mimesis vs. courtesy to the player

17 views
Skip to first unread message

Jim Aikin

unread,
Oct 24, 2006, 8:58:13 PM10/24/06
to
Sometimes NPCs need to make longish statements, for expository reasons. 'ask
butler about murder' might give you a rather long text, for instance.

One of my testers was bothered by the fact that if he asked an NPC about a
topic, he'd get the same block of text each time. This, of course, is a
crime against mimesis. In real life, the NPC would say something like, "I've
already told you all I know about that."

But what if the player is returning to a saved game after days or weeks and
wants to refresh her memory? In that case it would be rude in the extreme
for the game to refuse to print out the text block.

One could, of course, implement a 'repetitions on' verb, but that would be a
ton of extra work, given the number of NPC responses in my WIP.

How do others feel about this issue? I'm inclined to leave the repetitions
in the game, in spite of the breaking of mimesis. After all, how many
players (other than beta-testers) will type 'ask butler about murder' three
times in a row just to see what happens? But I'd appreciate some
feedback....

--Jim Aikin


David Fisher

unread,
Oct 24, 2006, 9:17:19 PM10/24/06
to
"Jim Aikin" <rai...@musicwords.net> wrote in message
news:FEy%g.17352$vJ2....@newssvr12.news.prodigy.com...

> Sometimes NPCs need to make longish statements, for expository reasons.
> 'ask butler about murder' might give you a rather long text, for instance.
>
> One of my testers was bothered by the fact that if he asked an NPC about a
> topic, he'd get the same block of text each time. This, of course, is a
> crime against mimesis. In real life, the NPC would say something like,
> "I've already told you all I know about that."
>
> But what if the player is returning to a saved game after days or weeks
> and wants to refresh her memory? In that case it would be rude in the
> extreme for the game to refuse to print out the text block.
...

> How do others feel about this issue?

Here is a helpful thread about this sort of thing ("City of Secrets
conversation interface"):

http://groups.google.com/group/rec.arts.int-fiction/browse_frm/thread/b553cc1abb3b7809/275fc8264cfff407

Not wanting to stifle any further discussion, though ... :-)

David Fisher


Jan Thorsby

unread,
Oct 24, 2006, 10:47:35 PM10/24/06
to
> ask butler about murder

"I didn't do it."

> ask butler about murder

He already told you he didn't do it.


Neil Cerutti

unread,
Oct 24, 2006, 11:14:57 PM10/24/06
to
On 2006-10-25, Jim Aikin <rai...@musicwords.net> wrote:
> Sometimes NPCs need to make longish statements, for expository
> reasons. 'ask butler about murder' might give you a rather long
> text, for instance.
>
> One of my testers was bothered by the fact that if he asked an
> NPC about a topic, he'd get the same block of text each time.
> This, of course, is a crime against mimesis. In real life, the
> NPC would say something like, "I've already told you all I know
> about that."
>
> But what if the player is returning to a saved game after days
> or weeks and wants to refresh her memory? In that case it would
> be rude in the extreme for the game to refuse to print out the
> text block.

Yup: "usesis" conflicts with "mimesis".

The compromise is as follows:

>ASK BUTLER ABOUT MURDER
He says: "A lot of text."

>ASK BUTLER ABOUT MURDER
"Again?" he sighs rhetorically. "Oh, very well."

"As I said before, a lot of text."

> How do others feel about this issue? I'm inclined to leave the
> repetitions in the game, in spite of the breaking of mimesis.
> After all, how many players (other than beta-testers) will type
> 'ask butler about murder' three times in a row just to see what
> happens? But I'd appreciate some feedback....

I think usesis has to trump mimesis here. Some acknowledgement
that you thought about the issue, as in my example, is good, but
that's as far as it ought to go.

A worth-the-effort solution might be a journal that keeps details
of all important conversations in a notebook the player could
consult whenever its available and visible.

And even if you don't do anything, that's fine, too.

But I think to refusing to reprint important conversation text is
right out.

--
Neil Cerutti

Andrew Plotkin

unread,
Oct 24, 2006, 11:34:57 PM10/24/06
to
Here, Jim Aikin <rai...@musicwords.net> wrote:
>
> How do others feel about this issue? I'm inclined to leave the repetitions
> in the game, in spite of the breaking of mimesis. After all, how many
> players (other than beta-testers) will type 'ask butler about murder' three
> times in a row just to see what happens?

Since "murder" probably has several synonyms in the topic list, most
players will. If you're trying a bunch of phrases, one will inevitably
turn out to be the same as one you've already tried.

The repetitions annoy me, but I agree that it's very bad to give
information once and then make it inaccessible after that. As others
have suggested, you can come up with a shorter paraphrase for the
repetition.

--Z

--
"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."
*
If the Bush administration hasn't subjected you to searches without a warrant,
it's for one reason: they don't feel like it. Not because you're innocent.

Jim Aikin

unread,
Oct 25, 2006, 2:17:36 AM10/25/06
to

> x butler
The butler is holding a smoking revolver and grinning evilly.

> butler, give me the gun
The butler has better things to do.

>oops revolver
(first wiping off the revolver with a handkerchief)
The butler hands you the smoking revolver. "I really didn't do it," he says.
"I'm sure you understand." The butler's left eye is twitching spasmodically.

Celestianpower

unread,
Oct 25, 2006, 4:33:44 AM10/25/06
to

Yes, it is definitely a problem.

What I did with Bible Retold is printed the information 4 times (on
four subsequent TALK TOs), getting more and more irate each time ("How
many times do I have to tell you!"), before the NPC refused to repeat
himself again. It's not the best solution probably, but it seems to be
somewhat of a compromise.

CP

Jacek Pudlo

unread,
Oct 25, 2006, 6:05:04 AM10/25/06
to
"Jim Aikin" <rai...@musicwords.net> skrev i meddelandet
news:FEy%g.17352$vJ2....@newssvr12.news.prodigy.com...

> Sometimes NPCs need to make longish statements, for expository reasons.
> 'ask butler about murder' might give you a rather long text, for instance.
>
> One of my testers was bothered by the fact that if he asked an NPC about a
> topic, he'd get the same block of text each time. This, of course, is a
> crime against mimesis. In real life, the NPC would say something like,
> "I've already told you all I know about that."
>
> But what if the player is returning to a saved game after days or weeks
> and wants to refresh her memory?

