re: Content

[WFreitag@aol.com]

>Just as you don't force people into "Quick,
>save the princess!" as they enter the theme park, at each stage in a
>specific plotline you don't force only 1 option on the player.  You give
>them several alternatives and they pick for themselves what they want to do.
>This is just simple branching.

Simple branching is too expensive. You say so yourself, a few paragraphs down 
(albeit in regard to computer games, not theme parks). Except in live games, 
choose your own adventure books, sometimes in IF, and occasionally in paper 
games, I've never seen a branching project stick to its guns and allow 
participants to freely choose preferred branches and disregard others. What 
the exceptions have in common is that they're the media in which the 
incremental cost of developing an additional branch is very low. In other 
media, branching schemes inevitably get subverted into "sooner or later, the 
player has to do everything" schemes. The much-feared combinatorical 
explosion of the number of branches never even has a chance to happen. By the 
time efficiency (content encountered in one pass / total content) drops below 
.5, someone, whether it's management or the creator's own ego, starts whining 
"but if they don't do that part, all this great content is going to waste." 
Then in come the locked doors and hidden keys.

>Massaging people to get them to replay would be a *big*
>skill.  In a theme park, I'd pay certain people a lot of money to reassure
>people, get them socially re-engaged, end their crankiness, and get them
>happy with something more to their liking.  That would be the customer
>service angle.

In live games we did exactly that (except for the part about paying them 
money). We called 'em gamemothers, and gamemothering quickly developed its 
own craft (plus a lot of really interesting war stories). In later years most 
events gave their gamemothers well-deserved equal billing with the authors 
and gamemasters. (Gamemother is a gender-neutral term, by the way.)

Problem: outside the realm of volunteer hobbyists, gamemothers are expensive 
because the typical reassure/reengage intervention takes time, and no economy 
of scale appears possible. A physical theme park _might_ be able to afford to 
add the cost of twenty minutes of gamemothering time average per customer to 
its cost of doing business (and from there to its ticket prices). For a 
hypothetical recreational virtual world operating on even lower margins, it's 
a deal killer.

I'm not saying you're wrong about the importance of live help in such a 
setting. What I'm saying is that the fact that you're right makes the 
business model even more dicey.

>Computer games are very much limited by their budgets.  Even if you
>understand how one could put things together with more branches, the
>production expense to polish all of those branches is prohibitive. 

Agreed, as discussed above. I have no reason to believe it wouldn't also 
prove prohibitively expensive for theme parks, present technology virtual 
worlds, or holodeck modules.

>Thus, I
>see the problem of an engine like the Erasmatron is not simply to provide a
>lot of automatically generated possibilities.  It must provide automatically
>generated *polished* possibilities.

Not necessarily. Linearity and the need for polish reinforce each other. If 
someone says "Come watch my presentation, I have something interesting to 
show you," then I expect a polished presentation and I'll be impatient with 
any unnecessary digressions, unclarity, or delays while the speaker stops to 
think. But if I were to sit down for an impromptu conversation on the same 
subject with the same person, I'd expect all those things. In fact, if the 
person's conversation started getting too polished, I'd think "hey, why is 
this guy lecturing me?"

Get the audience into the mindset of participants, and the demand for polish 
diminishes, even for those who don't participate actively (immersed 
spectatorship). Nothing in the entertainment universe was ever _less_ 
polished than some of my live role playing events. But no one ever asked for 
a refund. Participants, because they _were_ participants, got thoroughly 
involved and enjoyed themselves immensely despite the lack of polish. And, I 
think it's important to note, paid up to several hundred dollars (mostly in 
travel expenses) per event to do it, suggesting that it wasn't just 
good-natured tolerance of well-meaning amateurs that kept them happy.

>> But no attempt is made to steer an individual player
>> into a given path.
>
>Sure.  Why should you?  The idea that you can mindread the player, by
>measuring the number of mouseclicks he's making, the speed at which he's
>entering or exiting rooms, the number of buttons he's pushing or words
>he's using, is completely bogus.  You can't read the player's mind through a
>computer interface.  An expert human conversationalist can't often do it,
>sowhy would you expect a dumb computer to?  All you can do is pester the
>player with questions, eliciting his values, *asking* what he thinks about
>this and that, and hoping that he's answering in good faith.  And if your
>questions would be annoying right now, better not be pinging the player
>with them.  Let the player ping you.  Given them the interfaces to tell you
>what they want.  That's often as simple as having a branch point.

