Conflict between RPG players and numbercrunshers

[Hambil]

It'd easy to say a better AI is just to hard to make and we have to live with it. However, we've been living with a bad AI since Civ I. The rest of the game has always been good enough to make up for it.

The idea that the ai can't be improved though, I do not buy. Consider the tactics the AI needs to use in a twenty man raid in WoW. If that can be done, the AI can fight better IMHO.

 

[civvver]

Except civ is not an rpg. It's a emprie building simulation. I know there's a forum on here where people type out stories and tales corresponding to their games, but those might as well just be works of fiction on a blog, they can be completely abstracted from the game. The game has no characters to speak of, only immortal leaders, there is no plot. I fail to see how there are any rpg elements at all.

 

[Hambil]

I fail to see how there are any rpg elements at all.

Your lack of imagination is not my problem

Officially it's a 4x game. link I'm not going to pursue this discussion further as I think people are talking at each other instead of to each other.

 

[Veneke]

...

AI that is not trying to win is an AI that sits there and lets human players do their thing, whatever it is, without ever threatening them or causing to change their plans in any way. Many players don't care whether they lose or win and some get very upset when being beaten due to the play style they have chosen. So maybe in theory that specific group has no problem with the idea AI is capable of winning, however, in practice, such an AI completely lacks this capability.
I was not talking about bad coding, e.g. being unable to add the final SS part, tactical ineptness etc.

...

An AI not trying to win doesn't necessarily mean that they leave the human player alone. It's not a terribly good or enjoyable story without an antagonist (or four), provided of course that these antagonists develop as you might come to expect and it's possible to find friends amongst the AI to take on the 'bad guys' so to speak. In fact, I think I can say with a fair degree of certainty that there's no story at all without AI interaction with the human player. It might suit people wanting to play sim empires though, who I've come to accept are a different group entirely. As it happens, number-crunches more aptly describes that crowd than it does the people interested in beating the AI but the naming of these different groups across the board has, of late, degenerated rapidly into wholly inaccurate terms.

 

[The Pilgrim]

think we can agree on that.
But thats how ais in every (strategy)game act - thats not like its civ rpg related

I wouldn't say so. There are games with less shiny stuff and more competitive AI. Old games mostly, released during TBS golden age. Firaxis can make Civ5 AI more competitive (even Civ3/Civ4 was, IMO), but why would they if so many players think it's not about winning and AI should not try to do that. Call it RPG, simulation or a hybrid of whatever, Civ5 is not a pure strategy game either. If it was, all the resources put into graphics and creating narrative immersion would have been put into more competitive strategic/tactical AI.

 

[The Pilgrim]

An AI not trying to win doesn't necessarily mean that they leave the human player alone. It's not a terribly good or enjoyable story without an antagonist (or four), provided of course that these antagonists develop as you might come to expect and it's possible to find friends amongst the AI to take on the 'bad guys' so to speak. In fact, I think I can say with a fair degree of certainty that there's no story at all without AI interaction with the human player. It might suit people wanting to play sim empires though, who I've come to accept are a different group entirely. As it happens, number-crunches more aptly describes that crowd than it does the people interested in beating the AI but the naming of these different groups across the board has, of late, degenerated rapidly into wholly inaccurate terms.

Sure, it's an antagonist that is doomed to death a priory. And that's exactly kind of role AI is playing as far as many players are concerned. A prop, nothing more.
And again, I agree that terminology is not very accurate. But we get the idea, I think. There are those who play to win and want AI to do the same, so called number crunchers, and there are everybody else.

 

[whyidie]

Sure, it's an antagonist that is doomed to death a priory. And that's exactly kind of role AI is playing as far as many players are concerned. A prop, nothing more.
And again, I agree that terminology is not very accurate. But we get the idea, I think. There are those who play to win and want AI to do the same, so called number crunchers, and there are everybody else.

Maybe I don't know what box I fit into. I thought I am what has been termed an RPG/Narrative player. However, if the AI cannot win, even I would lose interest.

Overall I don't think too many people fall absolutely into one circle or another. I think for the vast majority there are areas of overlap (I think the AI at least putting up a struggle is one such area that the majority want) its just a matter of what the tipping point is for most players.

What mechanic is it that ultimately leads to dissatisfaction. I put myself in the RPG/Narrative camp purely because I rationalize parts of the game that a simulation player would find appalling.

