[All] Descriptions
Brian Ronk
short one liners, or long and extremely descriptive. Is there a
preference in how players would like them?
Eric Eve
news:1097416107....@c13g2000cwb.googlegroups.com...
> I'm curious how people like descriptions in their games. Are they
> short one liners, or long and extremely descriptive. Is there a
> preference in how players would like them?
I'm not sure how one can give a general answer to this; it depends
on the nature of the game, the style and skill of the writer and so
forth. It also depends what's being described (are you thinking of
rooms or objects, for example?). Personally I think that too many
one-line room descriptions in a game are likely to give it a rather
sparse feel (though that may be appropriate for a particular game),
while over-lengthy descriptions are liable to make the player feel a
bit bogged down (it's better to have a room description mention
objects which the player can then examine for more detail). Perhaps
about three or four sentences for a room description is a good
average to aim for -- assuming it's meaningful to suggest an average
at all. What you're aiming at ideally is to impart the information
you need to impart and create the atmosphere you want to create with
the maximum economy of words, where 'maximum economy' doesn't
necessary equate to 'minimum number' if brevity detracts from
clarity or mood.
But I feel I'm simply stating the blindingly obvious, so I'll leave
it to others to say something more insightful.
-- Eric
Tomasz Pudlo
> I'm curious how people like descriptions in their games. Are they
> short one liners, or long and extremely descriptive. Is there a
> preference in how players would like them?
I like my descriptions to describe *how* instead of *what* the PC sees. You
can do that by describing how light enters the location (Kaged) or make the
descriptions character-dependent, expressing the views and tastes of a
well-defined PC (Varicella). I also like descriptions that are spatially
specific, describing objects in relation to other objects.
"You are standing in the king's bedroom. You can see a king-sized bed. You
can also see the Sword of Porkhead here. The only exit is south."
How do I see these things? Is there a lightsource, a torch, a window
perhaps? Where exactly is the Sword of Porkhead? On the bed? On the floor?
Leaning against a wall? Things like that really help to visualise the world.
It's also much more fun to have a world described to you through someone
else's eyes than just being presented with a list of nouns to interact with.
As IF is slowly moving on from its Neanderthal past, descriptions like the
one above are becoming quaint relics from a happily bygone era.
Kevin Venzke
"Brian Ronk" <paladi...@gmail.com> wrote in message
news:1097416107....@c13g2000cwb.googlegroups.com...
> short one liners, or long and extremely descriptive. Is there a
> preference in how players would like them?
If descriptions (esp. room descriptions) are very long, and I get
stuck in the game, I will start to suspect that my problem is that
I've forgotten to examine something in some room's fine print.
Then I hope that the author has permitted "take all" (for instance)
to include absolutely everything implemented in a room.
(Side request: Please permit me to type "x purse and wallet." You
don't have to let me type "all." Rejecting multiple direct objects just
forces me to type "x purse. x wallet.")
Maybe a good length depends on how fast I'm expected to get
through each room. A long description won't be so bad if
everything mentioned is implemented. Otherwise I'll spend as much
time reading it as reading parser errors while I search for needles
in the haystack.
Kevin Venzke
samwyse
> I'm curious how people like descriptions in their games. Are they
> short one liners, or long and extremely descriptive. Is there a
> preference in how players would like them?
Great games tend to use descriptions of various lengths:
Short: "You are in a maze of twisty little passages, all alike."
Medium: "You're at a low window overlooking a huge pit, which extends
up out of sight. A floor is indistinctly visible over 50 feet below.
Traces of white mist cover the floor of the pit, becoming thicker to the
right. Marks in the dust around the window would seem to indicate that
someone has been here recently. Directly across the pit from you and 25
feet away there is a similar window looking into a lighted room. A
shadowy figure can be seen there peering back at you."
Long: the Volcano View
Brian C. Lane
Hash: SHA1
Yeah, Infocom really had a knack for getting it just right. Some
situations call for short others for long. But the descriptions should
make sure that what they are describing is necessasary for the story, I
really dislike it when a bunch of extra fluff is thrown in without any
useful purpose.
As with everything, especially writing, there is a balance that you have
to find for yourself and your game.
Brian
GUEtech.org admin
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFBaazWIftj/pcSws0RAkvrAJ4w0BWl0YLZ4iQz3Z232j3e+QLOowCbBK1m
eQjeZCADP5r3xzJW/5notLI=
=XlT4
-----END PGP SIGNATURE-----
ems...@mindspring.com
> short one liners, or long and extremely descriptive. Is there a
> preference in how players would like them?
It usually puts me off to encounter an extremely long description
(more than a page), but beyond that the word-count is not a big deal,
in my opinion. What's more important is the interactive complexity of
the description: I don't want to read about more things than I can
remember to examine.
