Artificial Intelligence

No, you can not.

An AI designed to "play to win" is by definition an AI that always does what if feels will give it the biggest chance of winning the game. If that means nuking a friend, so be it.

As soon as you deviate from this rule you are compromising and no longer have a pure "play to win"-AI.
I meant it's possible to code it that way.

If your point of view is that if a civ isn't "trying to win" of it doesn't go BALLS OUT TRYING TO KILL YOU IMMEDIATELY as soon as you start building the Utopia Project or building your spaceship parts, then we disagree on that principle. But it can certainly be coded so that the AI continues to primarily pursue their own victory conditions while you pursue yours, without coming after you as you near yours.
 
Personally, I don't care if they AI plays to win. I just want them to play competently and be more masterful with diplomacy. In AND for instance, the AIs using Ruthless AI would 1.) Ask other civilizatons to declare war (not just the player) before starting a war and 2.) Call for help when invaded.

Seems small but it had a big effect gameplay wise.
 
Personally, I don't care if they AI plays to win. I just want them to play competently and be more masterful with diplomacy. In AND for instance, the AIs using Ruthless AI would 1.) Ask other civilizatons to declare war (not just the player) before starting a war and 2.) Call for help when invaded.

Seems small but it had a big effect gameplay wise.
As long as they also teach them what "no" means and make them smart enough to stop asking me every... single... turn. :)
 
Personally i really prefer an AI which plays roleplaying historically the civilization or the leader over an AI which always try to win.
There should be a balance though because an AI needs to be also competitive.
If there is one thing which i really hate is being compelled to play at the last level of difficulty, where AI gets so much advantages that you are forced to make everytime the same tactics to fill the gap technologically and economically (early rush, cheesy tactics, exploit AI weakness and so on), to get a competitive game.
 
If your point of view is that if a civ isn't "trying to win" of it doesn't go BALLS OUT TRYING TO KILL YOU IMMEDIATELY as soon as you start building the Utopia Project or building your spaceship parts, then we disagree on that principle.

I think the way most (multiplayer) games deal with the scenario you are describing is by giving value to second place. For instance, it's evident one player is going to win very soon. The next strongest players, instead of trying to kill him, will fight each-other in order to get second place, because first place is out of reach.
 
I think the way most (multiplayer) games deal with the scenario you are describing is by giving value to second place. For instance, it's evident one player is going to win very soon. The next strongest players, instead of trying to kill him, will fight each-other in order to get second place, because first place is out of reach.
Didn't know that (I've never once played an MP game of any Civ). That sounds pretty cool! Kind of like a wildcard playoff spot. :p
 
I think the way most (multiplayer) games deal with the scenario you are describing is by giving value to second place. For instance, it's evident one player is going to win very soon. The next strongest players, instead of trying to kill him, will fight each-other in order to get second place, because first place is out of reach.

i think thats true with pretty much every multiplayer game that was multiple players playing against each other, racing games comes to mind easily.
 
Playing to win means you have a plan for winning and you try to reach it. That doesn't necessary means that's the best plan or what it use every feature available. In Civ4 AI don't have this.
 
The problem is that (other than domination) the victory conditions in Civ are essentially artificial milestones - they should mean nothing to the AI civs. The victory conditions are only there to give the player a goal to achieve, a concrete way to determine success or failure. They are a challenge issued by the game to the player: given this set of circumstances, can your civilization achieve any of these goals before another civ does. Achieving a victory condition is, quite literally, NOT the end of world - except to the player.

If this were not the case, then the diplomatic victory should be impossible to achieve - every civ should abstain or vote for themselves,

AI civs should not prevent their friends or allies from achieving Cultural, Diplomatic, or Spaceship Victories, because (from the AI perspective) they would benefit the AI in the future.
 
If this were not the case, then the diplomatic victory should be impossible to achieve - every civ should abstain or vote for themselves

That's why they have city-states. Regular civs will not vote for you in Civ 5.

And your general arguments are wrong. Which milestones should mean nothing to other civs - flying Alpha Centauri, building Utopia, or claiming world rule through diplomacy?
 
Would an ally really go to war to prevent a friendly nation from launching a Spaceship? Did the Soviet Union nuke the US when they were about to land on the moon (and they were enemies)?

Why should they mean anything to the AI Civs?
 
Would an ally really go to war to prevent a friendly nation from launching a Spaceship? Did the Soviet Union nuke the US when they were about to land on the moon (and they were enemies)?

There's a difference between science mission and space colonization. All victory conditions are world domination actually. That's from realism point of view.

And from gameplay point of view - you're playing against opponents, which meant to be equal. For "test of time" you have barbarians and city-states.
 
Top Bottom