Transport ships are a good flashpoint. To me, its less tasks so I don't mind we're not simulating them. Some find it to be an immersion breaking element.

Diplomacy is another flashpoint between simulation fans and narrative folk. I find that no game will be able to simulate the depth, complexity, and simplicity of diplomacy. I don't go in with a set perspective of "the AI should do this every time given these variables" because the most unpredictable part of the the narrative (human behavior) becomes programmatic. In this sense I want there to be some logic, but I don't want it to be too apparent or predictable. Once I've mastered it, it loses the fun.

Which maybe brings up a good area.Randomness. I like it in the right amount. Oops. Earthquake. Out of your control. Couldn't have planned for it. Disease, one you couldn't have planned for. Your game is impacted. To me, this is good. This is a twist in the story. To a gamist this represents an unpredictable outcome that destroys an otherwise perfectly planned and executed strategy. I think most gamists (number crunchers) consider it a waste of a good game.

 

[LoneRebel]

Let me guess, you're an RPG'er in your heart?

That's the difference between RPG/Historical simulation/Empire building simulation or a hybrid of all and pure strategy.

I prefer to avoid labeling myself. Although the way I see it, games should be played for fun. We play games to relieve stress. When a game becomes so stressful that it stresses you out more than work or whatever, it's time to stop.

All I'm saying is I don't really agree that the AI is not trying to win. You can say they suck at it, but to say that they're not trying to win - no. For someone who is not trying to win, the AI wins pretty often, especially at higher difficulties.

I'm also going to have to disagree with your statement that "numbercrunchers" are the only people who play to win and expect the AI to do the same. I don't consider myself a numbercruncher, and I play to win. I expect the AI to do the same, within reason.

The fact that the AIs are playing to win partly ruins the roleplaying experience because countries in real life are generally not playing to win by being the first to launch a spaceship, conquering the capital cities of all other major countries, or getting one of their citizens elected as UN Secretary-General. But it's fine with me.

To illustrate my point: Want to make an AI that really tries to win? Make one that doesn't cooperate with the human player in any way. It will never trade with the human player, never form alliances with the human player, never interact with the human except to hinder his progress. This AI's interactions with the human player are essentially limited to denouncing him, declaring war on him, spying on him, converting his cities, etc. And that would get old very quickly, wouldn't it?

 

[Veneke]

Sure, it's an antagonist that is doomed to death a priory. And that's exactly kind of role AI is playing as far as many players are concerned. A prop, nothing more.
And again, I agree that terminology is not very accurate. But we get the idea, I think. There are those who play to win and want AI to do the same, so called number crunchers, and there are everybody else.

But... it's not doomed to death in a priory. If we use the three groupings from that article, narrativists would be hugely disappointed if the AI could not win. There's no story without an antagonist to pose a challenge to the hero of the story as he sets about whatever it is he's setting about. There must be a realistic possibility that the hero can lose. If it was that the AI couldn't win none of the setbacks, advancements nor important turning points would hold terribly much significance - what kind of story is that? The AI is a prop, certainly, but it's a very important one - the most crucial one, frankly. If you removed the AI, if you removed the challenge it poses, and went about the game then I'd suggest that you're not really a narrativist, you're probably a simulationist - more interested developing an empire than crafting an exciting story. Of course, the development of an empire in and of itself can be a story, so there is an element of overlap there, but it would not be a primary goal of a simulationist to uncover a good story as the game progresses. It would simply happen across one.

There's certainly room for overlap between all three (or more or whatever) but in terms of the archetypes you are well off the mark when it comes to describing narrativists/role-players/etc as desiring an AI incapable of winning.

 

[The Pilgrim]

Overall I don't think too many people fall absolutely into one circle or another. I think for the vast majority there are areas of overlap (I think the AI at least putting up a struggle is one such area that the majority want) its just a matter of what the tipping point is for most players.

Absolutely. We're discussing extreme cases for simplicity purposes, but the distribution is obviously not discrete.

Once I've mastered it, it loses the fun.

You increase the difficulty.

Which maybe brings up a good area.Randomness. I like it in the right amount. Oops. Earthquake. Out of your control. Couldn't have planned for it. Disease, one you couldn't have planned for. Your game is impacted. To me, this is good. This is a twist in the story. To a gamist this represents an unpredictable outcome that destroys an otherwise perfectly planned and executed strategy. I think most gamists (number crunchers) consider it a waste of a good game.

