[MUD-Dev] narrative

[John Robert Arras]

On Thu, 15 Aug 2002, Brandon J. Van Every wrote:
> John Robert Arras:
 
>> I think the details are what separate good writing from bad
>> writing.  If you want to have compelling narrative, you have to
>> control the details.
 
> Yes!  As the writers say, the tale is in the telling.
 
> To be more mathematically rigorous about it, good writing is a
> global problem with tremendous complexity of elements.

I agree with this. You can't just focus on the details of the
world. You must have a global perspective and the ability to control
things from the world scale down to the details.

> Engineer-heads try to treat it like it's a collection of local
> problems, something you can just chop up into constituent and
> interchangeable components.  As you chop up the writing into
> components, you lose information.  Because the writing contains
> little information, it is not terribly compelling, it feels
> generic.  Some players are highly imaginative and will impose
> their own vision upon the components, but the vast majority of
> consumers will say "This is boring, this sucks."  A real writer
> integrates his components.  The integration is what you still need
> a human being for, and we will have that need until the advent of
> Strong AI.

However... If I had a really good simulation of the world and let it
run, then people experiencing the simulation wouldn't be bored. They
would find interesting things to do because the simulation is
complex enough. Getting a simulation this complex is probably as
hard as getting strong AI, but it would do what I want it to
do. This is where I'm trying to go.

>> On the other hand, if you want to have an interactive game, then
>> you have to lose control of the details. If you want a _really_
>> interactive game, then the player has to be able to affect more
>> than just the details. You probably have to let the whole story
>> go and let the "the story" be created on the fly. I don't see any
>> way to create a compelling "interactive narrative" since the two
>> ideas seem to contradict each other.

> Here, I think you're recreating the commonly received mythology.
> Storyline and interactivity, oil and water, blah blah blah.  An
> alternate thesis: interactivity is a psychological illusion.  You
> do not interact, you go exactly wherever the game designer already
> envisioned you could go.  What about Diablo II disproves this?
> Yet, people say Diablo II is interactive.  What about Donkey Kong
> disproves this?  Nothing.

I still think they're opposites. Diablo II isn't really interactive.
I am slowly coming to the realization that anything where there are
predetermined endpoints or directions I have to go in don't cut it
as "interactive" for me. I will discuss this in the end notes on
randomness.

> When you pursue interactive writing as techniques of illusion,
> possibilities open up.  You do not have to shunt a player down a
> path.  You can restrict him to an area, knowing that he will
> wander around probably for T time before running into the next
> item of interest.  You can control the size of area A and time
> variables T.  You can summon events E to push / pull / prod the
> player along.  Yet the player believes he has interactivity,
> because he has some freedom of movement.  People make the mistake
> of thinking they must grant players *all* freedom of movement.  It
> is only important to grant *some* freedom of movement, that is how
> the illusion of interactivity is maintained.

I want to create a world and let players play in it without me
trying to push or prod them to do anything. I think given enough
complexity and size in the simulation, they won't run out of things
to do since new things will always be happening and there will
always be reasons to go back to places they explore as the world
changes. If I know that what I'm doing here doesn't really matter
and I'm in a holding cell until the gate opens and I can go to the
next cell, I'm not feeling really interactive. The issue is then
that the player is reduced to being an observer (and I think players
can tell this) so the level of effort required to stave off the
boredom is much higher.

> Also, you maintain the illusion by capturing the player's focus.
> If he very much wants to read and follow up on some particular
> thing, then he won't notice that he's *willingly* following the
> storyline.  When you randomize storylines, you throw out the
> writer's best tool, his ability to be fascinating.  Random events
> are fascinating only on occasion.  It's the luck of the draw.
> Much is made of emergent behavior's potential; little is confessed
> of how fun these abortion experiments are in practice.  If you
> really wanted the experiment, you'd be happy as a research
> scientist.  You wouldn't need entertainment!  Entertainment is
> prepackaging, it's compressed experience.  Not the occasional
> exciting discovery in a sea of boring unknowns.  

What story? What focus? The player doesn't have to be an observer
that's eating up the story that you premade. The player can be a
participant. The problem here is that you want the player to follow
YOUR story. I don't think there's a story. Or at least there doesn't
have to be a story. There's a world. Just like our world. If you
make it complex enough, it won't get boring. That's all that I care
about.  Enough interesting things can happen that the players won't
get bored. So it's like research for me. I admit the reason that I
work on this is that it's the most interesting math problem that
I've ever seen. :) Secondly, it's about making the kind of game that
I want to play.

I am not interested in writing stories. I am interested in keeping
the player from being bored. I want to generate enough interesting
events in the world that the players will no get bored. That's
it. It's not the same thing as writing stories, but it doesn't mean
that it's automatically boring, either.

Let me end this with a question about randomness. What is the most
complex random "quest generator" or "event generator" that you can
think of for a world? I get the impression from your counterexamples
that you're thinking about this in terms of random encounters that
just get poofed out of nowhere and don't mean anything. I am not
talking about random encounters in this sense. Those things are
shallow and don't come close to the complexity of what I'm talking
about.

Imagine Diablo II where every map in the entire game is loaded into
memory at once and every creature on every map exists at all times.
The creatures work together within their local groups to try to take
over or rule their little part of the world. Imagine that they lash
out and raid other parts of the world and try to take over the whole
world. They build cities and take over cities and destroy each
others' cities and they make alliances with each other. Then,
imagine that the world is much larger. If I played in a game like
that where I started out as a nobody and the whole world was dynamic
and there were ways for me to change the world (slowly and with
great effort), I doubt that I would ever get bored. Since players
will always overrun AI's eventually, add in a few factions of
players that each try to take over the world and control the armies
and societies and kingdoms of the world, and you have a game that
won't get boring. Everything you do makes a difference and you
really don't know what's going to happen. And the game is large
enough that you can't completely change the world without a lot of
long-term effort being expended.

It's not storytelling or narrative or writing, but I also think that
it's not boring. This is the direction I want to take and this is
not just pie-in-the-sky talk. This is really what I've done with my
codebase and the main thing holding me back from opening it is that
I hate building zones and I don't have enough of a world created to
have the size and complexity that I want. It's about creating a huge
number of global events (that follow from within a simulation) so
that the players can't possibly predict or even know what to expect
far into the future. It's about making something where players can
really change the world that they're experiencing. It's not writing,
but that doesn't mean that it's boring, because players do get
strong interactivity.

John


_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev