[Announce] Rat In Control (was: Spatial cognition)
Mike Roberts
> It would take someone more clever than myself to invent an
> experiment to tease out whether my big-picture-map hypothesis
> or your path-finding hypothesis is the correct model (or if it's
> something else entirely).
...but why should I let that stop me? I've come up with an idea for a
little experiment - in the form of an IF game, of course - that might shed
some light on whether our internal mental models deal more in terms of
absolute or relative directions. The premise of the experiment is that we
ought to have an easier time using the model that's closer to our actual
mental model; and to put "easier" in quantifiable terms, I claim that if
something's easier, you can do it faster.
The experiment is a small adventure game called Rat In Control. It can be
played with either conventional compass directions or with relative
directions ("in front of you," "to your left," etc). It's exactly the same
game either way, and you can switch modes at any time. The idea is that
people can try the game both ways, and compare how long it takes them to
perform the same navigational task, in the same game, using the two modes.
The one that's faster is presumably the one that's easier to use, because
it's less work for our brains, which I claim means it's closer to our
internal mental model.
To reduce the bias from the long experience (most of) us have with the
conventional compass style, the game starts off with a "training" period
that lets you spend a while learning its map in the mode you want to try; so
if you play in relative mode from the start, you'll at least have a chance
to become accustomed to that mode. Once you reach a certain milestone in
the game, it switches to a series of timed "missions," which are just
navigational tasks that it asks you to perform while timing you.
If you want to try it out, I've just uploaded the game to the IF archive -
it's called "spatial.t3", and it'll make its way in the usual fashion from
/if-archive/unprocessed to /if-archive/games/tads.
I have no cognitive science training, so I've probably made all sorts of
errors and invalid assumptions in constructing the experiment. Even if it
has no scientific value, though, it might be interesting in a couple of
other ways. First, it's a rare chance to do an apples-to-apples comparison
of absolute and relative directions, to see how you personally feel about
the two styles. It always seems to me like the eternal controversy over
which style is better is based on very abstract ideas people have; but
matters of taste usually hinge on tiny practical things that the big
abstract arguments completely miss. Having a concrete way of comparing the
two modes head-to-head might bring it down to earth a bit. Second, the
source code (which I also uploaded - "spatial_source.zip", which should end
up in /if-archive/games/source/tads) includes a re-usable relative
directions module, which might be of interest to game authors who want to
experiment with the relative style in their own games.
If you feel like trying it out, I'd be interested in hearing about your
experiences with it. I'll be happy to compile timing data if people send
them to me (and I'll certainly anonymize these before publishing anything).
--Mike
mjr underscore at hotmail dot com
Mike Roberts
> If you want to try it out, I've just uploaded the game to the
In case it's not apparent, this is a tads 3 game, so you'll need a recent
TADS interpreter. You can get the latest for Windows and Unix from
http://www.tads.org/t3dl.htm (look for the Player Kit for Windows, or the
Unix source package); for Macintosh, look to http://www.hypertads.org.
Nikos Chantziaras
news:hO7la.199$y96...@newssvr16.news.prodigy.com...
> [...]
> ...but why should I let that stop me? I've come up with an idea for a
> little experiment - in the form of an IF game, of course - that might shed
> some light on whether our internal mental models deal more in terms of
> absolute or relative directions. The premise of the experiment is that we
> ought to have an easier time using the model that's closer to our actual
> mental model; and to put "easier" in quantifiable terms, I claim that if
> something's easier, you can do it faster.
I tried it. It didn't work out. I really tried, but I ended up drawing a
map (on paper), which means I just converted everything to the usual
compass-directions method. This "left" and "right" thing is very, very
confusing. Okay, this approach works in Doom and Quake (you don't care
about directions there), but in text games it just doesn't. You can't see
anything, after all. You just read descriptions and try to imagine where
you are, which means that this "built-in" mechanism of humans (if there is
one) stays inactive, since there's no input from your senses that could feed
it.
-- Niko
http://members.lycos.co.uk/realnc
Mike Roberts
> mjr:
> > The premise of the experiment is that we ought to have an
> > easier time using the model that's closer to our actual mental
> > model; and to put "easier" in quantifiable terms, I claim that if
> > something's easier, you can do it faster.
>
> drawing a map (on paper), which means I just converted
> everything to the usual compass-directions method. This "left"
> and "right" thing is very, very confusing.
You're confirming my hypothesis, that we model our surroundings with some
internal absolute orientation. If my reasoning is correct that easier means
closer to the internal model, then it would seem that, at least for you, the
compass direction model is closer to your mental model.
> Okay, this approach works in Doom and Quake (you don't
> care about directions there), but in text games it just doesn't.
> You can't see anything, after all. You just read descriptions
> and try to imagine where you are, which means that this "built-in"
> mechanism of humans (if there is one) stays inactive, since there's
> no input from your senses that could feed it.
I think you might be thinking that my hypothesis is that the relative model
is closer - it's the opposite, actually. My claim is that the built-in
mechanism thinks in terms of a big-picture map with a particular (but
arbitrary) orientation.
Note that I do think short-circuiting the visual sense is an important
element of the experiment. Text games make you construct mental imagery
from verbal descriptions, which no doubt has substantial visual cortex
involvement, but for processing higher-level abstractions rather than direct
sensory data. When you're dealing with direct visual input, I'd conjecture
that you process the sensory data into an abstract representation that maps
to your internal arbitrary big-picture orientation, and that this processing
happens at a pretty low level. When you bypass the visual input and use
mental imagery instead, I think you also bypass that pre-processing, so
verbal input is easier to deal with when it more closely matches the
higher-level abstraction of the internal mental model.
Mark J Musante
> You're confirming my hypothesis, that we model our surroundings with some
> internal absolute orientation. If my reasoning is correct that easier means
> closer to the internal model, then it would seem that, at least for you, the
> compass direction model is closer to your mental model.
I haven't tried the game yet.
However, I'd like to add a bit of data to this discussion. When I
play IF, I do maintain a mental map of the game (provided the map
is straightforward and not, say, a twisty maze of passages). But,
significantly, I consider 'north' to be 'ahead', 'south' to be
'behind', and 'east' and 'west' to be 'right' and 'left' respectively.
So, as I'm walking around, if I need to go s, se, s, w (for
example), I think of that as walking backwards, backwards and to the
right, backwards again, and then left. All the while my front faces
north and my back faces south.
Now if the game were to change things (as I imagine your test game
does), and reorient the world based on the direction I last went,
I believe I would get very lost very quickly, not having the
ability to transform text descriptions to the map in my head.
That being said, when I play a first-person 3D game such as Deus Ex
which has a compass built in, I've found that I only ever need
that compass when I'm told 'go west from such-and-such' or 'the
key is to the north of the shack'. If I don't need to know what
direction I have to go, I can navigate maps with incredible ease
just by getting a visual layout of my surroundings.
My impression is that I think of the direction I face as 'forward'
and the world's map turns in my brain as I turn my character. If
I know where I need to go, I don't think 'go forward a bit, turn
slightly left, go some more, turn right, then find the door and
enter it'. I just think 'I need to go that-a-way around the
building because the door is back there'.
Does that add anything to the discussion?
-markm
Mike Roberts
> I haven't tried the game yet.
>
> However, I'd like to add a bit of data to this discussion. When
> I play IF, I do maintain a mental map of the game (provided the
> map is straightforward and not, say, a twisty maze of passages).
> But, significantly, I consider 'north' to be 'ahead', 'south' to be
> 'behind', and 'east' and 'west' to be 'right' and 'left' respectively.
Yes, exactly - that's what I've saying about the mental model having an
absolute but *arbitrary* orientation. The compass directions are a bit of a
red herring, because to my mind they're not true compass directions in the
sense of being aligned with the Earth's magnetic or rotational poles;
they're just a convenient way for the author and player to share a common
reference frame with an agreed definition of "forward." I think the compass
convention came into being because just saying "forward" would confuse
players by leaving "forward" too undefined and nebulous; "north" clearly
conveys that we're talking about a bigger picture, and that "north" in one
location will be aligned the same way on the bigger picture as "north" in
the next location.
> That being said, when I play a first-person 3D game such as Deus
> Ex which has a compass built in, I've found that I only ever need
> that compass when I'm told 'go west from such-and-such' or 'the
> key is to the north of the shack'. If I don't need to know what
> direction I have to go, I can navigate maps with incredible ease
> just by getting a visual layout of my surroundings.
I'm extrapolating beyond what my experiment game could conceivably tell us,
but my theory is that your visual sense is in fact subconsciously creating a
mapping to a big-picture map in your head, and that's what makes the visual
navigation so easy. You could apply my thought experiment about pointing to
the street from your living room to Deus Ex equally well: after walking
around a bit in Deus Ex, in an area you know well, you could point
(virtually speaking) in the direction, relative to your current position and
orientation, in which you think some past landmark lies, right? My claim is
that you can do that, probably completely intuitively, because you've built
this mental model of the overall map, and you have a sense of where you are
in the model and how you're oriented relative to it. But as I said earlier,
that map has an arbitrary orientation, so there's no reason for "west" to
have any intuitive meaning in that map. You have to perform some conscious
work to relate your internal mental model to compass directions.
Ally
I think a 'relative' mode of navigation gets easier when the map "comes
with you" (as in a modern 3D game), but the IF geographic model is
nothing like that in that you more or less 'blink' from one location (or
scene) to the next.
Using 'right' or 'ahead' actually feels unintuitive to me, not because
it's something I wouldn't ever use in the real world, but because the
direction you face doesn't feature in room descriptions, nor do you see
the landscape move around you or have a clear idea as to where you're
going (I'm not happy with these 'traits', though).
I suppose vast, explorable areas are best traversed 'compass-style'
anyway... unless, maybe, you have plenty landmarks to walk toward, but
(unfortunately) you can't usually "walk toward the <landmark>" in IF.
Even if you could, it would still be 'moving toward things mentioned in
room descriptions', only they'd be destinations rather than 'exits'.
I even have a _very_ hard time finding my way through a "classic" Ultima
or Might and Magic dungeon, simply because there's no scrolling and
everything looks the same more or less. Unless I carefully build an
internal compass-based map and literally count my steps, I'm likely to
get lost once I've turned around the first corner.
But I guess I'm just bad at finding my way, regardless of the
environment. In physical reality I usually rely on landmarks: "once I'm
at the library, I know how to get to the cathedral" -- though not really
where the cathedral is in relation to other places in town. No landmark
often means I have no idea where I am.
Similarly, in IF I think along the lines of "going north from the spire
takes me to the bridge" unless I have a map at my disposal. Which is
where orientation actually becomes easier than in the physical world. I
can spend a long time trying to make sense of a 'mounted' map (on a
signpost or the back of a bus stop advertising 'wall'), i.e. one I can't
turn until it faces "my" direction. In IF it doesn't matter which
direction I'm facing.
So I'd rather eliminate "facing" and relative directions from the real
world than make them part of IF :)
~Ally
--
kitzapoo (at) gmx (dot) co (dot) uk
Michel Nizette
> If you feel like trying it out, I'd be interested in hearing about your
> experiences with it.
So, I did the experiment. Without any doubt, COMPASS mode feels more
user-friendly to me than RELATIVE mode, but I found that RELATIVE remains
reasonably playable after some getting used to it.
I also asked myself how I was mentally picturing the game world when I was
playing in one mode or another. Before doing the experiment, I'd have thought
that in both cases I'd picture it as a fixed map with north always pointing
upwards, exactly as I would have drawn it on a sheet of paper. In particular,
in RELATIVE mode, I'd have thought that I'd represent myself as an oriented rat
evolving in that static map, and that I'd more or less consciously compute
relative directions from the absolute ones and from the rat orientation.
However, by actually doing the experiment, I came to the surprising conclusion
that I do have this mental bird's eye representation of the world in COMPASS
mode only, not in RELATIVE mode. In RELATIVE mode, I discovered that the visual
impression that my mind was spontaneously elaborating was a first-person one,
much as though I were playing a car race game. It was the game world that
rotated around me, not the opposite.
So, it looks like my visual representation of the world was actually determined
by the navigation mode I was using, and that I wasn't unconditionally using the
simplest (whatever that means) of the two models. However, this conclusion
rests on the assumption that the visual representation *is* the model, which I'm
not sure is a valid assumption.
--Michel.
Mike Roberts
> So, I did the experiment. Without any doubt, COMPASS
> mode feels more user-friendly to me than RELATIVE mode,
> but I found that RELATIVE remains reasonably playable
> after some getting used to it.
That's pretty much in line with my experiences. I'm surprised at how
playable relative mode is, but it's definitely not as easy to use as the
traditional compass mode.
> I also asked myself how I was mentally picturing the game
> world when I was playing in one mode or another. [...]
> In RELATIVE mode, I discovered that the visual impression
> that my mind was spontaneously elaborating was a first-person
> one, much as though I were playing a car race game. It was
> the game world that rotated around me, not the opposite.
Interesting perspective. When we're in a real-life situation, or when we're
playing a graphical game with a first-person view, our visual sensory input
is obviously in terms of this first-person view. I've gone on at length
before about my hypothesis that our visual cortex translates this view to
the bird's-eye view with some global orientation, so I won't repeat that.
But I will add some speculation about the text-mode environment in RELATIVE
mode: my guess is that you do indeed build a mental image with a
first-person perspective, exactly as you report, and that your brain does
this specifically to take advantage of the visual cortex hardware to do the
rotations to the bird's-eye model. With COMPASS mode there's no need to go
through this intermediate step, since the game relates the environment in
the same sort of bird's-eye global terms that you use deep down. In other
words, I'm thinking the first-person mental imagery you create in relative
mode is a sort of intermediate computation step on the way to the compass
model. This is where my reasoning about the time measurements comes in - I
figure that if there is indeed an extra mental step (conscious or not) in
one mode or the other, then people ought to be consistently slower using the
mode with the extra step.
David A. Cornelson
news:hO7la.199$y96...@newssvr16.news.prodigy.com...
> A little while back, I wrote:
> > It would take someone more clever than myself to invent an
> > experiment to tease out whether my big-picture-map hypothesis
> > or your path-finding hypothesis is the correct model (or if it's
> > something else entirely).
>
> ...but why should I let that stop me? I've come up with an idea for a
> little experiment - in the form of an IF game, of course - that might shed
> some light on whether our internal mental models deal more in terms of
> absolute or relative directions. The premise of the experiment is that we
> ought to have an easier time using the model that's closer to our actual
> mental model; and to put "easier" in quantifiable terms, I claim that if
> something's easier, you can do it faster.
>
Not sure if this is relevent, but I originally wrote Cattus Atrox using
relative directions like RUN, JUMP LEFT, JUMP RIGHT, JUMP FORWARD, and the
beta-testers politely told me it was maddening.
I thought with the lions that typing relative directions would be scarier,
but in the end, the user has to be able to interact with the system in a
standard manner. Text adventures are all about standards where navigation is
concerned. I guess the only way to change this would be to implement very
strong games over and over that used relative navigation and then people
would know when they played that type of game what to expect.....but even
then many would balk.
My two cents.
Dave
Mike Roberts
> Text adventures are all about standards where navigation is concerned.
Probably so, but a major part of my thesis here is that there are good but
perhaps not completely understood reasons why the standards are as they are,
that they're not merely the results of inertial propagation of arbitrary
choices made long ago.
> I guess the only way to change this would be to implement very
> strong games over and over that used relative navigation.
In case I've created any impression I'm out to overthrow the compass regime,
I'm not - I'm just trying to explain it. My feeling is that even very
strong games using relative navigation wouldn't change a critical mass of
minds about it, because compass navigation is actually better, irrespective
of what people are accustomed to.
Valjean
still at least
75% in my brain) is going to go down like a lead balloon because it doesn't
conform to the old "Naughty Elephants Squirt Water" method of navigation.
Val
David A. Cornelson
I don't want to necessarily dispute that thought, but I have another.
It's my belief that people do things out of habit. Habits are formed largely
by copying others as well as repitition. If a person is placed into a system
that is different from what they're used to (breaking their habits), they
respond in various ways. Some people will be confused, some will be angry,
some will be curious, and some will be excited.
I give for example a US traveler, by himself, in a foriegn country. (this
happens to be me in 1995).
I was backpacking through Europe and started in London. Piece of cake.
Everyone speaks english and I was able to maneuver based on my own personal
experiences.
I then took the ferry down to France and a train to Paris. When I arrived I
went down a set of stairs to the Metro (the subway). I bought a 5-day pass
and then proceeded to stare at the Metro map on the wall.
Here is where habits break down to some degree. I did not know how to read
French and certainly proper French names were meaningless to me.
My response was excitement. I felt an enormous excitement at the prospect of
figuring out an entirely new set of rules for navigation. I had been used to
the colors of Chicago's subway system (the red line, the brown line, etc).
In Paris, the lines don't have colors or names. They have destinations,
which is one end of a line. It took quite a bit of practice for this to sink
in, but eventually it became second-nature.
Now I also ran into many other people in my travails and met varying types
of personalities. Some people simply broke down at the prospect of learning
a new system. I saw both young men and women nearly in tears, lost in
various locations in Europe. Some were very pragmatic and aggressive (they
read Frommers and several other guides cover to cover with a highlighter!)
and others were probably the smartest...they found people like me or the
pragmatists and followed us because we'd already figured out the system.
NOTE: The adventurers (me) and the pragmatists generally went well out of
our way to help the lost travelers. It became sort of an unwritten
task...."hey are you lost? where are you going? I think I can help..."
So there are two things I'd point out in attempting a non compass oriented
navigation system in IF. Some people will abhor it instantly and run
screaming. Some will get excited at the prospect of something new and
different and some will see the story as the main point and deliberately
"ignore" the navigation to finish the story.
But again, habits are more important than anything because after hanging
around me or a pragmatist for a day, the people that were once afraid became
emboldened and some grew into leaders. Once they got their bearings (with
the help of others) they seem to develop entirely new methods of attacking
problems.
I went on two trips like this to Europe in 1993 and 1995 and I observed
these behaviours both times. It may or may not be relelvent to IF
navigation, but it could be...
Dave
Michel Nizette
> I'm thinking the first-person mental imagery you create in relative
> mode is a sort of intermediate computation step on the way to the compass
> model.
Indeed I didn't have to re-learn the game map when I switched from compass mode
to relative mode, so that I was obviously using my previous knowledge of the
game world (which I had acquired, and probably stored, in "compass" form) in
order to generate the first-person views that helped me to travel quickly in
relative mode.
Anyway, I find it amazing that the navigation mode I am using exerts such a
drastic influence on my mental "camera", and thus on the way I feel immersed in
the scene.
--Michel.
OKB (not okblacke)
newsserver hasn't got that post so I'm replying to some other post.)
I found "Rat in Control" very interesting (and I think "Rat in
Control" is a super-cool title). The relative mode left me utterly
bewwildered. I managed to win more quickly than I had a right to just
by bumbling into the right places.
I probably had some subconscious mental image of the map from
relative mode, but to be honest, when I restarted in compass mode I
didn't feel I had ANY idea where anything was. I basically re-explored
the map, but this time I quickly got an idea of how it fit together and
I won much more quickly. I think the only direct way that my relative
playthrough helped was that it alerted me to which objects and locations
were important; it didn't form any model of the world in my mind,
though.
Anyway, I basically agree with your (MJR's) hypothesis about human
navigational cognition. I certainly perceive this as the way I
navigate, and there are a couple of examples that I think illustrate
what may be going on:
When I'm walking around the college campus here -- trying to get to
a building that I rarely go to, for instance, and whose location I'm
unsure of -- I tend to think of which general side of the campus it is
on. Is it on the "Santa Barbara" side or the "Isla Vista" side (those
being two communities on opposite ends of the campus). These boundaries
correspond roughly to west and east, but I never think of them in those
terms. I would imagine that most people navigate in this way: they
choose certain boundaries for the region they want to navigate in, and
identify them in terms of landmarks that are OUTSIDE the boundary. Then,
inside the region, they think in terms of directions relative to those
boundaries.
In fact, this is the same way that freeway road signs work (at
least here in California). When you get on the freeway in Southern
Calfornia, heading north, even if you start as far south as Los Angeles,
you don't just see a sign that says "north", because that wouldn't be
very helpful. Instead you see a sign that says "San Francisco", even
though San Francisco is hundreds of miles away. Nevertheless, since
most people (even those unfamiliar with California) are aware of where
San Francisco is relative to LA, they are able to mark it as one
boundary of the region they're in and orient themselves that way. (The
signs do say "north" as well, but almost never do they JUST state the
direction -- they almost always give a point of reference too.)
--
--OKB (not okblacke)
"Do not follow where the path may lead. Go, instead, where there is
no path, and leave a trail."
--author unknown
Jessica Knoch
>
> In fact, this is the same way that freeway road signs work (at
> least here in California). When you get on the freeway in Southern
> Calfornia, heading north, even if you start as far south as Los Angeles,
> you don't just see a sign that says "north", because that wouldn't be
> very helpful. Instead you see a sign that says "San Francisco", even
> though San Francisco is hundreds of miles away.