Yes yes yes. And spoken with the passion of someone who's been through "The 
Sounding" (in Greg Roach's Vortex) or some other similarly infuriating CD-ROM 
experience.

One example: in our design for an Arabian Nights world, we (Barbara and I) 
knew that temples would be important to the game milieu, but we needed 
something special that players could do there. The answer we came up with is 
that they can pray for things. If a player does so (it's completely 
optional), the world's event manager pays attention to what that player 
chooses to pray for and adjusts its future actions upon that player 
accordingly.

However, though I agree with you about vague psychological inferences based 
on mouse click speed or answers to intrusive questions, I think even a dumb 
computer can make useful rough inferences based on what a player actually 
_does_. Solve the puzzle, ignore the puzzle, or try and fail? Fight, sneak, 
bluff, or run? I'm not trying to peer into someone's soul, just guess with 
better than random-chance accuracy whether they like puzzles or not or 
whether they prefer to work cooperatively with others or not. Or rather, 
whether or not they feel like doing puzzles or working with others _today_.

Barbara and I have discussed how we would implement the Sorting Hat if we 
were to build an interactive world inspired by (but definitiely not based on, 
no sir, mister corporate attorney) a certain school for wizards. The four 
houses represent four fairly easily recognizable and distinct playing styles 
-- in fact, they map pretty close to those five paths in Spaceport Adeline. 
So how does the Hat know which house to assign someone to? The best answer 
we've come up with so far: First, ask them. But if they choose to let the Hat 
decide, and many will, then base the decision on how they've reacted to a few 
specific test situations they've encountered early on.

>> Every path competes for the attention of every player, on
>> the theory that players will choose the path that appeals to their style.

>Of course, another theory is that players will attempt to exhaust the
>available content. They'll clean out the dungeon levels in order to see all
>the nooks and crannies, so that they haven't missed anything.  Thus, you
>will get players that are going through certain paths, not because it agrees
>with their playing style, but because it's there.  If you can do anything to
>control/redirect their crankiness when they inevitably get pissed off,
>that's beneficial.

>Similarly, players will force themselves to go through things they feel
>"morally expected" to go through.  They may have adversities to "cheating,"
>or not want to appear to be a "wimp."  It takes either a certain maturity or
>personality to simply abandon things that frustrate or annoy you.  Some
>people will force themselves through it anyway, sometimes out of incredulity
>that anything could deliberately try to irritate them that much.

I think computer game players are trained to behave in these ways. Enough 
experiences with getting killed on level 28 because you didn't pick up the 
toothpick on the bathroom floor on level 4 will give anyone a mania for 
doggedly pursuing every possible path no matter how frustrating. It might not 
be training, it might instead be a process of selection, but either way I'm 
certain that computer gamers do differ from people who don't play computer 
games (even those who do play other kinds of games) in how they approach 
things. Yet another benefit of doing live games is that their audiences 
included all types, and the differences are wonderfully obvious. (Of course, 
individuals differ; obviously I'm speaking of the general case.)

Put a non-gamer into a fantasy world, give him a sword and shield, and ten 
minutes later inform him about a nearby dragon. The non-gamer will 
immediately rush off to slay the dragon. And will be extremely angry if the 
dragon kills him. Meanwhile the gamers will snicker at how incredibly stupid 
the guy was for doing such an incredibly stupid thing, how well-deserved his 
humiliating defeat and expensive death, how he'll know better next time, in 
the unlikely event that he comes back to play again.

I understand the gamers' point of view, but my sympathies are with the 
non-gamer. To me what's incredibly stupid is the idea that you're in a 
fantasy world with a sword and shield, and you're _not_ supposed to try to 
kill the dragon. Why punish the player for acting bravely, impulsively, and 
in a manner thoroughly consistent with the cinematic traditions of the genre? 
Why not reward him instead?

