next »

[Dave Ravel]

I love the example of Mount and Blade. The first dozen hours I spent in that game were magical, I was in this massive world fighting great battles, laying siege to castles, pillaging the countryside with massive warbands, and winning jousting tournaments. I think I stopped playing once I realized that even though the map was huge, the possibility space and significance of the interactions were pretty limited. Beyond trading, which was the only thing I left out above, there isn't much variety to what you do. Mount and Blade turned out to be a few good ideas replicated across the board in a non flexible fashion. Every king is fighting a war that never truly ends. Every king has one usurper that needs your help. I'd say that's a game that failed to live up to my expectations specifically because they avoided having any sense of story or consequences. There are plenty of things to be interacting with but there is no weight to the interactions. Why should I retreat to save a village being pillaged nearby when I could just ravage these trade caravans? There are no meaningful consequences since the war goes back and forth perpetually, the NPCs aren't very significant to gameplay or story, and the quests are just glorified fetch quests. The most significant choice you make is to either support a king or his usurper, but what is the difference between the different kings and usurpers?

This game was definitely given attributes of sandbox games that hurt it, namely favoring game balance at the cost of doing truly interesting things, such as conquering and exterminating the Rhodok Empire. The number of armies Rhodok has can never really be less than 2/3 to 1/2 of yours. You also can't kill Lords, which I assume is to protect gameplay balance. I think it comes down to the fact that the amount of change, which I believe is a very significant interaction, you can bring about is very small. The gameplay teaches you that you are no more important than any of the other hundred Lords in the game, and there isn't any narrative to tell you otherwise. I don't think an authored narrative would've been needed if the player's own story was interesting enough. Perhaps you could add in story elements that would add some meaning to the player's story, but that would feel kind of tacked on. The lack of weight and significance of choice undid this great sandbox experience.

I think you can approach narrative from a new direction, one that I was alluding to in my previous post. The concept of a unit operation is that an author gathers together elements to create a discrete unit of meaning, a specific thing or things that express something. Since the brain loves analogy and often has emotions/feelings triggered by things which are analogous to their own subjective experience the goal would be to use unit operations to create something that, through the players interactions, would express something important or meaningful. You can do this with traditional stories, or you can go the route Jon Blow is interested in and that is by infusing these units of meaning into the gameplay, to create significant interactions from a gameplay perspective rather than purely from a narrative aspect. Creating meaningful gameplay on a larger level than Gravitation is difficult, but I think it's possible. Maybe you can't express so much in so little, but the least you could do is create gameplay that informs the narrative, instead of being discordant or unimportant.

I'm going to return to my example of Braid as one of the few games I've played that has achieved a state of gameplay and narrative that are greater than the sum of their parts. Other examples of this would be Shadow of the Colossus, where the gameplay and narrative build off each other and makes your interaction with the game much more significant. Braid possesses six worlds, each world has a number of books telling small Calvinan chunks of story. Many people (including Paul I think) read the story as being separate from the game, it was just Tim's story, nothing more. The gameplay actually informs the story because the units of meaning that an individual world's story communicates are inherently tied into the time manipulation rules at work in that specific world. The gameplay informs the story and the story informs the player. If you think of the story as themes and ideas, a mental state if you will, then you can begin interpreting the gameplay itself as expressive, communicating other units of meaning strictly through the puzzles and time manipulation mechanics, which you can tune into if you are in a conducive mental state. This is one of the most interesting uses of story I've seen so far in a game, because even though most people distance Braid from the likes of Passage, Gravitation, Marriage, and Today I Die (there are many others) it is actually doing a lot of the same things, but on a subtler and larger level, and also with the aid of a narrative. Sadly this is an example of a non interactive story as far as player choice goes, but what it does possess is the need for subjective interpretation, which is I think a very interesting thing that both interactive and non interactive narratives can achieve.

