I have been wondering about more creative "uses" for an NPC than the
standard ones ...
* Talk about various topics to get information
* Give them something in exchange for something else
* Order them to do something for you
It would be great to interact in more interesting ways than these. Some
Develop a friendship with someone you see regularly during the game. There
might be a worker you see a few times; if you talk to him each time you see
him, he becomes more friendly. When you are cheated by a salesman later on,
he tells off the salesman and gets you a fair price.
Spread a false rumour about something to lure a thief to a certain location
so you can catch him. Not sure how you would express this, though ...
Blame someone else for something you did, to start a fight between two
people (tricky to express as well).
Ask an NPC to teach / train you how to do something they are obviously good
at (swordsmanship, art forgery, tie dying ...)
* Show them their hideous future.
* Enter their head.
* Sneak up behind and PLANT ONE ON THEM.
> It would be great to interact in more interesting ways than these.
> Some thoughts:
> Develop a friendship with someone you see regularly during the
> Spread a false rumour about something to lure a thief to a certain
> Blame someone else for something you did, to start a fight.
> Ask an NPC to teach / train you how to do something they are
> obviously good at (swordsmanship, art forgery, tie dying ...)
> Thoughts ?
This is one of those situations where I don't want to sit around
and discuss various random example ideas, I want to finish
writing the games I'm working on where I've tried to implement
Let me classify behaviour into two sorts: "commentary" and "utility".
Commentary is possibly the easiest. If your NPC follows the PC around
and makes little comments on his activities (nothing game critical
here) then it's nice if you can vary the comments according to the
NPC's knowledge of the PC (including what it has seen the PC do). For
example, when the PC gets stuck with a locked door, it could say
something like "Why don't you try looking under the doormat again."
(whether that worked before or not) Or if the PC tries to break open
doors it can comment "Violence isn't always the answer, you know." In
fact, as well as knowledge the NPC could start building up an opinion
about the PC, a bit like your idea with friendship but more sublty. The
NPC's comments could then reflect how the NPC feels about the PC based
on what the NPC has seen the PC do.
"Utility" behaviour then takes this further and makes it game-critical.
This means, however, that in order to not allow your game to get into
an unwinnable state you have to either provide multiple outcomes
according to NPC state or allow the PC to get the NPC into the right
state for the game to progress. The possibilities here I think are
quite exciting, especially with multiple outcomes. This'll probably
sound a bit contrived but as an example: you meet agent-x and convince
him that agent-y has double crossed him (a) or persuade him that
agent-y saved his life (b). Later on you meet them both. If you've gone
for (a), the two agents shoot each other, and you can now pick up their
kit to take the story further. If you've gone for (b) the two agents
now co-operate with each other, and with you, and you work with them in
order to take the story further.
I'm putting quite a lot of knowledge based "commentary" style behaviour
in my game (since this is easier to manage I think), but limitting
"utility" style to just annoying the NPC (who then goes off in a huff
and wont then do anything for you, however you can placate him by
simply finding him and apologising).
Re. uses of NPCs, I recommend having a look at
> Develop a friendship with someone you see regularly during the game. There
> might be a worker you see a few times; if you talk to him each time you see
> him, he becomes more friendly. When you are cheated by a salesman later on,
> he tells off the salesman and gets you a fair price.
It's possible to establish positive or negative relationships with
other characters in a number of games: see Varicella, City of Secrets,
Fallacy of Dawn (I think), and also any of the games listed as
"conversational" on the IF Ratings site:
Some of the romance games would fall under this category as well. (How
flexible the relationships are, and what effects they have later on,
obviously vary from game to game.)
> Spread a false rumour about something to lure a thief to a certain location
> so you can catch him. Not sure how you would express this, though ...
Yes, that's a bit vague.
> Blame someone else for something you did, to start a fight between two
> people (tricky to express as well).
I can think of at least a couple of games that have puzzles about
setting two NPCs against one another through some kind of trick.
> Ask an NPC to teach / train you how to do something they are obviously good
> at (swordsmanship, art forgery, tie dying ...)
I can think of a couple of games in which an NPC gives you some powers
or something along those lines; a scenario in which one trains you
interactively seems cooler but also trickier to carry off.
> "I am not a puzzle! I am a human being!"
> - Roger Giner-Sorolla, Crimes Against Mimesis
> I have been wondering about more creative "uses" for an NPC than the
> standard ones ...
> Ask an NPC to teach / train you how to do something they are
> obviously good at (swordsmanship, art forgery, tie dying ...)