To put this in very reductive terms, a description with too many
concrete nouns in it can be overwhelming, because (as a player) I have
to keep all those objects in mind as things to examine. Gareth Rees
gives a good example of why this is in his article on IF writing
(http://www.doggysoft.co.uk/inform/write/prose.html) though I disagree
that "fine writing" is identical with elaborate description. But
object density is still something to think about. Graham Nelson
suggests four interactive objects per room as a plausible number
(http://www.geocities.com/aetus_kane/writing/coa.html); I find this a
low estimate and seem to average closer to six to ten items per room,
but some of those items are usually parts of other objects, so that
they don't necessarily appear in the initial description. But it's
still probably a good idea to keep the number relatively low.
I also try for some variety between rooms. My impulse -- YMMV -- is
to try to keep major map junctures relatively simple; there might be a
thing or two to pick up, or some scenery to look at, but if there's a
room with five exits, I'm already going to be spending a lot of time
describing where they all go, and that's going to occupy a lot of the
player's memory on the first reading.
There's some flexibility in this. A large number of nouns is easier
to deal with where it is obvious that many of them are abstract or
nonphysical (darkness, anger, etc., are unlikely to be interactive
except in very surreal games); where the author manages to convey a
distinction between interactive items and scenery; or where there is
such a specific goal at every moment that, as a player, I know exactly
what I need to look at. If you try J. Robinson Wheeler's "Centipede",
you'll find that there are many more items than you have *time* to
look at, because the action keeps moving forward. If you don't want
to die, you'll concentrate on the important objects and ignore most of
the rest; having more things around than you have time to look at
contributes to the game's overall sense of haste and claustrophobia.
Here's a longish description that, IMO, does just about everything
right, from "Gourmet" by Aaron Reed:
===
Seating Area
Some laughed when you announced your intentions to open a five-star
gourmet eatery with a 1940s decor, but as you glance around fondly at
chrome-buffed tables, World War 2 recruitment posters, the big band
quintet in the corner, and the pneumatic tube food delivery system,
you wonder, who's laughing now? It all comes together to create an
ambiance that Joy Saunders of Haute Today described as "insultingly
passé." You console yourself with the knowledge that Haute Today has
really gone downhill in the last few years.
The kitchen lies west through a pair of classy swinging doors, a
unisex restroom may be entered to the north, and a small supply closet
lies to the south. The street entrance to your establishment lies past
some potted plants to the east.
A family of four sits at table one, and tables two and three are
unoccupied.
====
Part of the wordcount is devoted to PC attitude and experience. (What
player is going to try to examine Joy Saunders, or expect her to be
implemented? But this is useful information nonetheless.) A
manageable four interactive objects appear in that first paragraph --
tables, posters, band, pneumatic food delivery system -- but we're
given a sense of their relative positions in space and the mood evoked
by the combination. Exits appear separately, but aren't boring; the
ones you're likely to use (kitchen, bathroom, closet) are in a
different sentence from the one you can't use right now (street exit).
And the final line of the description focuses our attention on the
most important interactive item in the room, the family at table one.
Other references you may find interesting:
Mike Berlyn on description variety:
http://www.xyzzynews.com/xyzzy.17d.shtml. There're some good pointers
here, though I think I find his final example a little excessive.
Matter of personal taste.
Stephen Granade on what a description can accomplish within a game:
http://www.brasslantern.org/writers/iftheory/descriptions.html.
J. Robinson Wheeler on description styles, organization, and emphasis:
http://raddial.com/if/theory/territory.html.
Richard Bos
> samwyse wrote:
> | Great games tend to use descriptions of various lengths:
> |
> | Short: "You are in a maze of twisty little passages, all alike."
> |
> | Medium: "You're at a low window overlooking a huge pit, which extends
> | up out of sight. A floor is indistinctly visible over 50 feet below.
> | Traces of white mist cover the floor of the pit, becoming thicker to the
> | right. Marks in the dust around the window would seem to indicate that
> | someone has been here recently. Directly across the pit from you and 25
> | feet away there is a similar window looking into a lighted room. A
> | shadowy figure can be seen there peering back at you."
> |
> | Long: the Volcano View
>
> Yeah, Infocom really had a knack for getting it just right.
Erm... yeah, but not quite so right as C&W...
Richard
samwyse
Ditto.
I just sat down with a copy of the FORTRAN source and analyzed the room
descriptions. Here are the results:
1 Line 71 (includes the mazes and several DEAD ENDs)
2 Lines 32
3 Lines 22
4 Lines 5
5 Lines 4
7 Lines 3
8 Lines 1 (the end game)
11 Lines 1 (the end game)
19 Lines 1 (the volcano view)
So, it is obviously possible to create a compelling game with fairly
short descriptions. Admittedly, the technology of the time required
brevity, but Adventure turned the tables on the limits to create
something memorable. Let me explain...