Sadly I don't have an example of a game with an interactive narrative that achieved anything beyond fun dialogue trees to explore. I was thinking of including something like Pathways but even though you have a choice in the order of events you experience you really have no power over which events you experience, and I think that devolves into a very inflexible interactive story, because there is no exchange of elements to be included, just an interchange of the order, which could be meaningful, but in this case it really doesn't communicate anything differently based on the order. Oh god I feel like Socrates, asking a question and then rambling without ever really answering it. Oh well I have hope for this discussion and I look forward to your responses.

[Alex Vostrov]

Dragon Speech Part 4 of 5 "Characters" (1992)

(but, imho, human simulator == unethical. alienation through medial phantom pyschology... maybe it's good that isn't possible (yet). but i personally don't think characters in interactive "story spaces" would really need that level of realsitic psychological simulation. it would be enough to use simpler, mental states, which would have a function for the actual story. more authored than engineered...)

I'm quite familiar with Crawford's work, 0rel.

I'm not sure what you mean by "medial phantom pyschology."  Perhaps you could elaborate your thoughts a bit?

[0rel]

I'm not sure what you mean by "medial phantom pyschology."  Perhaps you could elaborate your thoughts a bit?

that was more a personal side note. assuming, one day it would be possible to simulate how humans behave (look, speak, act, think, learn...) through advanced AI/audio-visual techniques, and new sort of more natural of interaction device, that would be a really dangerous innovation, in my eyes. however, that's pure fiction... still, i think that the "interactive storytelling fraction" is clearly heading in that direction, because having a tool to create artificial actors (with artificial brain, which could be filled with a arbitrary memory), procedural character modeler and things like that, would immediately enable very interesting (but mentally/socially highly dangerous) kinds of "story spaces". play in an artificial 'Big Brother' show with your AI friends, or other open social simulations... personally, i would be all against that sort of non-linear computer dramas.

on the other hand, there is the clearly authored way of story telling. like in mostly all games so far, where the player doesn't expect the characters to behave on their own will. he knows that he plays in a written story, and that things have to make sense in some way. - the story is medium for communication.

in truly non-linear, interactive stories, though, the simulation would have to be engineered in a very open/procedural way, so that the author(s) would loose so much control, that the game would become a reality of its own... but i think, that good stories must always communicate human values in some way, through well chosen characters/dialogs/places/plots....
a game like 'The Sims', as one of the interesting examples, builds on a clearly engineered model of modern, civilized life. the system tells the player what's good/bad through the rules/values. now, if the players experience their "own" stories in the game, they will never be really authored/meaningful, in so far, that they have no real plot, and conflict, deeper content... but they will always be a secondary effect of the underlying "engineered reality model"...

hmm... i didn't want to write that much... but now i even think that these kind of open story games could be interesting as well, when the underlying model would be made in an poignant way. maybe it's exactly what "dynamic meaning" is about... i don't really understand it yet...

The biggest problem is that a linear narrative is not amenable to player modification. In other words, linear stories don't listen or think. What they do very well and powerfully is speak. [...] The carefully crafted sequence of events is impossible if the player can just mess with it at will.

I see that these two forces of narrative and agency are fundamentally in conflict.  Therefore, mixing them is a bit like mixing water and oil.

that made me really think more about the topic. that's kind of the center of the problem, for me as well.

[Paul Eres]

I feel a bit like you're ignoring the specific points of my argument.  If my definitions are unsatisfactory to you, you should point out why.  Dismissively waving it away doesn't communicate much.  My reasoning was a bit more developed than "a pure linear story is not interactive."  I'm making a bit of effort to understand your viewpoint, Paul.  So far, I haven't gotten anything that I can work with.

I feel the same way. I think it's weird to say that you don't understand my viewpoint when I already wrote several pages about it (such as the quotes quoted in the first post of this thread). I'm not exactly sure how to be more clear than that.

It's also weird that you say I don't understand your viewpoint; I think I do. I don't agree with it, but that's not the same as not understanding it. I didn't intend to dismissively wave it away; I doubt anyone in this forum has read more about interactive storytelling or tried it more than I have, so it's not that I'm being dismissive of the idea.