I had that in my game, but it really was minimal NPC interaction. You
pay the swordmaster, then get an info screen of you training. When
you're all trained up you get the magic sword (the actual thing you're
> Thoughts ?
There's a game I've had in head for years. I enjoyed the Craig Shaw
Gardner fantasy series about the Cineverse, and universe incorporating
the themes of Golden Age American movies, that is until they started to
be be able to use bad words on screen. One of the main characters is a
Sidekick, and he explains the Sidekick's duties, like moving the plot
along and helping the hero, etc.
This lead me to think about a game with the PC as a Sidekick, but the
Sidekick to a particularly inept Hero. The Hero is going to do stuff,
often really dumb stuff (but heroic, of course) that will get the Hero
or you, or other people killed. You have to set things up and trick the
Hero into doing things the right way. You're not able to order the Hero
around, of course, or ignore orders too blatantly. It all has to be
done in a Sidekicky manner. If you haven't guided the Hero down the
correct path, then the Hero will do things the Hero way when events
The Hero NPC for this would be, to say the least, complex. So would the
interaction with the Hero.
If televison's a babysitter, the Internet is a drunk librarian who
won't shut up.
-- Dorothy Gambrell (http://catandgirl.com)
Let's say we have two city-states -- Bellona and Shangri La. They don't like
The player starts out in Bellona and is of relative anonymity. At the start,
a few default variables are attributed to him -- a 11-place political
scaled, leaning towards either Bellona or Shangri La; a list of political
crimes; the number of people who personally know the main respective to both
city-states. All of these start realtively neutral.
Let's say the main goes to Shangri La. People will generally treat him as
neutral, unless he let's on that he's originally from Bellona. At this
point, some of Shangri La's variables come into play -- how prone people are
to gossip; their general other flows of communication rates (WTF butchered
that sentence ecksdie); the city's current dislike towards the other. These
primarily effect how quickly the main's political affiliation spreads, and
how severely the affect of different feality effects conversation results.
Also, the number of people who know the main reflects his ability to
positively sway other's opinions about him.
Ultimately, NPCs approach you differently depending on these variables. I
think it should be clear to see how it could easily be broadened to take in
several other factors and whatnot. If you've ever played Daggerfall, it sort
of works like this, conversation-wise.
> There's a game I've had in head for years. I enjoyed the Craig Shaw
> Gardner fantasy series about the Cineverse, and universe incorporating
> the themes of Golden Age American movies, that is until they started to
> be be able to use bad words on screen.
*GREAT* books. Another idea would be to have the PC play the hero
Captain Crusader, and do an Edifice-style language puzzle in the Art
The subtitle reads: "Your sister has dog-breath. Am I Captain Crusader?"
> Default User wrote:
> > There's a game I've had in head for years. I enjoyed the Craig Shaw
> > Gardner fantasy series about the Cineverse, and universe
> > incorporating the themes of Golden Age American movies, that is
> > until they started to be be able to use bad words on screen.
> GREAT books.
I enjoyed them, although I'm usually more of a regular science fiction
guy (no need to attempt a definition of what THAT means).
> Another idea would be to have the PC play the hero
> Captain Crusader, and do an Edifice-style language puzzle in the Art
> Film world.
I think that's a great idea ...
It reminds me of a (true) story about a man in the Russian army at a time
when positions of high rank were assigned according to political
connections, rather than actual ability. He was aware of an impending
attack, but he couldn't just go up to the generals and tell them what to
do - so during his normal job of delivering supplies, he would mention
things to the commanders like, "Oh, you're troops must be about here by now"
(pointing to a position on the map) - "I guess you'll be sending them some
reinforcements from town X, then ?" The commander would reply, "Um, yes,
that's right. That's exactly right."
(Details of the story may be incorrect, but that was the general idea).
Give it your best shot: How would you implement this in IF ?
Myself, I'd implement it by writing an actual game
and not dreaming about one on the newsgroups.
It's one thing to talk about some implementing some
game idea, and quite a different thing to actually
implement it. Without some real example to back it
up, all such talk is just wish fulfillment. (And there's
also the chance that it could spoil some games in
A better way to discuss "uses for NPCs" -- by which
I mean a more interesting, useful and rigorous way --
is to talk about NPC uses in the games you've played.
