[MUD-Dev] believable NPCs (was Natural Language Generation)

[Michael Sellers]

JC Lawrence wrote:
> Michael Sellers <mike at onlinealchemy.com> wrote:

>> Believable characters OTOH -- with or without NLG -- will make
>> gameplay much more immersive and consequential.  Imagine asking a
>> tavern keeper NPC for info on where to buy good weapons.
>> Consider how his responses might differ based on whether he's
>> seen you before, whether you tipped well last time, whether and
>> how he's heard other NPCs (and PCs) talk about you, and, say,
>> whether he heard you were flirting heavily with the blacksmith's
>> daughter (whom he happens to be in love with), or were the one to
>> save her from ruffians.  This goes way beyond script- or
>> knowledge-based responses, but provides much more interesting
>> player-related context for even the simplest menu-driven
>> interactions.

> I'm not convinced.

> Many text MUDs had two modes: long descriptions and short
> descriptions.  The standard practice was for players to turn on
> long descriptions only when exploring a new area, but to otherwise
> play with the faster and more efficient short descriptions (even
> if they were on fast connections).  This was true even on the
> games that dynamically generated or customised the long
> descriptions.  As such they sacrificed emotive richness and
> presentation quality for goal efficiency.  There are/were
> (partial) exceptions like Raph's Legend where long descriptions
> were often structurally necessary for effective ongoing play, but
> I'd argue that they're contrived examples where necessary
> datacomms were inlined with the long descriptions to enforce long
> descriptions rather than being a quality explicitly selected by
> players.

Part of the challenge then is to make sure any "flavor" provided by
believable NPCs isn't obstructive, but that it is integral to the
gameplay.  Yes, some goal-oriented players will complain because now
they have to actually interact meaningfully with people, but you
have to decide if the Diablo market is your main market or not.

I'd argue that this is a different experience entirely from the old
"long descriptions" in text muds, as those added nothing qualitative
to the gameplay; after the first time you read some lush description
it's just more yada yada yada that doesn't change anything.

> Coming from the other side the ideal form would be to have capable
> actresses and actors taking the roles of NPCs with solid character
> definitions etc.  The (only) reason we don't due this is that the
> economics are prohibitive.  Ergo, we try and simulate humanity
> through computation in the form of NPCs.  I don't see that the
> simulation has intrinsic value.

For some games it won't.  But if a MUD or MMP game could get the
effect of actors on staff 24/7 who never break character and who
don't offer the same stilted dialogue over and over again, and where
such NPCs were integrated with the gameplay itself, I suspect that
would add significantly to the players' experience.

> At root the problem would seem to be a dichotomy in player goals;
> parallel to the description length field choices: the tourists who
> are there for the experience/ride, and the achievers (only partly
> Bartle-sense) who are there to accomplish named and known goals.
> Tourists seem to correlate highly with high socialiser quotients,
> and to a large extent play for the turgid feedback loops of human
> emotion and reaction among the players as set against the game
> backdrop/scenery.  (cf roleplaying) Conversely, in the most
> extreme form the GoP players like the thematic flavour (cf spice
> in food), but the shortest route to the cheese comes first and
> drives visible motivation and activity.  Between such audiences
> the rich NPC characters are scenery for the social tourists,
> acting more as catalysts and stage props for the human
> interactions they seek, and for the GoP players the rich NPCs are
> annoyingly detailed and complex barriers in forming their
> optimised route to the cheese.  This latter is especially true in
> the case of treadmills where repetition plays such a large role.

You seem to be arguing that repetitious treadmill play is a good
thing.  I disagree.  IMO we have it only because the alternatives
are prohibitively expensive or have not been successfully
implemented.  Clearly there's an ultra-hard-core audience for whom
anything getting in the way of their cheese is an annoyance; my bet
is that they are a smaller segment than any other except perhaps
dedicated PvPers.  From this POV, people who don't want to interact
with anything more than a cardboard cutout of an NPC may find
themselves in a subculture analogous to text MUDders -- defiant in
their niche, but largely in a backwater as games and play styles
change.

> Its very easy to get hooked up in the concept of a virtual world.
> The intellectual delight and attraction of such concepts are huge.
> Engineers especially take an almost masturbatory intellectual
> delight in trying to assemble a virtual world which fully
> simulates everything from the physics models to the social
> structures, societies and individuals in them.  "But of course
> that's also what the whole world wants!  Isn't it just obvious?
> Its just so cool!  Look, there are real NPC people with real
> believable NPC relationships and prejudices, and real clouds in
> the sky and real deer in the fields that breed and have real baby
> deer!"  The very definition of engineer-toy cool!

That's one way to look at it.  Another way to look at is that from
Barbie Fashion Designer to Tamagotchi to The Sims, the more
believable the characters, the greater number of people become
deeply hooked (though Masahiro Mori's "uncanny valley" looms too).
It's not about uber-simulation (for me anyway); it's about human
emotive connection.  Those who want to stay on the "kill monster get
gold" treadmill may not see the value in this immediately; they are
quite literally like the lab rat pressing the bar repeatedly to get
the random-scheduled food pellet, so their needs are essentially
satisfied.

>   From the player perspective in terms of the actual goals that
>   player can be _seen_ to pursue (social, cultural, in-game, etc),
>   what is the actual function of an NPC?

The answer to that will be based on how much we as developers can
break out of our own twenty-year-old molds of gameplay.  Can
believable NPCs add to the gameplay experience in terms of the
players' goals?  I think so, immensely so.  But we'll see.

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