But again, to summarize, this is what I mean: I felt that Blow was recommending against using traditional storytelling methods because he felt that they are fundamentally incompatible with interactive media and that if a game developer wants to evoke emotion they should not use linear stories. I disagree with that, because I feel that interactivity can be used and in fact has been used to make traditional storytelling methods better, and that interactivity doesn't replace storytelling, but adds to it.

[Alex Vostrov]

I feel the same way. I think it's weird to say that you don't understand my viewpoint when I already wrote several pages about it (such as the quotes quoted in the first post of this thread). I'm not exactly sure how to be more clear than that.

I just haven't been able to build a mental model of your ideas.  I understand that you think that narrative and interactivity are not in conflict.  You think that they can co-exist and enhance each other.  My problem is that from my point of view they very obviously DO collide.  My issue then is trying to understand why you think they don't.

I'll re-read your comments to see if things are clearer a second time around. 

It's also weird that you say I don't understand your viewpoint; I think I do. I don't agree with it, but that's not the same as not understanding it. I didn't intend to dismissively wave it away; I doubt anyone in this forum has read more about interactive storytelling or tried it more than I have, so it's not that I'm being dismissive of the idea.

I haven't claimed that you don't understand the issue.  It sounds to me like we have very different basic assumptions, which makes communicating difficult.  What I was bothered by is that I laid out my chain of reasoning and instead of addressing its specific flaws you argued against abstract theorizing.  That doesn't help me to understand WHY you see my reasoning as wrong.

Thanks for being patient, I know being in discussions like these can be a bit frustrating sometimes.

But again, to summarize, this is what I mean: I felt that Blow was recommending against using traditional storytelling methods because he felt that they are fundamentally incompatible with interactive media and that if a game developer wants to evoke emotion they should not use linear stories. I disagree with that, because I feel that interactivity can be used and in fact has been used to make traditional storytelling methods better, and that interactivity doesn't replace storytelling, but adds to it.

What I'm wondering is why you think this.  Haven't you said before that maximizing player choice will essentially un-make the game?  If this is the case, shouldn't it destroy linear narrative first, due to the inflexibility of pre-specified plot?

[0rel]

'Indigo Prophecy' was one of the most remarkable games, i've played so far, which conveyed a quite striking story in an interactive way. the interactivity was always integrated into the story, and still very varied. i mean, it was even promoted as first interactive film...

although i quite liked it, there's to mention that, on the downside, the interaction is completely laid out, and you have almost no real freedom (pseudo non-linear). also, the button hitting in the action scenes is kind of stupid... otherwise it works pretty well, imo. the interaction makes it more immersive, but otherwise it's a linear story.
(would be worth trying out @topic...)

whether linear or not, as human story telling medium, i think, games have at least one large problem: dialogs, the primary way of human interaction. you can't really play with the characters in an intuitive way without using words... in a story like in 'Half-Life 2', where the characters and the environment look and behave very realistic, that flaw is visible at a its best: the protagonist will always be silent during the whole long story. that's not really a big problem in an action game, but in a game like 'Facade', i think it's really not acceptable... typing in words or choosing between possible answers works to a certain degree (esp. when the graphics/sounds are not too realistic), but i think, that's one of the bigger problems with stories in games... expressing emotions through body language/facial expressions and even the simulation of mental processes can already be done in a convincing way by the industries, but the actual interaction with the persons not really, as far as i know...

so, a simple conclusion could be: "don't do interactive stories with words.". all three requirements: representation of mental states (expression through audio-visuals) / some kind of mental processing (behavior of the actors) / and even direct interaction (without spoken language in that case), could be achieved...

And the reason I think it's silly to say "if you want to make a game with a story, write a book instead! stories in games can't be as good as books!" is obvious if you think about it. Compare with these: "if you want to make graphics, paint a painting instead! graphics in games can't be as good as paintings!" or "if you want to make music for a game, just put it on a CD instead and release it as an album! music in games can't be as good as music released as an album! Or better yet, perform it live in front of an audience!"

that's great! that's actually what makes games so interesting to me... when all elements can be put together in a new interactive way, the whole can be more than the sum of its parts... also the story.