This might require more discipline and effort than
hypothetical flights of fancy, but it's generally
worth it. The most influential theory articles --
such as Craft of Adventure, Crimes Against Mimesis,
Emily Short's IF essays -- are full of examples from
real games. You can always complement your real
examples with hypothetical ones, of course, but it's
best to choose them carefully and keep them
I think that you'd have to use a menu conversation system, otherwise the
possibilities are too great; a command line would have too many
pre-requisites for the author. Still, I'll give it a shot.
First, you (the author) would have to somehow instruct the player that
it's possible to spread a rumor: TELL BOB ABOUT JIM'S DRINKING PROBLEM.
Getting to this point requires teaching the player both that Jim has a
problem and that telling Bob is useful within the game. Then, you need
to teach the player that false rumors are possible: TELL JIM ABOUT
BOB'S GAMBLING PROBLEM. This requires that the player know that
GAMBLING is a valid problem within the context of the game (where DRUG,
TOBACCO or OVEREATING may not be). It also requires knowlege that
setting up this expectation in Jim will lead to a useful result. This
last part may be as hard as the other three put together, since
something like "Pat tells you that spreading a rumor about Bob will help
you win" is unlikely to be satisfactory.
A simpler solution might be to allow the player to give mis-information.
The tatooed man looks around, then whispers, "When is the gold
TELL HIM 5:00 might lead to the thieves being arrested, TELL HIM 7:00
might lead to the player ending the game on a remote Bahama beach.
Heh. My best shot would be in the form of a game, wouldn't it?
But given that I don't have time to do that: there are a lot of
possible approaches to this problem, but the first task would be to
make your idea a good deal more precise.
To come at it from the perspective of puzzle design: is this the main
puzzle in your game, or one puzzle of many? How much time do you want
the player to spend on it, and how complex do you want it to be?
There are lots of possible lightweight implementations, but they're not
very interesting. So let's say it's a fairly significant part of the
game. In that case, we don't want it to be too trivial (e.g., picking
an obvious selection off a conversation menu); which introduces some
Are there going to be other similar puzzles in the game (involving
rumors and communication)? If so, is it worth working out a whole
system for modeling rumors in your hypothetical group of NPCs? How does
this system work, and how do we make it so that the puzzle is more than
just picking the right items from a list? An interesting approach might
be to allow the player to say only things from an inventory of facts,
and center the puzzle on, say, choosing to share the combination of
facts that will mislead the ultimate recipient of the rumors. But I can
think of a bunch of other possible ways to go at this. We'd want to
design our system so that it also accommodated the other puzzles we had
in mind (or design the other puzzles to fit the system).
If on the other hand this is a puzzle of an isolated type and there are
no other rumor-spreadings in the game, the challenge is different: how
do we let the player know that this is an option, given that it's a
pretty non-standard thing to try? And how do we let him express it
without committing ourselves to adding an entire complicated model to
the game that won't otherwise be useful? If most of the other puzzles
involved (say) rigging security systems and manipulating computer
files, I might skip the conversation system entirely and try for a way
of framing the puzzle within the same model: make it be about hacking
into and modifying information in a database, sending the thief an
elaborate code, or something along those lines.
Once you're done thinking through all that, you may have an idea
specific enough to be implemented.
Finding a note in one place and leaving it another. Cutting notes in
half and taping the parts together. Recording over part of an
answering-machine message. Finding a note in an envelope with a
photograph, and swapping it for a different photograph. Writing on the
back of a photograph.
DON'T BELIEVE HIS LIES.
"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."
If the Bush administration hasn't thrown you in military prison without trial,
it's for one reason: they don't feel like it. Not because you're patriotic.
>Myself, I'd implement it by writing an actual game
>and not dreaming about one on the newsgroups.
>It's one thing to talk about some implementing some
>game idea, and quite a different thing to actually
>implement it. Without some real example to back it
>up, all such talk is just wish fulfillment.
Frankly David, I think you should ignore this advice entirely. I find
the topics that you raise and the discussions that follow to be both
entertaining and enlightening. Just recently, you introduced a thread
concerning save points, and the ensuing discussion gave me some radical
new insights about my own project. I was so struck by them that I had
to get up from my chair and wander around the room for a good twenty
minutes to take it in.
You do the community a great service by starting interesting
discussions and challenging widely held assumptions, and I hope you
continue to do so unselfconsciously. Critical thinking is quite
abdundant here, but enthusiasm and creativity are sometimes in short
Though I dread and fear having to parse things like
>CUT THE YELLOW NOTE IN HALF WITH THE CUT RUNNING BETWEEN THE WORDS FAKE AND DIAMOND.