I first played Adventure on an 80x24 glass TTY over a 300 baud
connection. At that speed, an 80 character line takes about 2-1/2
seconds to be transmitted. The volcano room was only reached after
exploring dozens upon dozens of rooms with short descriptions. Entering
it, the description slowly filled the screen, line by line by line,
taking about 40 seconds to complete. Today, you might think that to be
intolerable. But that first time I entered the room, before even half
of the text had been transmitted, I think that I had stopped breathing,
captivated by a description that by its very size created the illusion
of being even more detailed than it really was.
David Jones
>
> (Side request: Please permit me to type "x purse and wallet." You
> don't have to let me type "all." Rejecting multiple direct objects just
> forces me to type "x purse. x wallet.")
But "x purse. x wallet" is shorter than "x purse and wallet". You're
wasting your own time. :)
David Jones
David Jones
> connection. At that speed, an 80 character line takes about 2-1/2
> seconds to be transmitted. The volcano room was only reached after
> exploring dozens upon dozens of rooms with short descriptions. Entering
> it, the description slowly filled the screen, line by line by line,
> taking about 40 seconds to complete. Today, you might think that to be
> intolerable. But that first time I entered the room, before even half
> of the text had been transmitted, I think that I had stopped breathing,
> captivated by a description that by its very size created the illusion
> of being even more detailed than it really was.
Absolutely bang on. Finding the volcano was a major emotional event
(for a video game) and the most important device in acheiving that was
the contrast, in length, of the room description. I felt that the
authors had managed to achieve a response in the player, me, that the
protagonist would feel. One of shock and awe and of having to stop
and take stock of the situation. I think, for me, that level of
emotion response has been achieved about 3 or 4 times whilst playing
video games (text adventures included).
Amusingly, these days there might be some mileage to be had in
suddenly dropping down to one line descriptions.
drj
Kevin Venzke
"David Jones" <d...@pobox.com> wrote in message
news:3e298a5b.04101...@posting.google.com...
Wouldn't have noticed that. But if the verb isn't "examine," it's not
the case anymore. A number of games seem to be disallowing multiple
objects for many verbs.
Kevin Venzke
Ice
"x purse and wallet" is ambiguous under many IF grammars that support
command catenation. For example, does this imply:
"x purse and x wallet"
or
"x purse."
"wallet."
What if I name an object "leave"?
x purse and leave
Does this resolve to examine purse, and then leave the room? Or
examine purse and examine leave? The word 'and' cannot be used in
multiple contexts and remain fully useful. It must either be used for
catenation of sentences, or as an expansion of a verb. If we assume
it is always an expansion, then the resolution is possible.
eg.
examine bob and joe and frank
expands to:
examine bob / examine joe / examine frank
It is hardly the objective of the parser to accept the minimal form of
all commands; it is usually some intermediate form of the expansion
that the player will use.
Cedric Knight
> Spider and Web (Inform): "take door and jump" produces "You can't see
> any such thing."
TAKE DOOR THEN JUMP is broken up as two commands, though. In the next
Inform library release TAKE DOOR AND JUMP will be too.
That is, unless there also happens to be a jump lead in the room.
FWIW I'm not in favour of EXAMINE BREADFRUIT AND POMEGRANATE because the
descriptions might become muddled.
CK
Ice
> "Ice" <icedr...@icedragon.net> wrote in message
> news:b37d7688.04101...@posting.google.com...
> > > Wouldn't have noticed that. But if the verb isn't "examine," it's not
> > > the case anymore. A number of games seem to be disallowing multiple
> > > objects for many verbs.
> > >
> > > Kevin Venzke
> >
> > "x purse and wallet" is ambiguous under many IF grammars that support
> > command catenation. For example, does this imply:
> > "x purse and x wallet"
> > or
> > "x purse."
> > "wallet."
>
> objects is because the parsers cannot handle them (...anymore?).
I think you misunderstood my point. You can make a parser handle it
(kind of). But it doesn't change the nature of the grammar.
Redefining your rules are the only ways to remove the fact that
multiple parse cases exist. Better parsers will notice this problem
earlier (ie. maybe our parsers are getting better, not worse), and it
can be a hassle to resolve if you plan to complete parsing before
doing object resolution. Almost all ambiguous grammars have instances
of valid parse tables, so I would expect your examples to work in any
parser that accepts one valid table for an ambiguous set.
http://en.wikipedia.org/wiki/Ambiguous_grammar
Andrew Plotkin
>
> You seem to think that the reason people are disallowing multiple
> objects is because the parsers cannot handle them (...anymore?).
> objects that the player may not know about yet.
Actually, while Inform can be rigged to handle multiple objects, I
avoid it because it's a headache. It causes me to have to worry about
the action for one object invalidating, or interfering with, the
action for another object on the same turn. Life is easier if one turn
is one action.
(Of course I have to worry about that for "take" and "drop", but the
more verbs allow multiple objects, the bigger the headache gets.)
--Z
"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."
*
* Make your vote count. Get your vote counted.