I feel that interactivity can be used and in fact has been used to make traditional storytelling methods better, and that interactivity doesn't replace storytelling, but adds to it.

the second link after typing "interactive storytelling" into google lead me to this paper, from this page...
i've read only some bits of it, but it seems to go into that direction.

[Paul Eres]

re Alex:

Ah, understandable.

I think that you've only played one of the games on my top 10 list of favorite stories in games may contribute to you not knowing how interactivity can add to stories. But you've probably least played some games you enjoyed because of their stories.

I'm also troubled by your rejection of empiricism. I believe empiricism is the only trusted way to deal with reality: the evidence of the senses and the collection of data is all we have, the conceptual mind is impotent in comparison and mostly produces fantasy and illusion. But that's a different subject. But don't take my rejection of deduction in general as a rejection of your line of reasoning in particular: I reject lines of reasoning period. I don't feel that lines of reasoning are to be trusted when they contradict experience and empirical data.

In particular, it's a bad idea to me to reject the idea that interaction and traditional storytelling can work together just on the basis of definitions of interactivity and stories when clearly there are working examples (many, many games) where interactivity and storytelling are not in conflict. Rejecting empirical data like that strikes me as akin to a religion, where you take on faith that interactivity and storytelling contradict and ignore all evidence to the contrary or interpret it in such a way that they work with your definitions, usually by saying the people who enjoy them and find no problems with them are actually deluded into liking things which they should not like. And you do seem to take that line of approach with the "this game's story would be better in a book or movie" and other constructs -- just dismissing that they exist and saying they shouldn't because of some imaginary contradiction between interactivity and storytelling, even though millions of people are enjoying partially interactive traditionally told stories every single day.

Conceptually, I think the key ingredient you're missing from your model of my model is that I don't conflate story and plot. Stories don't necessarily have to contain a set plot, they can have a partially set one or even be plotless. Stories also include characters and setting. A lot of people say that World of Warcraft for instance "has no story" -- but it does, it just doesn't have a plot. But it has a setting -- one I find kind of cliche and boring, but it's a setting -- and it has a lot of characters, most of them players (which makes the story interactive). So I feel that if you think of story as characters + setting + plot rather than just plot, you won't see a contradiction between interactivity and stories.

[Lucaz]

On interaction damaging storytelling, I don't think so. Any good narration that's important in a game is merged with the gameplay. Even if the player interaction weakens the story, it strenghtens the storytelling, as the player is now a more integral part of it, and it also strenghtens the gameplay as it makes it more meaningful and deeper. As Paul says, storytelling is made multiple parts, so weakening the story doesn't necesarilly weakens the narration.

Also it has to be remembered that a story that works in a game isn't necesarilly good when looked at in the same you look at a book or movie, and vice versa.

@Orel: One way to go about procedural dialogue is to have an ample but limited array of generic subjects to talk about. The dialogue syste, in Maupiti Island will help me explain what I mean. The main character doesn't has any dialogue line, but you have a list of subjects to ask about, that includes every character, event, place and moment. Something like that, maybe with the plus of a text parser, and some slightly complex algorithm to get the character reactions, would allow for an ample and worthy dialogue. Maybe no the deepest one, but good enough.

[Lyx]

It's a very hard problem and a lot of people working in the industry don't naturally lean toward solving it.  What is needed is to develop algorithms to simulate a human being.  You need to have characters that have emotions and can act on them in a plausable fashion.  We've barely started on that problem.

What i was trying to show was that in some cases, it is not difficult at all. When it comes to simulating such things, the desired "deepness" matters. In a sandbox game with a very big world for example, it isn't necessary to have "deep" interactions - or at least, not often. Take for example a quest template in a roguelike - it doesn't need to have deep dialogue. It doesn't need to have a complicated backstory. All it needs is to have shallow variety. And if possible, it needs to have consequences.