Saying such things to game developers has elicited reactions ranging from 
stony silence to accusations of trying to deliberately sabotage their 
projects with my obviously bad advice. The exact same reactions one could 
imagine upon advising a medieval lord to let the peasants farm their own 
land. Including "Oh, no, they don't _want_ to do that, they're happy the way 
things are." Which may very well be superficially true, for both the peasants 
and the gamers. Change is frightening, after all. Or maybe it's just that the 
developers share some sort of Orwellian social agenda, conditioning 
intelligent young people who might otherwise become dangerously creative to 
toe the line, to always choose the cautious approach.

>One of these days I'm going to write a game where death is simply another
>state of existence and the shocking movements between life and death are
>merely a great circle.  It isn't what I actually believe, but it's a way of
>getting around the authorial crutch of "death is punishment."

In live games we usually tried to avoid death, because it would have been 
folly to tell people who set aside three days and traveled hundreds of miles 
to participate, "you're dead, go home" two hours into the game. A few deaths 
per game we could handle in other ways, with miraculous recoveries, letting 
the players return in new roles and so forth.

But for one event we had a setting in mind where lots of deaths were likely 
to occur. So we wrote half the game to take place in the Astral Plane. The 
player whose character _started_ the game dead was a bit surprised at first, 
but he soon had plenty of company. Living and dead had separate concerns but 
were not totally isolated from one another. Since by the rules of the game 
the ghosts could not be prevented from listening in on any conversation among 
the living, the balance was self-regulating. Factions who killed too many 
enemies found themselves literally haunted, and riddled with unstoppable 
security leaks.

We didn't close the circle and allow Astral players to reincarnate (except in 
a few special cases), but in this case it didn't matter because the game was 
a finite duration anyway.

>> Yeah, that much can be done now (or ten years ago). No story
>> engine, fractal
>> narrative, or artificial intelligence required. BUT... it's not
>> good enough.
>
>The framework is plenty good enough.  What isn't good enough, is the talent,
>budget, and production time allocated to such projects.  We don't have
>sufficient amplifying tools to pull these things off.

Nope. Even with the talent and budget, they're not good enough. Well, OK, 
they sound cool, they are cool, and, if built, they could revolutionize 
digital entertainment. 

But nonetheless, as far as I'm concerned they've all already been built and 
tested. Arabian Nights first ran as a live game in 1987, and Star Saga One, 
using many of the same techniques, came out as a commercial computer game in 
1990. (The latter remains, to my knowledge, the most sophisticated text IF 
ever.) To me this is all old news. I wrote thousands of pages of interactive 
world scripts in the 80's, and spent every spare dime staging them. Turning 
them into digital virtual worlds would be creatively challenging in the 
manner of adapting a book to film, but nothing structurally new, and still 
subject to the same audience-limiting shortcomings that motivated us, in 
1992, to move on.

Not that I wouldn't happily build new media-rich 3-D online versions of 
Arabian Nights or The Vaeringlied or Spaceport Adeline or even Star Saga, 
warts and all, using the craft I've already developed, if someone gave me the 
bucks to do so. Meanwhile, I prefer to work on smaller projects that might 
point the way to solving the problems I ran into ten years ago when trying to 
advance that work farther. I have little motivation to work on low-budget 
home-brewed versions of things I did low-budget home-brewed versions of in a 
different interactive medium fifteen years ago.

>> So, hello AI, hello story engines, hello fractal narrative, hello
>> interactive storytelling problem...

>It's very much an engineer's way of facing a fundamentally human problem.

That's OK with me. Engineers face fundamentally human problems all the time.

>Oh, I think the consensus *does* say that [Myst] appeals to people.  But
>the appeal is expressed in the pejorative.  "Those stupid people who like
>coffee table books!"  "What a bunch of mindless assholes clicking on pretty
>pictures!"  The error in such pejoratives is they don't answer why other
>Myst-like games with pretty pictures *didn't* sell.

Good points. An an even worse error in such perjoratives: they preclude any 
learning of how to build on the successful elements, to create something that 
might sell even better. In a way they represent wilful refusal to learn.

>I agree that puzzle bottlenecks are stupid.  I used to be so radical about
>this that I used to say that all puzzles have to go.  My views are more
>moderate now, but they're still "puzzles are not allowed to be impediments."

Works for me. Many people do find puzzles compelling, but even for those 
people, they're more compelling when they're not impediments.

- Walt