I remember that. Also interesting are the signs they put up around the
495 loop that circles Washington D.C. Because sometimes you have to
go north (or south) to get east (or west), they often put the names
of cities on the signs so that a driver chooses between "495-Arlington"
and "495-Springfield" or whatever. Unfortunately, a large chunk of
people see those signs and think, Springfield? Do I have to drive
through that to get to Maryland, or is it the other one? (Of course
either will get you there eventually; it's just a question of time
and looping around through the traffic.)
At some on-ramps (very VERY few that I have seen), they have these
clever signs that say "Inner Loop" or "Outer Loop," so that the
driver unfamiliar with the local landmarks can choose Inner or Outer
(which will always be a constant, even as you go from facing north
to east to south, you're still on the Inner Loop). Of course, the
first time I saw that choice (inner/outer), it took me a full five
seconds to figure out whether I wanted to be driving clockwise or
counter-clockwise.
--
Jess K., sorry for the OT, but it *is* interesting.
L. Ross Raszewski
<jessan...@mindspring.com.invalid> wrote:
>
>At some on-ramps (very VERY few that I have seen), they have these
>clever signs that say "Inner Loop" or "Outer Loop," so that the
>driver unfamiliar with the local landmarks can choose Inner or Outer
>(which will always be a constant, even as you go from facing north
>to east to south, you're still on the Inner Loop). Of course, the
>first time I saw that choice (inner/outer), it took me a full five
>seconds to figure out whether I wanted to be driving clockwise or
>counter-clockwise.
>
One of my favorite road signs is at the end of I-70 at I-695
Baltimore. Two arrows, one pointing lest, one pointing right. Above
each is the legend "I-695 to I-95"
Greg Ewing (using news.cis.dfn.de)
> I probably had some subconscious mental image of the map from
> relative mode, but to be honest, when I restarted in compass mode I
> didn't feel I had ANY idea where anything was. I basically re-explored
> the map, but this time I quickly got an idea of how it fit together and
> I won much more quickly.
I think perhaps a maze isn't the best setting for this experiment,
because mazes by their very nature tend to be difficult to learn
to find your way around. It's possible that a large part of the
reason for the above experience is simply because it's easier
to learn the layout of a maze by looking down on a map of it,
rather than wandering around inside it.
It would be interesting to repeat this using a more "normal"
setting, e.g. a house, whose layout is easy to memorise. Then
maybe you'd be able to measure the difficulty of the navigation
methods for getting around the map once you've learned it, as
opposed to learning the map in the first place.
--
Greg Ewing, Computer Science Dept,
University of Canterbury,
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg
Mike Roberts
> OKB (not okblacke) wrote:
> > I probably had some subconscious mental image of the
> > map from relative mode, but to be honest, when I restarted
> > in compass mode I didn't feel I had ANY idea where
> > anything was.
>
> because mazes by their very nature tend to be difficult to learn
> to find your way around. It's possible that a large part of the
> reason for the above experience is simply because it's easier
> to learn the layout of a maze by looking down on a map of it,
> rather than wandering around inside it.
>
> It would be interesting to repeat this using a more "normal"
> setting, e.g. a house, whose layout is easy to memorise.
Actually, the game's map isn't a maze, or at least isn't meant to be
(perhaps you've tried it and disagree, though). It's meant to be an
entirely typical adventure game setting - a "normal" sort of setting. I
specifically gave every room some significant landmark, and tried to make
the layout at least plausible as a real-life scenario. The layout is
non-mazelike in that it's pretty simple, is substantially open, and where
there are obstacles they're not just arbitrary walls but are meant to make
some kind of sense in the scenario.
Peter Killworth
is far from new... consider Jonathan Mestel's Cambridge game Xenophobia
from the late 70's... it's in the archive. (It's also an excellent
game!)
Peter Killworth
-----------------------------------------------------------------------
Prof. Peter D. Killworth, James Rennell Division for Ocean Circulation
and Climate, Southampton Oceanography Centre, Empress Dock, Southampton
SO14 3ZH, England.
Tel: +44 (0)23-80596202 Fax: +44 (0)23-80596204
Email: P.Kil...@soc.soton.ac.uk
Web: http://www.soc.soton.ac.uk/JRD/PROC/people/pki/pki.html
Editor, Ocean Modelling: http://www.elsevier.com/locate/omodol/
-----------------------------------------------------------------------
David Thornley
Ally <kitzapooR...@gmx.co.uk> wrote:
>
>I think a 'relative' mode of navigation gets easier when the map "comes
>with you" (as in a modern 3D game), but the IF geographic model is
>nothing like that in that you more or less 'blink' from one location (or
>scene) to the next.