Perhaps i wasn't making it clear enough in my post, that i was making a 180 degree turn and looking at the topic from an opposite angle. Before, we were talking about generating deep NPC interaction. But in the post, i was talking about having a high quantity and variety of interaction, but each of them not very deep. For games like knytt stories, mount'n blade or roguelikes, it isn't necessary to have deep complicated interactions - rather, the magic could come from being able to interact with many inhabitants in many different ways - and those interactions then influencing each other via their consequences, so that some kind of emergent story is created.

Let me explain this with my favorite target (as in "i like to bash them for the problem, because it is so ridicously trivial in their case") - quests in roguelikes. A roguelike has all its items and npcs nicely tagged and classified in all kinds of ways, and it has sophisticated means to generate those. Correct me if i'm wrong, but to me this practically means that the developer - practically (not necessarily technically) - has a SEARCHABLE DATABASE of modular game content right there in front of him. Often, even all important entities in the gameworld are accessible regardless of the player location. The developer also already knows how to create meaningful items and npcs via placeholder variables. The way how this works, is that a content-entity template has no static definition. Rather, parts of it consist of variables which define *selective randomization*. It works like a fill-in-the-blanks puzzle.

In other words: He has already all the prerequisites there, to generate meaningful quests. Whats missing? What would be left to do? Well...

1. The mental step to recognize, that you can do the same with quests.

2. The intention of doing it.

3. The additional step to recognize, that you can do this in a more complex way than is typically done in some games. You dont need to stop at randomizing some NPCs, monsters and items. You can also vary the greeting a bit (heck, you can even look at the player stats and make it apply to the player - i.e., different greeting depending on fighter or mage), vary trivial parts of the phrasing a bit, vary a minimalistic (1-3 sentences) backstory a bit. Heck, you can even go as far as not just generating fire-and-forget quests that way, but even full multiplart subquests - you'd just need to first create a few subquests by hand, then look at how they are structured and recognize that they are kinda modular structured, et voila, just vary the modules a bit.

4. Someone talented and creative in coming up with ideas... seriously, i think that THIS in practice is the most difficult part of it all - not the technical requirements, but finding someone who is talented in designing modular quests.

5. Going one step further: Make the result of the quest modify variables in the gameworld. Then make it so that which quests are spawned depends on those variables in the gameworld. Depending on how complex you go there, this way you get the mentioned "emergent mainstory".

So, yes, if we're talking about deep emotional interaction and story, then the problems which you mentioned are faced. But they are mostly irrelevant if all you want to do, is generate some varied but shallow textual content.

- Lyx

[Lyx]

P.S.: Again, i just remembered that there was a game which did just that. It is among the gradfather family of sandbox games. It was Elite II / Frontier :-) Check out the blackboard :-) However, it stopped at adding the "consequences" and "emergent mainstory" aspect.

P.S.2: This by the way is the main reason, why none of the pseudo-remakes made me go "wow!". The X-Series and the like didn't impress me. Why? Because they felt sterile and cold. They felt as if you were interacting with boring machines all the time. It felt lonely and dead. In frontier, the character portraits were selectively randomized. The quests were in their phrasing selectively randomized. The game was given a feel of being part of a LIVING world, full of lifeforms, not machines. None of the remakes managed to create this impression. The coffeebreak world of "Adventures in infinite spaces" feels more alive to me, than all elite/frontier remakes taken together.

[Alex Vostrov]

I think that you've only played one of the games on my top 10 list of favorite stories in games may contribute to you not knowing how interactivity can add to stories. But you've probably least played some games you enjoyed because of their stories.

I've played many games where the story kept me interested.  The problem was that in many of them, the actual game was merely an obstacle between me and the rest of the story.  This is a bit of a different topic, but it's something that's of great concern to me.  I'm troubled by the possibility that a lot of our design tricks are nothing more than operant conditioning.  We don't want to be addicting our audience to our games, a topic that Jonathan also talks about, by the way.

I'm also troubled by your rejection of empiricism. I believe empiricism is the only trusted way to deal with reality: the evidence of the senses and the collection of data is all we have, the conceptual mind is impotent in comparison and mostly produces fantasy and illusion.