Very good point. RNG must die!

All I'm saying is I don't really agree that the AI is not trying to win. You can say they suck at it, but to say that they're not trying to win - no. For someone who is not trying to win, the AI wins pretty often, especially at higher difficulties.

It really depends on the way you play. In practice it sits there, waits for the tech tree to end and once it happens builds the spaceship and UN. If human player hasn't won by then, good. Otherwise - oh well... The only thing affected by a difficulty is the time that takes to finish all techs. The fact AI can't win by culture, domination or diplomacy speaks for itself.

I'm also going to have to disagree with your statement that "numbercrunchers" are the only people who play to win and expect the AI to do the same. I don't consider myself a numbercruncher, and I play to win. I expect the AI to do the same, within reason.

I really don't know how to phrase it better, especially since the distinction is so vague. What I mean is that winning is the ultimate goal. The only goal. Every decision made is supposed to bring the player closer to that goal and pretty much everything that fails this criteria can be and is sacrificed, whether it's wonders, diplomacy or whatever the circumstances dictate. Real number crunchers (unlike me) take it even further trying to win by the earliest date possible every time. I don't think this applies to most of the narrativists/rpg'ers/etc. But I will stop trying to predict what these guys like or expect. The range of preferences is indeed too wide.

To illustrate my point: Want to make an AI that really tries to win? Make one that doesn't cooperate with the human player in any way. It will never trade with the human player, never form alliances with the human player, never interact with the human except to hinder his progress. This AI's interactions with the human player are essentially limited to denouncing him, declaring war on him, spying on him, converting his cities, etc. And that would get old very quickly, wouldn't it?

If AI was really Intelligent, it wouldn't, just as MP doesn't. That also wouldn't mean a total lack of cooperation. It can be mutually beneficial sometimes. I guess. It would require a whole different AI and balancing.

There's certainly room for overlap between all three (or more or whatever) but in terms of the archetypes you are well off the mark when it comes to describing narrativists/role-players/etc as desiring an AI incapable of winning.

I don't disagree.

 

[BSPollux]

There's no story without an antagonist to pose a challenge to the hero of the story as he sets about whatever it is he's setting about. There must be a realistic possibility that the hero can lose.

Lets talk about 'winning' in Civ and the climax of a story.

Imagine this: You got a huge empire, its happy, wealthy and educated. Then the little, backwards empire of 'whatever' wins diplo victory.

Its totaly anticlimatic. He shouldnt win, easy as that. If you wanted to stop his vote you could have declared war on him and his citystate allies and killed him with zero threat to yourself. Like an execution. After the vote, you still could do that. Nothing changed. But he 'won'. Its just a label, but its no good ending to the game. But the AI played by the rules and beat you. You definately lost and the AI won, but all it did was to stop the game, without a climax. The AI being a challenge didnt help that game, it ended it on the most boring way.

'Fairness' is a concept ment for games where equal participants challenge each other. But this is not the case in Civ, the AI is no human, its only a tool. Theres no need for the AI to follow the same rules as you do and it doesnt need to have the same goals. Its just one obstacle the game puts in your way. It should be good in opposing the hero, but it shouldnt try to be the hero.

 

[LoneRebel]

I really don't know how to phrase it better, especially since the distinction is so vague. What I mean is that winning is the ultimate goal. The only goal. Every decision made is supposed to bring the player closer to that goal and pretty much everything that fails this criteria can be and is sacrificed, whether it's wonders, diplomacy or whatever the circumstances dictate. Real number crunchers (unlike me) take it even further trying to win by the earliest date possible every time. I don't think this applies to most of the narrativists/rpg'ers/etc. But I will stop trying to predict what these guys like or expect. The range of preferences is indeed too wide.


If AI was really Intelligent, it wouldn't, just as MP doesn't. That also wouldn't mean a total lack of cooperation. It can be mutually beneficial sometimes. I guess. It would require a whole different AI and balancing.

Ah, thank you for clarifying your earlier comments.

Basically, for number crunchers, winning isn't everything, it's the only thing.

The main reason I play games is to have fun. Of course, winning is a big part of having fun. But not the only part.