Type RESTART followed by YES. That's what I'd do with a novel I return to
after having dropped it in mid-plot for weeks.

> In that case it would be rude in the extreme for the game to refuse to
> print out the text block.

It would be silly in the extreme for the player to assume that the game
somehow "knows" he hasn't played it for weeks.

> One could, of course, implement a 'repetitions on' verb, but that would be
> a ton of extra work, given the number of NPC responses in my WIP.

Flag each and every NPC response so that you can do this:

ASK HECTOR ABOUT HELMET

blah blah blah

ASK HECTOR ABOUT HELMET

You recollect verbatim from your god-like memory:

"blah blah blah"


> How do others feel about this issue? I'm inclined to leave the repetitions
> in the game, in spite of the breaking of mimesis.

How pedantically should an IF game mime reality? How do you feel about a
game that lets you open and close a door 50000 times without advancing
in-game time by so much as a minute? An atrocity against mimesis or a
necessary evil?


Conrad Kayne

unread,
Oct 25, 2006, 6:18:24 AM10/25/06
to
IMO, it doesn't irreparably break mimesis to repeat blocks of PC/NPC
conversation. After all, when reading a novel, one often skims back
over sections already read in order to get some earlier plot-point
clear or simply to savour the writing over again. (at least I do!)

But if you do consider it a dilemma, another way round it is to have
the ASK only print the exchange once, and on subsequent ASKs, inform
the player of alternative ways to access either the same text or the
gist of it at his leisure. Or implement a notebook for the player to
consult. Or just print a short precis if you don't want to repeat the
whole section of dialogue.

If it's one of those "I have no more to say"-type responses, the only
sensible way around it is to vary the responses randomly. The more
alternatives you can invent, the merrier. Nothing breaks mimesis like a
short stupid phrase repeated ad nauseam, or to have the parser shout at
the player "How many times do I have to tell you!" (because that's the
author qua game-author addressing the player qua player, a definite
no-no.)

Many's the game I've deleted in disgust just because of this sort of
behaviour.

Regards,

ck

Robin Johnson

unread,
Oct 25, 2006, 6:27:09 AM10/25/06
to
Jim Aikin wrote:

> Sometimes NPCs need to make longish statements, for expository reasons. 'ask
> butler about murder' might give you a rather long text, for instance.
>
> One of my testers was bothered by the fact that if he asked an NPC about a
> topic, he'd get the same block of text each time. This, of course, is a
> crime against mimesis. In real life, the NPC would say something like, "I've
> already told you all I know about that."

In my opinion, courtesy to the player wins outright.

But in a situation like this, I'd probably make the NPC say only a
summary (perhaps with some randomised elements, for flavour) on
subsequent ASKs. An NPC saying a long monologue is basically a
cutscene, and I wouldn't repeat a cutscene in-game.

Robin Johnson

A.P. Hill

unread,
Oct 25, 2006, 9:24:56 AM10/25/06
to
First of all, you break memesis when you say, "ask butler about murder"
Who the fuck talks like that? IF is a memesis breaker right off the
fucking bat. So don't worry about it. Until you fix..."ask butler
about murder" just worry more about laying down a good story within
the current medium.
Your thread should be, how do you make a better language than whats
available. Everything else is just circle talk. I tend not to talk
during groupsex.

A.P. Hill

Rockersuke Moroboshi

unread,
Oct 25, 2006, 10:32:47 AM10/25/06
to

"Jim Aikin" <rai...@musicwords.net> escribió en el mensaje
news:FEy%g.17352$vJ2....@newssvr12.news.prodigy.com...

> How do others feel about this issue? I'm inclined to leave the repetitions
> in the game, in spite of the breaking of mimesis. After all, how many
> players (other than beta-testers) will type 'ask butler about murder'
> three times in a row just to see what happens? But I'd appreciate some
> feedback....

My personnal take on this:

>NPC, tell me it all

>Oh, well, first there was God, and He said "Let there be light"... later
>came the dinosaurs... (several screens of text with the universal History
>triumphally ending with the release of Inform 7 ;-p ) ...and that's it.

3 weeks later player returns to the game

>NPC, tell me it all

> Hey man! You really DO want me to tell it ALL again? (yes/no)

>yes

>Ouch... well, first there was God, and...

The courtesy bit is alerting somehow the player that an already known long
piece of text is going to be displayed and letting him decide wether he
feels like going through it again or not.

Jim Aikin

unread,
Oct 25, 2006, 11:45:37 AM10/25/06
to
"A.P. Hill" <aph...@altavista.com> wrote in message
news:1161782696.6...@e3g2000cwe.googlegroups.com...

> First of all, you break memesis when you say, "ask butler about murder"
> Who the fuck talks like that? IF is a memesis breaker right off the
> fucking bat. So don't worry about it. Until you fix..."ask butler
> about murder" just worry more about laying down a good story within
> the current medium.

Good point.

> Your thread should be, how do you make a better language than whats
> available. Everything else is just circle talk. I tend not to talk
> during groupsex.

How do I make a better language than modern English? People have tried. I
prefer to tilt at smaller windmills.

--JA


Jim Aikin

unread,
Oct 25, 2006, 11:54:59 AM10/25/06
to

"Andrew Plotkin" <erky...@eblong.com> wrote in message
news:ehmm11$cjb$3...@reader2.panix.com...

> The repetitions annoy me, but I agree that it's very bad to give
> information once and then make it inaccessible after that. As others
> have suggested, you can come up with a shorter paraphrase for the
> repetition.

Coming up with shorter paraphrases is not hard to do, it's just a lot of
extra work. I have _ten_ NPCs in this story, one of whom can transition from
one state to another so radically that she's coded as two separate objects.
My original list of topics that some or all of them would need to be able to
respond to had about 25 items. My testers have added 25 more, all of them
quite reasonable. So that's upwards of 400 separate responses, even if we
grant that the maid is not a very interesting or articulate character and
can reasonably ignore a lot of topics.

I think what I need to do first is finish and release the project. All of
these refinements, along with some others that have been suggested, can be
added to version 2.

--JA


Jim Aikin

unread,
Oct 25, 2006, 11:56:37 AM10/25/06
to

"Celestianpower" <celesti...@gmail.com> wrote in message
news:1161765224....@m73g2000cwd.googlegroups.com...