Ah, but when we stand in the domain of fantasy and illusion such tools are appropriate no?  I'm not 100% sure what you mean by empericism when you talk about games, but I can venture a guess.  I suspect that what you mean is that we should rely on examining what has and hasn't been succesful in the past.  My concern with this approach is that it imposes incremental innovation, at best.

To transcend the limitations of current games, we need to leap over them conceptually.  This approach is not so disconnected from reality as you might think.  Let me give an example from physics.  What is the electro-magnetic field?  It is a thing that we cannot directly see or touch, and yet the field is an idea that informs our understanding of the world.  Using this concept, we can make predictions about observations that have not yet been done.  Any time a physicist calculates on paper the force between a wire carrying a current and a charged particle, he is utilizing conceptual reasoning.  Neither the particle, the wire nor the field are strictly "real" objects.  They're merely ideas that are convenient to us.  We could just as easily imagine the wire as a collection of atoms or the atoms as more elementary particles.  Without such abstractions we would not be able to make generalized predictions about the world.

It is no different when we talk about games.  We can imagine them to have abstract qualities such as "interactivity" or "linear narrative."  We can then employ our understanding of these concepts to make predictions.  These predictions still need to be tested empirically, but to reject this method in general would be unwise, I think.

In particular, it's a bad idea to me to reject the idea that interaction and traditional storytelling can work together just on the basis of definitions of interactivity and stories when clearly there are working examples (many, many games) where interactivity and storytelling are not in conflict. Rejecting empirical data like that strikes me as akin to a religion, where you take on faith that interactivity and storytelling contradict and ignore all evidence to the contrary or interpret it in such a way that they work with your definitions, usually by saying the people who enjoy them and find no problems with them are actually deluded into liking things which they should not like.

I've never claimed that such approaches cannot function.  To the contrary, I have played many games where I greatly enjoyed the story element.  Sometimes, a well-made story has touched me emotionally.  I do claim however, that they're deeply conflicted works and that we must resolve that conflict for the sake of the medium.

And you do seem to take that line of approach with the "this game's story would be better in a book or movie" and other constructs -- just dismissing that they exist and saying they shouldn't because of some imaginary contradiction between interactivity and storytelling, even though millions of people are enjoying partially interactive traditionally told stories every single day.

I believe that one should work with the grain of a medium, not against.  Just as a skilled woodworker embraces the irregularities of his material, we should embrace the peciluar properties of ours.  Every artist in other media has an innate understanding of their strengths and limitations.  If we ever want to reach the same level, we must gain this wisdom as well.

Conceptually, I think the key ingredient you're missing from your model of my model is that I don't conflate story and plot. Stories don't necessarily have to contain a set plot, they can have a partially set one or even be plotless. Stories also include characters and setting. A lot of people say that World of Warcraft for instance "has no story" -- but it does, it just doesn't have a plot. But it has a setting -- one I find kind of cliche and boring, but it's a setting -- and it has a lot of characters, most of them players (which makes the story interactive). So I feel that if you think of story as characters + setting + plot rather than just plot, you won't see a contradiction between interactivity and stories.

Your definition of a story is different than mine, but once we get past that minor detail, I agree with you.  There's nothing stopping interactive stories from having great characterization and setting.  As a matter of fact, I suspect that eventually our ability to describe character will be greater than static media.

[Alex Vostrov]

Let me explain this with my favorite target (as in "i like to bash them for the problem, because it is so ridicously trivial in their case") - quests in roguelikes. A roguelike has all its items and npcs nicely tagged and classified in all kinds of ways, and it has sophisticated means to generate those. Correct me if i'm wrong, but to me this practically means that the developer - practically (not necessarily technically) - has a SEARCHABLE DATABASE of modular game content right there in front of him. Often, even all important entities in the gameworld are accessible regardless of the player location. The developer also already knows how to create meaningful items and npcs via placeholder variables. The way how this works, is that a content-entity template has no static definition. Rather, parts of it consist of variables which define *selective randomization*. It works like a fill-in-the-blanks puzzle.