Regarding an AI trying to win, it should not cooperate with the human player unless its own interests are also advanced. I think the Civ 5 AI does a reasonable job at this. I would even say it's a little extreme the way the AI acts friendly to you and then betrays you the very next turn.

 

[whyidie]

Lets talk about 'winning' in Civ and the climax of a story.

Imagine this: You got a huge empire, its happy, wealthy and educated. Then the little, backwards empire of 'whatever' wins diplo victory.

Its totaly anticlimatic. He shouldnt win, easy as that. If you wanted to stop his vote you could have declared war on him and his citystate allies and killed him with zero threat to yourself. Like an execution. After the vote, you still could do that. Nothing changed. But he 'won'. Its just a label, but its no good ending to the game. But the AI played by the rules and beat you. You definately lost and the AI won, but all it did was to stop the game, without a climax. The AI being a challenge didnt help that game, it ended it on the most boring way.

'Fairness' is a concept ment for games where equal participants challenge each other. But this is not the case in Civ, the AI is no human, its only a tool. Theres no need for the AI to follow the same rules as you do and it doesnt need to have the same goals. Its just one obstacle the game puts in your way. It should be good in opposing the hero, but it shouldnt try to be the hero.

A good narrative doesn't have to be a "heroic epic". Tragedy can be just as compelling.

In your scenario, you present a result. The interesting bit to me is how did it happen ? How does a backwards empire turn into a diplomatic powerhouse ?

 

[Red Pearl]

It is also true that Brave New World is adding layers to the diplomacy and to the diplomatic victory condition. So if someone in BNW wins with diplomacy there will have been a build up to it and it will still be impressive even if they end up with just 1 city.

 

[Veneke]

Lets talk about 'winning' in Civ and the climax of a story.

Imagine this: You got a huge empire, its happy, wealthy and educated. Then the little, backwards empire of 'whatever' wins diplo victory.

Its totaly anticlimatic. He shouldnt win, easy as that. If you wanted to stop his vote you could have declared war on him and his citystate allies and killed him with zero threat to yourself. Like an execution. After the vote, you still could do that. Nothing changed. But he 'won'. Its just a label, but its no good ending to the game. But the AI played by the rules and beat you. You definately lost and the AI won, but all it did was to stop the game, without a climax. The AI being a challenge didnt help that game, it ended it on the most boring way.

...

I'd argue that it's probably a very good ending. It depends on the huge empire's story mind you - why didn't it wipe out that backwater empire? In order for this huge empire to create its utopia did it anger every other civ, obliterate the cities of that backwards empire that won making it the face of the rest of the world?

There's no reason why the player must be the hero, he can be the villain or an anti-hero too. Victory isn't terribly important and it doesn't even really need to be the climax of the story. He may be the leader of the rest of the world but what happens then? You could end the story there or you could carry on as the huge happy empire spreads its ideals to everyone, whether they want them or not!

 

[Red Pearl]

@Red Pearl: I'm not terribly familiar with WoW but ME1 is a RPG, definitely. Civ really isn't. It can be roleplayed but I don't follow the comparison between ME1 and Civ 5.

The civilization series has historical civilizations and leaders. So if you play something as was historically then you role playing.
Arabia: focus on a classical/medieval war and then switch to focusing on trading.

 

[Veneke]

Aye, I get that (though I wouldn't necessarily believe that it's restricted to that) but there's a distinction to be drawn between that and a role-playing game though.

 

[GhostSalsa]

There's no reason why the player must be the hero, he can be the villain or an anti-hero too.

So true. It's very rare to win on Immortal or Deity without being the villain. You start out as the backwater civ and have to take down and starve and dismantle all these largely perfect, utopian societies where happiness and growth are off the charts.

Another reason why I prefer to take scrappy underdog civs into my Immortal games...

 

[Red Pearl]

CiV is not a RPG. I am not number cruncher seeing as I do a very small amount of simple math. But I would put myself in number cruncher category in this thread. Number cruncher in this thread mean that one's primary focus is to win.

 

[Hambil]

CiV is not a RPG. I am not number cruncher seeing as I do a very small amount of simple math. But I would put myself in number cruncher category in this thread. Number cruncher in this thread mean that one's primary focus is to win.

Incorrect. It means one's primary focus is to win by crunching numbers.