>
> What I did with Bible Retold is printed the information 4 times (on
> four subsequent TALK TOs), getting more and more irate each time ("How
> many times do I have to tell you!"), before the NPC refused to repeat
> himself again. It's not the best solution probably, but it seems to be
> somewhat of a compromise.

One of the comp games has a hilarious, over-the-top version of this
technique. If you try opening a certain door or performing a certain action,
the game will give you ten or twelve increasingly irate responses. (And
brilliantly, it turns out that there's a sort of in-game justification for
its doing so.)

--JA


Jim Aikin

unread,
Oct 25, 2006, 12:00:37 PM10/25/06
to

"Jacek Pudlo" <ja...@jacek.jacek> wrote in message
news:kFG%g.20828$E02....@newsb.telia.net...

> "Jim Aikin" <rai...@musicwords.net> skrev i meddelandet
> news:FEy%g.17352$vJ2....@newssvr12.news.prodigy.com...
>
> Type RESTART followed by YES. That's what I'd do with a novel I return to
> after having dropped it in mid-plot for weeks.

But should the novelist FORCE the reader to start over at the beginning?
Speaking as a novelist, I'd say no. The reader should be entitled to pick up
where she left off, and to flip back to an earlier page and refresh her
memory along the way if she needs to.

Also, more than a few IF puzzles are solved in ways that people couldn't
repeat, because they don't know what they did! Forcing them to start over in
that situation would be extremely cruel.

> Flag each and every NPC response so that you can do this:
>
> ASK HECTOR ABOUT HELMET
>
> blah blah blah
>
> ASK HECTOR ABOUT HELMET
>
> You recollect verbatim from your god-like memory:
>
> "blah blah blah"

Amusing, but real characters don't have godlike memories. If the player
character is supposed to be a real person, giving her total recall of
earlier conversations (even if they CAN be reconstructed verbatim, which is
not guaranteed, because the conversation block may have been assembled on
the fly while testing several variables) would in itself break mimesis.

--JA


JDC

unread,
Oct 25, 2006, 1:10:18 PM10/25/06
to

Jim Aikin wrote:
>
> Amusing, but real characters don't have godlike memories. If the player
> character is supposed to be a real person, giving her total recall of
> earlier conversations (even if they CAN be reconstructed verbatim, which is
> not guaranteed, because the conversation block may have been assembled on
> the fly while testing several variables) would in itself break mimesis.

This is a really good point about repeated conversation in general,
namely that the state may have changed from the initial conversation.
As you point out, to recall the earlier conversation verbatim (or even
a summary) might require saving the state of previous conversations, a
nasty business. And what about the following:

> ask alice about bob
"Bob is wonderful! I hope he asks me to marry him soon!"

(later in the game, after Alice has caught Bob in bed with her sister)
> ask alice about bob
"I hope he dies in a horrible, horrible way, preferably involving a
giant squid!"

So what do you want the response to be now if you ask again?
> ask alice about bob
"You've asked me that twice now! FIrst, before I realized what utter
slime he was, I told you that he was wonderful and I wanted to marry
him. Then, once I came to my senses, I told you how I wanted him to die
horribly."

This seems rather less than optimal, somehow, and I think here it would
be best to just repeat the last response.

-JDC

n2...@yahoo.com

unread,
Oct 25, 2006, 2:07:55 PM10/25/06
to
Rockersuke Moroboshi wrote:
> "Jim Aikin" <rai...@musicwords.net> escribió en el mensaje
> >NPC, tell me it all
>
> > Hey man! You really DO want me to tell it ALL again? (yes/no)
> >yes
>
> >Ouch... well, first there was God, and...
>
> The courtesy bit is alerting somehow the player that an already known long
> piece of text is going to be displayed and letting him decide wether he
> feels like going through it again or not.

While I feel that this is good general answer, and possible even
the right answer for Text-to-Speech. Do any of the other answers
for different reasons change, when facing the prospect of
attempting to be TTS friendly?

Or, not having tuned my ears to a visually impaired
game session, am I assuming a bandwidth problem?

should VERBOSE and other in game variables be affected by a -TTS start?

José Manuel García-Patos

unread,
Oct 25, 2006, 2:08:46 PM10/25/06
to

> How do others feel about this issue? I'm inclined to leave the repetitions
> in the game, in spite of the breaking of mimesis. After all, how many
> players (other than beta-testers) will type 'ask butler about murder' three
> times in a row just to see what happens? But I'd appreciate some
> feedback...

What about implementing ASK NPC _AGAIN_ ABOUT TOPIC? (Maybe it's already
implemented. I had never thought about it myself before, but it looks like
the obvious solution to me.) It wouldn't break mimesis, and the answer
would go straight to the point the player wants to know about. You could
even start the response with something like "As I already told you..." to
make it more realistic. Or you could complicate it by merging several
responses already given by ASK NPC ABOUT, if there were more than one.

All The Best.
José Manuel García-Patos
Madrid

steve....@gmail.com

unread,
Oct 26, 2006, 3:17:21 PM10/26/06
to
Jim Aikin wrote:
> Sometimes NPCs need to make longish statements, for expository reasons. 'ask
> butler about murder' might give you a rather long text, for instance.

This is no doubt true, but you probably want to maximize for
interactivity, which means breaking up the information such that pieces
come frequently, and through various interactions, rather than more
rarely in large chunks.

In other words, where you have a large chunk of text, you probably want
to break it up into smaller bits of text, and present that information
as a result of varying types of interaction.

So, instead of learning everything from asking the butler about the
murder, you could learn some from that, some from eavesdropping in
another scene, some from the butler's diary, some from his ledger, some
from asking the butler about something seemingly unrelated, some from
the state of the butler's overshoes, and so on.

This by itself goes a long way to solving your problem. Dropping a lot
of text is fine, especially if you want to break it up into several
different replies to the same question. You might have one single
information-point that the butler-interaction needs to communicate to
the player, and you'll want to phrase the same information differently
over the course of multiple replies.

> One of my testers was bothered by the fact that if he asked an NPC about a
> topic, he'd get the same block of text each time. This, of course, is a
> crime against mimesis. In real life, the NPC would say something like, "I've
> already told you all I know about that."

Mimesis isn't really the word for it: we're probably intending
'realism' by the word, but I don't think that 'crime against realism'
is really the problem either (although a tester not-yet-disabused of
these fallacies might indeed put it that way) -- for really we're not
looking for realism in IF. We're looking for variation and interest,
multi-dimensional characters and so on. It's a crime against
interactivity and good IF writing, really.

