they have no hope, no logical reason and even admit they will lose?

The problem is that they tried to make it "like a player", except a real player would quit the game at that point. So the AI should just be a little more role-playish in that situation instead of spiting the player, especially when it has no solid reason to spite the player compared to other civs. Few real life civilizations have started a war because they knew they would lose.

 

[Chessack]

The problem is that they tried to make it "like a player", except a real player would quit the game at that point. So the AI should just be a little more role-playish in that situation instead of spiting the player, especially when it has no solid reason to spite the player compared to other civs. Few real life civilizations have started a war because they knew they would lose.

This is a good point.

I would add a few more. That 4-city civ that has no hope of winning a military victory might, if it was not stupid enough to attack the 28-city human, be able to sneak in a Cultural victory instead. Why not program the AI to concentrate on alternate victory methods (assuming those are enabled)? After all isn't that what a human player would do? Sure if victory is impossible you would resign, but in the early going if you see you are outmatched for one option, you might try a different option. Heck I do this now... I might start a game planning for a world-conquering military victory, but as the map shapes up, decide that doing so would be much harder than say a 3-city empire with a cultural win. Or after a long period of play and racking up lots of city-state allies, I might decide to go for a UN win.

In fact the last 2 games where I won with Policies and the UN, in both cases I did NOT have the strongest army, and I would probably have lost of the leading AI had not been wasting its time fighting some of its neighbors (admittedly, somewhat incited by me, LOL). I had been going for the policy win but not the UN one... I switched to it when I saw that my research and production would probably let me pull it off, and before the AI could either build Apollo or conquer the remaining capitals (including mine).

The point is, in a normal game (one where the victory conditions are all available), the AI should be willing (and able) to shift gears. If it sees itself falling way behind on tech, stop trying for a science or even military victory, and swap over to Cultural or Diplomatic instead. If it sees the other guy way ahead on policies, shift to a tech strategy... Etc, etc.

But it doesn't really seem to do that. It seems to just DOW whenever it tries to beat you.

 

[Carnifaux]

Then, just as I was about to join them in a war against a common enemy, suddenly, they declare a suicide war against me. With their pikeman, they attack my musketman who were fortified on hills inside of forts. My ally city-state proceeds to decimate their lines from behind. Within one turn, they lost 75% of their military (half of which I provided). Sigh. A few turns later they beg for peace and I am merciful. Now I have set about builiding them back up.

You are a more forgiving and patient sort than I am, I would've made no mistake about puppeting the cities the first time.

 

[Strategist83]

I think too many of us assign human-like reasoning power to the AI civs. All they are -- bottom line -- are a few thousand lines of computer code.

This is the exact situation. The problem is, of course, that we IMO are in our good right to expect a little more from those 1000 lines of code. CiV is supposed to be a 'flagship' title and has sold millions of copies. I will reinstate my opinion (which I must have hinted at the first 100 or so times already in my posts on this board): Firaxis didn't do well enough to begin with and their follow-up patches leave much (much indeed) to be desired.

As soon as a solid competing company (ie. Blizzard) turns their attention to the turn-based strategy genre we should all dump Firaxis' offerings and move on to some actual decent software. The only reason why we are here is because of Firaxis' near-monopoly on the turn-based strategy scene. Unfortunately, the likelihood of others turning to TBS games seem about as great as the chances of winning in the lottery.