They say that The AI will now try to win a cultural victory in BTS, but I think that isn't enough. I think that they should program the AI to try to win any one of the six victories, not just one or two, because then they would have no way of winning aside from those two, and time. But hey, I think it would be cool if the AI wanted to try to pull off a conquest victory. I'd stop them, but it would make the game more interesting. Who here agrees?

I would love, love, love for the AI to go for conquest and domination victories! Being a Warmonger, it bugs me how peaceful the AIs are with each other, I would love for some of them to go after each other, resulting in bloody, World Wars! Muahaha, I love the idea!

I think you're missing the point... my understanding of the description is that the AI will now go for ANY victory. The only reason they listed Cultural is because it's the hardest and requires the most planning.

I certainly hope and would assume that each leader will have a "favorite" victory type...though that will beg the question, will each leader only persue one victory type? I hope not.

I think you're missing the point... my understanding of the description is that the AI will now go for ANY victory. The only reason they listed Cultural is because it's the hardest and requires the most planning.

I didn't see it that way, but still, I would like to see other civs going for domination or conquest.

I would love to have the AI trying a conquest victory on me. Instead of always having me going around the world conquering them, how about the AI conquering me?

Da_V_Man:
Gaius Octavius:
I think you're missing the point... my understanding of the description is that the AI will now go for ANY victory. The only reason they listed Cultural is because it's the hardest and requires the most planning.
I didn't see it that way, but still, I would like to see other civs going for domination or conquest.

Gaius is right. BetterAI (or some form of it) has probably been incorporated into BtS so the AI likely go for different victories. I'd also point out that the article which gave that piece of information was a magazine preview - not an interview with Firaxis staff - who tend not to be less careful with their choice of words; noting cultural victories and failing to mention other types.

A thought just occurred to me: if BetterAI is included, then the leaders will now know how to use their UU's properly (at least in theory). You know what that means, right? I pity the poor fool who gets stuck next to Augustus without iron or copper...

if BetterAI is included, then the leaders will now know how to use their UU's properly (at least in theory).

I suspect more importance was placed on giving the AI adequate understanding of how to make use of the new features.

An Ai conquest or domination would be amazing to see. The closest I've seen in warlords was an AI Ghengis Khan controling 45 percent of the land. Unfortunately he stopped there, and modern armor have a way of defeating cavalry. It just doesnt make sense, you'd think horses would be immune to heavy cannon fire, but I guess they aren't.

I certainly hope and would assume that each leader will have a "favorite" victory type...though that will beg the question, will each leader only persue one victory type? I hope not.

Humm, that sounds interesting and makes total sense!
Like Isabella aiming for the new religious victory, or Genghis with an horde to take over the world :king:
Anyway, I think it'd still be possible to see, say, Montezuma going for spaceship.
They said the AI is greatly developed :goodjob:

Humm, that sounds interesting and makes total sense!
Like Isabella aiming for the new religious victory, or Genghis with an horde to take over the world :king:
Anyway, I think it'd still be possible to see, say, Montezuma going for spaceship.
They said the AI is greatly developed :goodjob:

I like this idea too. Alex and Monty are already psyched to attack people, so why not let them go loose and actually conquer something for once? I would imagine it wouldn't be hard to implement a favorite victory, much like the civic feature, as the civs are already programmed with specialized characteristics and playing styles.

I do think the AI needs to be a little more aggressive with the war thing, though. I almost always have to orcestrate elaborate methods to force wars between AI civs, otherwise many of my games would simply involve me racing towards cultural or SP victory while the AI plods along quietly. If Firaxis could improve ANYTHING to the current incarnation of civ, this would be it.

I think it would kind of suck if the AI played like multiplayer humans, where they rush your capital ASAP. If this was true, there would be one golden rule: "Never explore with your first unit!"

Padraig:
I certainly hope and would assume that each leader will have a "favorite" victory type...though that will beg the question, will each leader only persue one victory type? I hope not.

I'm probably in the minority, but I hate this idea. I want civs to go for victories based on the circumstances they find themselves in. If Hatshepsut finds herself the master of two or three vassals, I want her to consider going for a domination victory - not pressing on with a cultural one.

I'm probably in the minority, but I hate this idea. I want civs to go for victories based on the circumstances they find themselves in. If Hatshepsut finds herself the master of two or three vassals, I want her to consider going for a domination victory - not pressing on with a cultural one.

Don't worry; it sounds like it would be extremely hard to implement that idea. It stands to reason win goals will be based on circumstances, not a pre-set plan.

Culture victories are my favorite cause I like to flip cities and take over one or more continents or islands that way. Usually, I turn off the Time and Space Race victories in Warlords to get that far. I mostly play always peace games, so that pretty much throws out the Conquest victory. I just learned how to use the Spy to cause riots in neighboring cities to hasten cultural conversion with Amra's mod so I'm excited about all the new Spy features coming into BtS. Now I'll have to be more wary of my neighbors even in peaceful games. :eek:

I think you're missing the point... my understanding of the description is that the AI will now go for ANY victory. The only reason they listed Cultural is because it's the hardest and requires the most planning.

I agree.

I certainly hope and would assume that each leader will have a "favorite" victory type...though that will beg the question, will each leader only persue one victory type? I hope not.

I can the AI having a preference for a certain vc, but I can't see them going for only a specific one.

I'd also point out that the article which gave that piece of information was a magazine preview - not an interview with Firaxis staff - who tend not to be less careful with their choice of words; noting cultural victories and failing to mention other types.

:lol: Boy, are you right!

I just hope the AI are setup so they don't overextend themselves. While it would be fun to watch Genghis trample the world only to have his empire collapse in on itself, I don't want to see it everytime.

Did they(Firaxis) first said that AI does not need anymore bonuses? Now they are saying that AI does not need "so" much bonuses. Hopefully AI is really much better than before, so when you change difficult level it also influences to AI's behaviour. Or am I wrong? :)

No I think, the way the AI will work will be

Medium= Smartest AI with very minimal bonuses (the type they are trying to minimize I think, are Bonuses that made certain game concepts almost irrelevant for the AI.. Barbarians, WW, etc.)

Medium->Low=Increasing human v. AI Bonuses, slight decrease in AI 'optimum behviour' ie smartness
Medium->High= Increasing AI v. human Bonuses

The higher levels will need bonuses, because they might get an AI to beat a Competent human, but not to beat a Good human.

Changing AI behavior with Level is rarely a good idea, because that means you have to spend time programming a 'stupid AI'. That is only worthwhile on the lower levels.

Programming a stupid AI when you have a smart one is easy in a turn based game. Just make the AI pick moves it doesn't consider optimal some of the time. That's how chess computers work.

Of course, if the move the AI considers optimal isn't very good, you don't end up with a good AI anyway.

I recall something about this being discussed with whomever did the AI in RoN (I think), and how they did it was NOT a matter of making lower level AI intentioanlly stupider, just that as you get into the higher levels of AI it makes more checks and refinements to its overall strategy (i.e. will be more adaptive to what is currently going on).

Well there are of course more ways than one to hamper the AI. Normally when programming an AI I aim for an unbeatable strategy (not possible in a game like Civ of course), and then add handicaps. Increasing strategies for higher level AI is a good strategy also since lower level AIs will require less processing time. It's similar to chess AI again in that the look ahead is reduced in low levels to save processing time. High level chess computers often take a long time to choose a move which is often not practical in a game where people don't want to be waiting for ages for an AI to move. Both AI strategies are valid though.