> But what if the player is returning to a saved game after days or weeks and
> wants to refresh her memory? In that case it would be rude in the extreme
> for the game to refuse to print out the text block.

Well, shuffled event list sounds like the best solution. (The butler
says one of maybe 4 different things, chosen randomly each time through
the order.) Each variation can provide a sufficient version of the the
important bit of information (assuming we're talking about a mystery or
a puzzler), so the player who asks once gets the meat, while a player
who asks multiple times will get the reward for exploring the
interactivity of the piece.

d...@pobox.com

unread,
Oct 26, 2006, 5:09:42 PM10/26/06
to

On Oct 26, 8:17 pm, steve.bres...@gmail.com wrote:
> Well, shuffled event list sounds like the best solution. (The butler
> says one of maybe 4 different things, chosen randomly each time through
> the order.) Each variation can provide a sufficient version of the the
> important bit of information (assuming we're talking about a mystery or
> a puzzler), so the player who asks once gets the meat, while a player
> who asks multiple times will get the reward for exploring the
> interactivity of the piece.

Having to type
ASK BUTLER ABOUT MURDER
4 times in a row sounds about as thrilling as having to type
SEARCH THE CHEST
4 times in a row. Patently.

drj

d...@pobox.com

unread,
Oct 26, 2006, 5:11:58 PM10/26/06
to

On Oct 26, 8:17 pm, steve.bres...@gmail.com wrote:

> Jim Aikin wrote:
> > One of my testers was bothered by the fact that if he asked an NPC about a
> > topic, he'd get the same block of text each time. This, of course, is a
> > crime against mimesis. In real life, the NPC would say something like, "I've
> > already told you all I know about that."Mimesis isn't really the word for it: we're probably intending
> 'realism' by the word, but I don't think that 'crime against realism'
> is really the problem either (although a tester not-yet-disabused of
> these fallacies might indeed put it that way) -- for really we're not
> looking for realism in IF. We're looking for variation and interest,
> multi-dimensional characters and so on. It's a crime against
> interactivity and good IF writing, really.

Indeed. It's more usually a crime against physics (ie simulation),
immersion, or plausibility. But hey, we can call text adventures IF so
mimesis seems fine to me.

drj

Default User

unread,
Oct 26, 2006, 5:33:35 PM10/26/06
to
d...@pobox.com wrote:

To me, that's worse than a "guess the verb" puzzle. It's "guess the
number of times to ask".

One way to solve it is to have the NPC give a relatively short
response, then have the PC ask follow-up questions. That avoids that
large expository dump and rewards persistent and (relatively)
intelligent questioning. Pick out new keywords based on what you have
so far.


Brian

--
If televison's a babysitter, the Internet is a drunk librarian who
won't shut up.
-- Dorothy Gambrell (http://catandgirl.com)

steve....@gmail.com

unread,
Oct 26, 2006, 5:44:31 PM10/26/06
to
d...@pobox.com wrote:

> Having to type
> ASK BUTLER ABOUT MURDER
> 4 times in a row sounds about as thrilling as having to type
> SEARCH THE CHEST
> 4 times in a row. Patently.

You miss the point. We want to satisfy as many players as possible, of
course. Certainly 'Patently.'-esque players like you, but other types
also.

One player (like you I guess) will want to ask the question once and
from the one answer receive enough information to solve the game to his
satisfaction. I don't see why anyone would ever wish to write for a
reader like this, but other writers are apparently more charitable than
I. (On the other hand, if you need my point repeated to you, you
probably won't understand what the butler says about the murder the
first time 'round either, and will want the same main point repeated in
different ways, as I initially suggested.)

Another player (more like what I take to be the intelligent and
interested player) might want to ask several times, or as Andrew says,
will accidentally ask the question several times, in an attempt to
better problem-solve, or better, appreciate the author's writing and
its interactivity.

> [A 'crime against mimesis' is] more usually a crime against


> physics (ie simulation), immersion, or plausibility.

Probably most useful to call it one or the other, as appropriate,
thereby identifying which fallacy we're belaboring straight away, and
with greater precision to boot.

> But hey,
> we can call text adventures IF so mimesis seems fine to me.

You might disagree, but I never thought of 'interactive fiction' as an
unforgivable abuse of either term. You can call it whatever you like,
but be aware that jargon or other shorthand very often occludes and
collapses meaning, whereas taking the time initially to call what one
means by the right name -- this avoids lots of confusion from the
start. It's not always easy to figure out the correct formulation of a
question, but one may remark when certain jargony words are discovered
to be inherently misleading.

Emily Short

unread,
Oct 26, 2006, 5:48:07 PM10/26/06
to

Default User wrote:

> d...@pobox.com wrote:
> > Having to type
> > ASK BUTLER ABOUT MURDER
> > 4 times in a row sounds about as thrilling as having to type
> > SEARCH THE CHEST
> > 4 times in a row. Patently.
>
> To me, that's worse than a "guess the verb" puzzle. It's "guess the
> number of times to ask".

There's a bit like that in Anchorhead, and I got completely stuck at
that point in the game, because I hadn't realized I could keep pursuing
the same line of inquiry. Other players I talked to found it
frustrating, too. So I'd say it's probably a good idea to avoid this,
or -- failing that -- make it very clear on the first question that
there is more information that might be available with further
application.

Default User

unread,
Oct 26, 2006, 6:21:11 PM10/26/06
to
Emily Short wrote:

I have played games where you could change the tone in which you ask
the question, and get differing results with "friendly" versus
"aggressive" or something like that. That's somewhat more legitimate.

Also, if you tell the NPC something or show it something, then you
might expect different results (those pictures with the three midgets
and the yak might get butler's tongue moving). Or bribe the NPC. It
could be legit too if then NPC "gets to know you", but that's kind of
tricky. Maybe when the bartender stops calling you "stranger" or
something.

steve....@gmail.com

unread,
Oct 26, 2006, 6:32:58 PM10/26/06
to
Emily Short wrote:
> Default User wrote:
> > d...@pobox.com wrote:
> > > Having to type
> > > ASK BUTLER ABOUT MURDER
> > > 4 times in a row sounds about as thrilling as having to type
> > > SEARCH THE CHEST
> > > 4 times in a row. Patently.
> >
> > To me, that's worse than a "guess the verb" puzzle. It's "guess the
> > number of times to ask".
>
> There's a bit like that in Anchorhead

Oh my GOD I cannot believe you're missing the point so badly. NOBODY
suggested that you should have to ask the same question several times
before receiving the necessary information. The point was that the
necessary information should be given from the beginning, and though
varied, repeated each time. Obviously the world is populated by readers
who require the same point be repeated over and over.