Do you mean something like the following?

Let's say we have five types of creatures in the game, 4 roles and 3 types of objects:

Creatures: Human, yeti, orc, elf and demon
Roles: Pirate, king, warrior and mage
Objects: Sword, ring and chest

Then, if you wanted to generate a quest, you'd do something like this:

Hi wanderer.  A terrible thing happened recently!  A <role> of the <race> has recently stolen our only magic <object>.  If you return it to us, we will be very grateful.

Maybe once you did the quest, the questgiver would like you more and give you different quests.  The guys that you killed would like you less.

Do I have the right idea, or were you thinking about something else?

[Alex Vostrov]

On interaction damaging storytelling, I don't think so. Any good narration that's important in a game is merged with the gameplay.

The big problem is that we have developed narration and gameplay to different levels.  We've had thousands of years to develop our narrative skill.  We've only had 30 years or so to develop procedural expression.

The topics that we can express best procedurally are resource management and power projection through space.  These are the subjects that we've spent the most energy on.  So, if you have a narrative about the pain of losing a loved one and the gameplay is about shooting monsters in corridors, is it any surprise that the result is schizophrenic?

If you want to have the two merge together, we need to be able to talk about meaningful topics procedurally.

[Lyx]

@Alex:

Yes, i mean that approach, but my most important point (besides of "consequences" and "emergence" - will get back at that later) was that you can do more than that. Developers of such games often are focussed on their entities and rules in the gameworld and (!!!) consider purely story-related aspects as "unimportant". Think about that: A developer wants to generate stories, and then considers just that aspect ... unimportant?!? I suspect that even at this point, it may still not be obvious what i'm talking about. I'm talking about all this (marked in red):

"Hi wanderer.  A terrible thing happened recently!  A <role> of the <race> has recently stolen our only magic <object>.  If you return it to us, we will be very grateful."

I mean all those things which you didn't make a variable, because it makes no difference to the "gameplay". The thing is: This IS important! To the gameplay, it may make no difference which phrasing you use to explain the quest-mechanic - but from a story-POV, it does. It not just does because reading the exact same phrasing again and again is repetitive (though, thats relevant too).

To understand this, think about what the above quest-description is. It is a story-summary. Now, think about popular stories in books and movies. Compare their "main plots". You will see a damn lot of repetition. Many stories and books - when looked at from a gameplay-mechanic perspective - are identical - and yet people dont interprete them as identical?!? When i point this out to game developers, i often get the reply:

"We are not interested in adding stuff which makes no difference to the gameplay. The fact is: There aren't that many types of generic quests. So its not really easy to add more variety".

Well doh, You missed the point of a story, hacker! Differences in all the minor details IS what gives stories variety. Plus: Things which are meaningless to the gameplay, may very well make a difference to the story. When a player does a quest and is interested in story, then he doesn't just want to know from where to where he needs to carry an item - he wants to know WHY, for WHOM, etc. - even if that makes no difference to gameplay-mechanics.

To summarize: If you only care about gameplay-mechanics, then you will create something which is only about the gameplay mechanics. With this, i do not seek to imply that story-aspects should be SEPERATED from gameplay mechanics, so that both may be in conflict. No, i just mean that to a story, more than gameplay mechanics matters - and that therefore, in a story - even if its just a few sentences long - there are aspects which may not change the gameplay, yet still change the perceived "whole". What i'm talking about here, is a synthetic process: Two agents interact and the interaction creates a higherlevel entity - the relationship. In this case, the gameplay aspects and the story aspects are perceived as one compound entity. If you vary the story-aspects you get more variety in compound entities. You just shouldn't rely ONLY on variation in story-related aspects, because if in the game you have only story variation but no gameplay variation, then the player notices that, and therefore again perceives story and gameplay aspects as seperate.

[Alex Vostrov]

By the way, I think that this article by Jason Rohrer has relevance.  It's almost a year old, so you might have seen it already.  If not, you should check it out.

http://www.escapistmagazine.com/articles/view/issues/issue_155/4987-The-Game-Design-of-Art