Also, trying to provide a sensible response if the player cuts the
notes into ransom-letter confetti and reassembles them to say FEED ME,
But I find this an appealing idea. I'm just not quite sure what game
system (short of a human GM) could deal with it sensibly.
> DON'T BELIEVE HIS LIES.
You'll note that I'm not criticising David's posts in general,
which are indeed useful and interesting, but the thrust of this
> Just recently, you introduced a thread
> concerning save points, and the ensuing discussion gave me some radical
> new insights about my own project. I was so struck by them that I had
> to get up from my chair and wander around the room for a good twenty
> minutes to take it in.
Good for you. I look forward to playing your game. (Though you should
probably check the names of the contributors to the "Savepoints in
IF" thread. Credit where it's due.)
> You do the community a great service by starting interesting
> discussions and challenging widely held assumptions, and I hope you
> continue to do so unselfconsciously.
As do I. But if I happen to dislike one of those discussions or hold
one of those widely held assumptions, I'll respond appropriately.
Critical thinking is quite
> abdundant here, but enthusiasm and creativity are sometimes in short
I have plenty of enthusiasm for IF. This is where I get my enthusiasm
to post about my lack of enthusiasm for idle "what about a game
Have you any evidence that creativity is in short supply?
To be honest, I'm not at all comfortable with your conversational
style, as it resembles an agressive trial lawyer during
cross-examination, but that does not give me the right to insult you.
My remark about creativity and enthusiasm being a lacking commodity was
not directed at you specifically, but rather echoes the opinion that
I-F has entered a stagnant period. In retrospect, I can see that it was
a poor choice of words, and I apologize for not being clearer.
>David Fisher wrote:
>> "I am not a puzzle! I am a human being!"
>> - Roger Giner-Sorolla, Crimes Against Mimesis
>> I have been wondering about more creative "uses" for an NPC than the
>> standard ones ...
>> * Talk about various topics to get information
>> * Give them something in exchange for something else
>> * Order them to do something for you
>Re. uses of NPCs, I recommend having a look at
>> Develop a friendship with someone you see regularly during the game. There
>> might be a worker you see a few times; if you talk to him each time you see
>> him, he becomes more friendly. When you are cheated by a salesman later on,
>> he tells off the salesman and gets you a fair price.
A rather successful implementation of this was in Frederick Pohl's
Gateway. There was a hermit/castaway you encountered in a remote
planet, whose family offered a prize for his safe return. His decision
about whether to accompany you back to earth depended on the actual
solutions of many problems you encountered in the planet. Each problem
had two solutions, an easy one which angered the hermit, and a
difficult one which pleased him. It was a very well thought
meta-puzzle, I think.
> I have plenty of enthusiasm for IF. This is where I get my enthusiasm
> to post about my lack of enthusiasm for idle "what about a game
> where...?" threads.
Personally, I think this thread was more useful than the savepoint one.
Most people think that better NPC interaction would improve games. But
what does "better" mean? What are the likely avenues of exploration?
Implementing new ideas is one way to explore, but there's nothing at
all wrong with brainstorming some concepts too. If that gives someone
working a new game a bit of inspiration, it's beneficial.
In short, I think you were way off base.
Thanks for that, Emily. I find the questions you bring up and the logical
approach you have to these kinds of issues to be very helpful -
That's a very clever idea - I'll check out Gateway.
Did you find that it was clear enough what the consequences would be to
solving puzzles the easy way, or did you have to restart the game once you
discovered what the "catch" was ?
I seem to remember it was pretty clear from the beginning (from the
conversations with the hermit) that he was ambivalent about returning
to civilization, and I think most of the "easy" solutions were
followed by him expressing his displeasure. Not too sure about the
details tho, been too long.
I've no doubt that lots of people enjoy "what-if" discussions,
which is why they continue to appear in r*if despite the whinges
of various people since long before I showed up here. But I
reserve the right to carp about them every now and then.
In general, I'm interested in descriptive IF theory ("how IF is
and has been written") and not at all interested in prescriptive
IF theory ("how to write IF") but the latter is on-topic here
and arguably useful. Also, there's a fine line and frequent
overlap between the two approaches: the recent "newbie
transcripts" thread saw a bit of both. There's also a fine line
between "how to write IF" and "how my dream piece of IF would
be written". When prescriptive theory discussions stay on one
side of the line, I'm merely uninterested and don't contribute;
when they cross over into idle speculation, I'm actively
irritated by it and occasionally say so.
Amusing coincidence that S&W and Following are both from 1998.