But while we're on the subject, yes I hated that similar thing, in one
of Baggett's games, where you had to perform multiple 'search'es before
finding the scroll on the wizard's corpse (or something like that).

d...@pobox.com

unread,
Oct 26, 2006, 6:55:44 PM10/26/06
to

On Oct 26, 10:44 pm, steve.bres...@gmail.com wrote:
> d...@pobox.com wrote:
> > Having to type
> > ASK BUTLER ABOUT MURDER
> > 4 times in a row sounds about as thrilling as having to type
> > SEARCH THE CHEST
> > 4 times in a row. Patently.
> You miss the point. We want to satisfy as many players as possible, of
> course. Certainly 'Patently.'-esque players like you, but other types
> also.

I don't think everyone wants their works to be maximally popular. Nor
would I appreciate it particularly. Sure, I enjoy the odd Hollywood
summer blockbuster, but I also have pretentious french stuff on my
shelves of DVDs.

>
> One player (like you I guess) will want to ask the question once and
> from the one answer receive enough information to solve the game to his
> satisfaction. I don't see why anyone would ever wish to write for a
> reader like this, but other writers are apparently more charitable than
> I. (On the other hand, if you need my point repeated to you, you
> probably won't understand what the butler says about the murder the
> first time 'round either, and will want the same main point repeated in
> different ways, as I initially suggested.)
>
> Another player (more like what I take to be the intelligent and
> interested player) might want to ask several times, or as Andrew says,
> will accidentally ask the question several times, in an attempt to
> better problem-solve, or better, appreciate the author's writing and
> its interactivity.

I don't mind pursuing a line of inquiry by asking a NPC more than one
question. But I would like that to be a little more involved than ASK
X ABOUT MURDER repeated. I have to say I don't have a problem here
with the NPC heavy games that I've played (City of Secrets mostly, to
be honest). ASK X ABOUT MURDER might be followed by ASK X ABOUT WAGES
or ASK X ABOUT BLACKMAIL.

If an NPC has something more to say, please don't put it in a topic
about which you've already asked them; that really is as bad as
requiring people to SEARCH CHEST multiple times.

Plotkin raises the valid point that a player is likely to perform the
_action_ of ASK X ABOUT MURDER without realising it, so the designer
should certainly code sensibly for that case. But _requiring_ the same
NPC to be pumped about the same topic is bad form (unless the topics
change state by, for example, the NPC or the PC acquiring some new
knowledge).

As it happens I don't expect to have to ask an NPC about a topic only
once. I often repeat my questions because I know that _some_ games
will require this device, ditto about examining things and searching
things. And I'll often repeat myself just to see what the designer
implemented (which is a pretty latent motivation).


>
> > [A 'crime against mimesis' is] more usually a crime against
> > physics (ie simulation), immersion, or plausibility.
> Probably most useful to call it one or the other, as appropriate,
> thereby identifying which fallacy we're belaboring straight away, and
> with greater precision to boot.

I'm basically with you, use the more precise term, but I have got used
to at least reading and understanding "crime against mimesis".

>
> > But hey,
> > we can call text adventures IF so mimesis seems fine to me.
> You might disagree, but I never thought of 'interactive fiction' as an
> unforgivable abuse of either term. You can call it whatever you like,
> but be aware that jargon or other shorthand very often occludes and
> collapses meaning, whereas taking the time initially to call what one
> means by the right name -- this avoids lots of confusion from the
> start. It's not always easy to figure out the correct formulation of a
> question, but one may remark when certain jargony words are discovered
> to be inherently misleading.

Agreed. Sorry, cheap pot-shot.

As it happens I think it's misleading to call ADVENT interactive
fiction, whereas the term seems more appropriate for something like
City of Secrets. You might notice that I prefer to use "work" when I
think I might offend by using the term "game".

drj

steve....@gmail.com

unread,
Oct 26, 2006, 8:17:35 PM10/26/06
to
d...@pobox.com wrote:
> If an NPC has something more to say, please don't put it in a topic
> about which you've already asked them[.]

I think you can only mean that if an NPC has something *of essence* to
say, don't bury it. Obviously, that was one major point of my initial
post. Yes, don't bury it, and what are the strategies for not burying
essential information.

>From the follow, I conclude you're not speaking generally against
alternate prose upon repeated queries...

> Plotkin raises the valid point that a player is likely to perform the
> _action_ of ASK X ABOUT MURDER without realising it, so the designer
> should certainly code sensibly for that case.

... that is, I assume you're not arguing that each topic should be
answered with the same prose each time. Good, and I agree.

====

This raises an interesting possibility for further discussion. If we
consider the information separate to the topics, we can engage in some
drama management.

Say we have seven crucial points that the player must digest before
solving the mystery successfully.

(Obviously we don't want to say these points only once. Please nobody
respond to this post by arguing something inane like this.)

Though we might not want to make obvious the seven crucial points, we
might also dislike forcing the reader down each nook and cranny of the
space of interaction.

Maybe what we can do instead is write several versions of each point's
discovery, and repeat them according to some plan, within the framework
allowed by the player's interaction.

So for example, one of our golden seven is that "the butler *claims*
that he was away from his room during dinner." (In fact he wasn't, a
fact we can glean elsewhere, and this leads to the resolution of the
mystery by some ingenius double-turn.) We want to produce this point
repeatedly, in case our feeble-minded player didn't notice the point
the first few iterations, or in case (as Jim proposes) they haven't
played the game in a week and forgot the salient points.

How do we produce the point repeatedly? Perhaps not by saying it
repeatedly when the player asks the butler about his whereabouts during
dinner, but by saying it periodically, and innocuously, in different
ways, when the player happens to ask the butler about his room, or
about his love of chess, or asks the maid about the dinner, etc.

Obviously one wouldn't want to beat the facts to death, so the
frequency of each fact can be kept separately from the conversation
response, and the player only gets a 'butler did it' golden-seven point
every n-questions.

There's at least one problem with this theory, namely what to do if the
'n' hasn't arrived but the player thinks he's exhausted each of the
avenues where a golden-point has been hidden.