>
>Using 'right' or 'ahead' actually feels unintuitive to me, not because
>it's something I wouldn't ever use in the real world, but because the
>direction you face doesn't feature in room descriptions, nor do you see
>the landscape move around you or have a clear idea as to where you're
>going (I'm not happy with these 'traits', though).
>
and that what you see on the screen depends on where you are and where
you're facing. I also assume that there is a certain amount of overlap
in the visual cues. At least, this is true of Myst and Riven, which
are the only games of this sort I've ever played. (The biggest problems
I had with Myst navigation are when I couldn't clearly perceive
the overlap.)
This means that such games are much more like real life in some ways,
but real-life navigation is not necessarily the best model for IF.
I believe this is what Mike's exploring.
FWIW, the last time I had to enter directions in "right", "left",
"forward", etc., in a text game, was in a comp game with a maze.
I read the spoilers for the maze, entered, got lost, and gave up
there, giving the game a score of 1.
--
David H. Thornley | If you want my opinion, ask.
da...@thornley.net | If you don't, flee.
http://www.thornley.net/~thornley/david/ | O-
Mike Roberts
> Haven't played the game, but the idea of non-compass relative
> directions is far from new...
In fact, that's part of my argument that compass directions are a better
natural fit for people than relative directions. If relative directions had
never been tried before in an actual game, it would lend credence to the
claim that compass-style navigation is purely an artifact of historical
inertia; but since several alternative styles have been tried, and really
only one stuck, I think it's more likely that the one that stuck is the one
that works better for more people.
> ... consider Jonathan Mestel's Cambridge game Xenophobia
> from the late 70's...
But does it let you switch between compass and relative directions
throughout the game, to compare their relative usability within a single
setting? (The single setting is important because it controls for other
unrelated factors that might affect ease of navigation.) That's the key
piece of the experiment here.
Michael Laurino
<mjrUND...@hotmail.com> wrote:
>"Peter Killworth" <P.Kil...@soc.soton.ac.uk> wrote:
>> ... consider Jonathan Mestel's Cambridge game Xenophobia
>> from the late 70's...
>
>But does it let you switch between compass and relative directions
>throughout the game, to compare their relative usability within a single
>setting? (The single setting is important because it controls for other
>unrelated factors that might affect ease of navigation.) That's the key
>piece of the experiment here.
>
Well, how about (good ol') "Mystery Mansion"? On the grounds of the
mansion, you can use compass directions from the start; inside the
mansion, only relative directions are available until you find the
compass. As long as you carry it, you can use both compass and
relative directions in all locations.
(Whenever I restarted that game, the first order of business was to
find the compass so I wouldn't have to keep using the relative
directions -- which relates to the original question, I guess.)
Gene Wirchenko
>Mike Roberts wrote:
>
>> If you feel like trying it out, I'd be interested in hearing about your
>> experiences with it.
>
>So, I did the experiment. Without any doubt, COMPASS mode feels more
>user-friendly to me than RELATIVE mode, but I found that RELATIVE remains
>reasonably playable after some getting used to it.
[snip]
>So, it looks like my visual representation of the world was actually determined
>by the navigation mode I was using, and that I wasn't unconditionally using the
>simplest (whatever that means) of the two models. However, this conclusion
>rests on the assumption that the visual representation *is* the model, which I'm
>not sure is a valid assumption.
What about when the angles are not 90 degrees? If I have a set
of instructions internalised for how to get to a place, I may think of
turns in terms of *roughly* so many degrees to the left or right.
After following a chain, I may not know which way I am facing
according to the chain. (Think of accumulated error.)
When I was in China, the university that I worked in was divided
into two parts by one road. The part were referred to as the east
side and the west side. These were not the actual directions though.
They were more like northeast and southwest, except that they were
more like 30 degrees east of north and 30 degrees west of south.
Using north and south would have been more accurate!
When I went for walks, I did not use compass directions because
there was no assurance that I would be in a NSEW grid. Instead, I
kept track of turns and lengths. If I got lost, I would backtrack.
If I found a location that I already knew, I updated my set of
connections.
.
Sincerely,
Gene Wirchenko
Computerese Irregular Verb Conjugation:
I have preferences.
You have biases.
He/She has prejudices.