But with further thought I think the algorithm can be fixed to
accomplish this. My only point here is that the conversation and the
disclosure could be tracked separately.

Basil Chillingworth

unread,
Oct 26, 2006, 9:10:03 PM10/26/06
to
steve....@gmail.com wrote:

> Emily Short wrote:
> > Default User wrote:
> > > d...@pobox.com wrote:
> > > > Having to type
> > > > ASK BUTLER ABOUT MURDER
> > > > 4 times in a row sounds about as thrilling as having to type
> > > > SEARCH THE CHEST
> > > > 4 times in a row. Patently.
> > >
> > > To me, that's worse than a "guess the verb" puzzle. It's "guess the
> > > number of times to ask".
> >
> > There's a bit like that in Anchorhead
>
> Oh my GOD I cannot believe you're missing the point so badly. NOBODY
> suggested that you should have to ask the same question several times
> before receiving the necessary information. The point was that the
> necessary information should be given from the beginning, and though
> varied, repeated each time. Obviously the world is populated by readers
> who require the same point be repeated over and over.

Do you have an advanced degree in Bitchiness, or are you just an
enthusiastic hobbyist?

> But while we're on the subject, yes I hated that similar thing, in one
> of Baggett's games, where you had to perform multiple 'search'es before
> finding the scroll on the wizard's corpse (or something like that).

Don't we all (hate to frisk wizards's corpses for scrolls)?

hmbe...@gmail.com

unread,
Oct 26, 2006, 9:57:57 PM10/26/06
to
Your, uhm, 'challenge' lies in that you are thinking like a programmer,
or a mathematician or something, instead of thinking like a writer, or
an author, or a player.

You are also depreciating and condescending on the player, not
appreciating the effort that they are putting forth and viewing them
not as people enjoying an experience but rather like mice to be led
around your maze and mde to jump through you hoops.

These two factors will produce a very stilted, awkward,
artificial-feeling game. That kind of game where even though
programatically everything looks fine, there's just that feeling of
'uhm' that doesn't let you get into it. I've seen it happen a lot, both
in IF and other types of games. Just overthinking the whole thing.

You already have everything you need to produce realistic conversations
even with repetitions. Just go up to someone and ask them something,
for example 'where's your car?', they'll say, 'it's over there'. Then
five minutes later go up and ask them, 'sorry, where did you say you
car was again', they'll say, 'dude, I already told you, it's over
there.'

So you note that down in your little notepad and you go 'oh my god I
can't believe it, I actually know how someone answers the second time
someone asks them the same thing! Im a frickin GENIUS!' So then you
code the butler so that the second time, he says, 'Sir, I already told
you, I am innocent.

Then you go back to you friend and go, 'hey, youre not gonna believe
this but I forgot again, where's your car?!' and they'll go, 'haha is
this a joke?? you're joking right? It's right there.' And the fourth
time, by then he'll probably be over it, he'll roll his eyes and play
along with your research and go, 'over there'.

You code all that into your game. Wow what an accomplishment.

So as you see, there's no need to make complicated random algorthmic
solutions, there's no need to overcomplicate and overthink things, and
there's CERTAINLY no need to bget snippy and condescending and go "Oh
my GOD blablabla" and throw poorly veiled insults at perfectly smart
people who are making perfectly valid points about a less-than perfect
idea.

For the record, having the player get different responses to the same
line of inquiry without making it CLEAR that there's no more
information to be had (as would happen if the player got randomized
responses that all said the same thing) is a BAD IDEA. BAD. As a
player, you will of course feel forced to keep asking, since you're
getting different responses, only without getting any new information.
Only when you start getting repeat responses (which could actually take
a long time since they are *randomized*) wil you realize the NPC never
said anything beyond the first text, and then you will really feel
gipped.

Of course, as long as you keep equating the players and fellow authors
with lab rats, you'll never understand this.

steve....@gmail.com

unread,
Oct 26, 2006, 10:45:03 PM10/26/06
to
hmbe...@gmail.com wrote:
> Your, uhm, 'challenge' lies in that you are thinking like a programmer,
> or a mathematician or something, instead of thinking like a writer, or
> an author, or a player.

Good theories of game design requires consideration of programming,
design and writing, which obviously require consideration of the
player. If any of my hpyotheses on game design fail on any of these
levels, please point out where and how.

> You are also depreciating and condescending on the player, not
> appreciating the effort that they are putting forth and viewing them
> not as people enjoying an experience but rather like mice to be led
> around your maze and mde to jump through you hoops.

That's a wonderful way to look at it, very useful actually.

First, though, I don't think that sort-of *slanting* the game towards
narrative (or puzzle) disclosure, drama management as I've suggested,
is itself condescending. (I'm not taking the attitude, say, that the
player is just too dumb to figure it out without our quiet help.) This
particular theory of drama management is just a tool, just an option.
It might make a game easier, probably would I think, but then one could
always come up with a harder game to begin with, or tamper with the
disclosure-algorithm such that it effectively makes the game harder.
That "disclosure-algorithm" is worth thinking about, if it's a good
idea to make the game more consistently reflective of the author's
aims.

The problem with IF is that you never know what the player is going to
do, so tools and techniques which help you control disclosure are worth
discussing. I guess such tools *can* mean you're treating the player
like a lab-rat, by which I think you mean, you're giving them the sense
of interactive control, but separately controlling the disclosure in a
way that undermines the significance of the interactivity. I don't
think that necessarily follows, but I definitely think it's a real
risk, and it's great you bring it up. People who talk about drama
management should definitely bear this in mind!

> These two factors will produce a very stilted, awkward,
> artificial-feeling game.

Yeah, well, it's true with computers, any degree of awkwardness is
possible. The technique I mention is a possible way to reduce the
artificiality. Misused, it might increase, but used well, it might
decrease the sense of artificiality.

> That kind of game where even though
> programatically everything looks fine, there's just that feeling of
> 'uhm' that doesn't let you get into it. I've seen it happen a lot, both
> in IF and other types of games. Just overthinking the whole thing.

Could you be more specific?

> [H]aving the player get different responses to the same


> line of inquiry without making it CLEAR that there's no more
> information to be had (as would happen if the player got randomized
> responses that all said the same thing) is a BAD IDEA.

Well, I personally don't favor information-oriented games. The
structure is that the information is always somewhat hidden, and you
never know what you might have missed. That's basically unavoidable.
Should that be minimized? I don't know that it can be. I think the
not-knowing, the unclearness, is at the heart of that sub-genre.

> As a
> player, you will of course feel forced to keep asking, since you're
> getting different responses, only without getting any new information.

You make it sound difficult to understand on a meta-level what's going
on. I didn't imagine it so.

> Only when you start getting repeat responses (which could actually take
> a long time since they are *randomized*) wil you realize the NPC never
> said anything beyond the first text, and then you will really feel
> gipped.

It depends on the presentation, I would think. If it well written,
they'll probably appreciate the efford, or so one would hope.

But this is all well beyond my point. As I said,

JDC

unread,
Oct 26, 2006, 10:48:22 PM10/26/06
to

hmbe...@gmail.com wrote:
[responding to a previous post]

> Your, uhm, 'challenge' lies in that you are thinking like a programmer,
> or a mathematician or something, instead of thinking like a writer, or
> an author, or a player.
>
> You are also depreciating and condescending on the player, not
> appreciating the effort that they are putting forth and viewing them
> not as people enjoying an experience but rather like mice to be led
> around your maze and mde to jump through you hoops.

This made me think of a big difference between IF and static
literature: In IF the player is responsible for moving the story
forward by his actions, whereas a reader can just keep reading if he
has missed something. I have had the experience of finishing a novel
and realizing that I had missed some very significant point until the
end, but this can be okay if it all makes sense in the end. (Slight
spoilers for "Lolita" ahead). When I read "Lolita", for instance, I
didn't really pay too much attention to the Foreward, thinking (I
guess) that it was there as a framing device or for atmosphere or
something. When I got to the end I felt positively stupid that I hadn't
figured things out earlier, since the Foreward really spells it out
pretty clearly (at least in retrosepct).

So in static fiction, I would generally feel it is my fault if I have
missed something. But in a work of IF, if the player misses something
he may be unable to advance in the game, so that it comes grinding to a
halt (or he needs to use a walkthrough). You don't necessarily have
that "it's all explained the end" sort of experience. An IF author
probably has more of a responsibility to make sure that the player
hasn't just missed something, either by repeating the information or
providing it in multiple ways, allowing multiple solutions to puzzles,
etc. Not necessarily everything, but enough so that the player can
reach the end, where hopefully the rest falls into place.

-JDC

Jim Aikin

unread,
Oct 27, 2006, 1:09:26 AM10/27/06
to
<steve....@gmail.com> wrote in message
news:1161890241.2...@i3g2000cwc.googlegroups.com...

>
> In other words, where you have a large chunk of text, you probably want
> to break it up into smaller bits of text, and present that information
> as a result of varying types of interaction.

Maybe. While I agree with the general principle that long expository lumps
are not desirable in IF, breaking essential information up and forcing the
player to figure out how to get at it is not always going to be appreciated.
Even if the player figures it out the first time ... my original question
was about repetition, not expository lumps. So if the player figures out how
to get all of the relevant facts once (via some conversational trajectory or
gambit) should we force her to remember two weeks later what gambit she used
(assuming she even knew what it was she did right the first time)?

Seems to me the courteous thing is to present the information in a concise
manner (which might, on rare occasions, be several paragraphs in length) and
then get on with the game.

> Well, shuffled event list sounds like the best solution. (The butler
> says one of maybe 4 different things, chosen randomly each time through
> the order.) Each variation can provide a sufficient version of the the
> important bit of information (assuming we're talking about a mystery or
> a puzzler), so the player who asks once gets the meat, while a player
> who asks multiple times will get the reward for exploring the
> interactivity of the piece.

I'm not convinced. First of all, you're suggesting that I do four times as
much writing! My project is way too verbose as it is.

Second, what you're suggesting perhaps falls under the heading of what
Fowler -- I'm pretty sure it was Fowler -- called "elegant variation."
Elegant variation is a school of writing that holds (quite amateurishly and
perhaps unconsciously) that the reader hates repetition. An author who is
enamored of elegant variation won't refer to Janet throughout the text as
"Janet." He'll refer to her first as "the mother of three," then as "the
auburn beauty," then as "the Swarthmore graduate," then as "Bob's wife." The
poor befuddled reader can't tell whether one woman is being referred to, or
five.

Plus, the poor IF player now has to read all of the extra text to make sure
that the damned author didn't slip a bit of vital information into the third
repetition. That's cruel. I prefer to confine my cruelty to the puzzles I
design.

--Jim Aikin


steve....@gmail.com

unread,
Oct 27, 2006, 11:00:30 AM10/27/06
to
Jim Aikin wrote:
> <steve....@gmail.com> wrote in message
> news:1161890241.2...@i3g2000cwc.googlegroups.com...
> >
> > In other words, where you have a large chunk of text, you probably want
> > to break it up into smaller bits of text, and present that information
> > as a result of varying types of interaction.
>
> Maybe. While I agree with the general principle that long expository lumps
> are not desirable in IF, breaking essential information up and forcing the
> player to figure out how to get at it is not always going to be appreciated.

I wasn't recommending the use of force. Providing extra is good, and
putting some version of the essential in each fragment has some obvious
benefits.

> Even if the player figures it out the first time ... my original question
> was about repetition, not expository lumps.

My answer is that they're related problems.

> So if the player figures out how
> to get all of the relevant facts once (via some conversational trajectory or
> gambit) should we force her to remember two weeks later what gambit she used
> (assuming she even knew what it was she did right the first time)?

It should be easy to review the facts, which appear in various form but
consistently and repeatedly.

> Seems to me the courteous thing is to present the information in a concise
> manner (which might, on rare occasions, be several paragraphs in length) and
> then get on with the game.

That's one alternative, but does not exclude the other.

> > Well, shuffled event list sounds like the best solution. (The butler
> > says one of maybe 4 different things, chosen randomly each time through
> > the order.) Each variation can provide a sufficient version of the the
> > important bit of information (assuming we're talking about a mystery or
> > a puzzler), so the player who asks once gets the meat, while a player
> > who asks multiple times will get the reward for exploring the
> > interactivity of the piece.
>
> I'm not convinced. First of all, you're suggesting that I do four times as
> much writing!

Not at all. I'm suggesting that you do exactly as much writing as
before, but break it up so the player encounters the text in more
various interactions. Basically I'm recommending you make the game more
interactive, but with the same amount of prose.

> Second, what you're suggesting perhaps falls under the heading of what
> Fowler -- I'm pretty sure it was Fowler -- called "elegant variation."
> Elegant variation is a school of writing that holds (quite amateurishly and
> perhaps unconsciously) that the reader hates repetition. An author who is
> enamored of elegant variation won't refer to Janet throughout the text as
> "Janet." He'll refer to her first as "the mother of three," then as "the
> auburn beauty," then as "the Swarthmore graduate," then as "Bob's wife." The
> poor befuddled reader can't tell whether one woman is being referred to, or
> five.

That's not what I'm suggesting.

> Plus, the poor IF player now has to read all of the extra text to make sure
> that the damned author didn't slip a bit of vital information into the third
> repetition. That's cruel. I prefer to confine my cruelty to the puzzles I
> design.

I don't think it should be difficult to telegraph to the player what
the essential information is, and what's decorative and for the
enjoyment of the interaction -- particularly if your essential
information is repeated in various forms. They'd quickly get the hang
of things, I expect.

Glenn P.,

unread,
Oct 30, 2006, 1:30:04 AM10/30/06
to
On 25-Oct-06 at 12:58am -0000, <rai...@musicwords.net> wrote:

> Sometimes NPCs need to make longish statements, for expository reasons.

> > ASK BUTLER ABOUT MURDER

> ...might give you a rather long text, for instance.

> One of my testers was bothered by the fact that if he asked an NPC about
> a topic, he'd get the same block of text each time. This, of course, is
> a crime against mimesis. In real life, the NPC would say something like,
> "I've already told you all I know about that."

> But what if the player is returning to a saved game after days or weeks
> and wants to refresh her memory? [ Snip... ]

Probably the most elegant way of handling this, it seems to me, would be
to set up some way that the game can tell whether the current session is
a "restored game" or not; and, if it is, clearing the variable which
determines whether the relevant information is or is not "already given".

Even better would be if -- in ADDITION to the above -- the game were able
to determine HOW LONG after the "save" it has been restored. For example,
if the game determines that it is running a "restored" session, AND that
more than twenty-four hours have elapsed since the "save", THEN (but ONLY
then) would it clear the relevant variable.

That way, whenever the player restores a game after a substantial amount
of time, information can be repeated, as if it is being provided for the
very first time.

Care should of course be taken that this is the ONLY effect on a restored
game. In fact you might want to set up a variable SPECIFICALLY to govern
the giving of information, and that ONLY, to make sure that nothing else
is affected.

From a TECHNICAL standpoint, I'm not sure such a thing is even possible.

The "SAVE" command would probably need to be "hacked" such that it assigns
a value to a special variable just before the actual "save" procedure is
begun, so that the variable will "alert" the game to the restored status
once it has been restored.

-- %%%%%%%% "Glenn P.," <C128UserD...@FVI.Net> %%%%%%%%
_____ -----------------------------------------------------------
{~._.~} "...Just *where* are we going?"
_( Y )_ "To the Tower! To Rassilon! The single greatest
(:_~*~_:) figure in Time Lord history!"
(_)-(_) "Is that where he lives?"
========= "Uh, not exactly, Brigadier: it's his tomb."
========= --"Dr. Who" ("The Five Doctors")

:: Take Note Of The Spam Block On My E-Mail Address! ::

Andrew Cowper

unread,
Oct 30, 2006, 6:16:01 AM10/30/06
to
"Jim Aikin" <rai...@musicwords.net> writes:

> Sometimes NPCs need to make longish statements, for expository reasons. 'ask

> butler about murder' might give you a rather long text, for instance.


>
> One of my testers was bothered by the fact that if he asked an NPC about a
> topic, he'd get the same block of text each time. This, of course, is a
> crime against mimesis. In real life, the NPC would say something like, "I've
> already told you all I know about that."
>

I should just repeat the text. No one complains when the same text
comes out after you type look again and again, and you could say that
continually looking would reveal different details each time you do
it.

--
Andrew Cowper

Pick a different user name to email me.

--
Posted via a free Usenet account from http://www.teranews.com

quic...@quickfur.ath.cx

unread,
Oct 31, 2006, 11:36:37 AM10/31/06
to
On Mon, Oct 30, 2006 at 06:16:01AM -0500, Andrew Cowper wrote:
> "Jim Aikin" <rai...@musicwords.net> writes:
>
> > Sometimes NPCs need to make longish statements, for expository
> > reasons. 'ask butler about murder' might give you a rather long
> > text, for instance.
> >
> > One of my testers was bothered by the fact that if he asked an NPC
> > about a topic, he'd get the same block of text each time. This, of
> > course, is a crime against mimesis. In real life, the NPC would say
> > something like, "I've already told you all I know about that."
> >
>
> I should just repeat the text. No one complains when the same text
> comes out after you type look again and again, and you could say that
> continually looking would reveal different details each time you do
> it.
[...]

And it would be extremely annoying otherwise. As a player, when I LOOK
around, LOOK AT or EXAMINE something, I expect the PC to do what is
reasonable, that is, look at that thing carefully enough to notice
anything unusual that he/she will ever notice, at least until something
changes in the game. Hiding details that requires the player to guess
that the solution is to EXAMINE something multiple times is a big
turn-off.

I'd say that for NPC conversations, having the information readily
available (once the player has done what is needed to access that
information) is a must. Ideally, the game would summarize what the
player has already learned:

>ASK BUTLER ABOUT MURDER
The butler has already told you what he knows about the murder,
that XXX was at the scene when YYY happened, and [...]

Repeating the entire text again is acceptable, although less ideal. Not
repeating the text at all is unacceptable: I've encountered too many
games that not only expects you to remember every last detail of what an
NPC said, but also frequently uses the clear-screen function so that I
can't scroll back to review previous text. Very annoying.

As for author effort, I appreciate that summarizing a conversation takes
a lot more effort; but then attention to detail and polish is what makes
a good game an excellent one. If the required effort becomes
prohibitive, maybe one could employ devices such as the PC keeping a
notebook that can be consulted at will, or summarize only key
conversations and simply repeat less important ones verbatim. In any
case, never annoy your players by withholding past information with the
expectation that they have photographic memory.


QF

--
There are 10 kinds of people in the world: those who can count in
binary, and those who can't.

Reply all
Reply to author
Forward
0 new messages