I've seen some discussions of the 3fold founder on interpreting what the
triangle means. It seems to me to be easier to understand in terms of a
"projection" of a 3-dimensional space.
Imagine 3 orthogonal axes labeled with your 3 favourite terms for the 3fold
concepts (I'll use world, story, game). The points of interest are in the
range 0..1 on each axis. For a given GM decision (about a resolution point,
since that's apparently where the 3fold is to be used), a "1" for a given axis
X means "the decision was made according to X-based principles", and a 0 means
"the decision contradicted X-based principles". A decision made according to
"world" principles that contradicts story and game principles would thus be
(w,s,g) = (1,0,0).
One could characterize a given GM's style by taking an average of all the
resolution points s/he ever decides. Thus 30% world, 50% drama, 20% game
would be (.3, .5, .2).
The threefold then represents a projection of "style" points onto the plane
defined by w+s+g=1.0. I think this plane can be thought of as pretending each
decision is made purely according to one viewpoint, so that all "averages"
must have w+s+g = 1.0 (any increase in "story" principles requires a decrease
in one of the other areas).
However, it's not necessary to presume the three are mutually exclusive.
A decision made according to principles common to all three views would be
(1,1,1) for example. This point projects to the centre of the usual
triangle. As does, unfortunately, (0,0,0) -- the impossibly chaotic campaign
where all decisions contradict all 3 sets of principles.
The illumination I received from the 3D viewpoint is this: perhaps a lot of
3fold discussions have bogged down because the 2d triangle projection throws
away some information available in the 3d version -- people talk past each
other because their shared communication model -- the triangle -- is sometimes
inadequate to represent their different points of view.
The 3d version is harder to visualize, but contains more information, since it
allows for representing decisions consistent with more than one of the "axes".
--
"Yo' ideas need to be thinked befo' they are say'd" - Ian Lamb, age 3.5
http://www.cs.queensu.ca/~dalamb/
> Imagine 3 orthogonal axes labeled with your 3 favourite terms for the
3fold
> concepts (I'll use world, story, game). The points of interest are in the
> range 0..1 on each axis. For a given GM decision (about a resolution
point,
> since that's apparently where the 3fold is to be used), a "1" for a given
axis
> X means "the decision was made according to X-based principles", and a 0
means
> "the decision contradicted X-based principles". A decision made according
to
> "world" principles that contradicts story and game principles would thus
be
> (w,s,g) = (1,0,0).
Would a "0" necessarily mean that that leg of the model was *contradicted*
or just not considered?
>
> One could characterize a given GM's style by taking an average of all the
> resolution points s/he ever decides. Thus 30% world, 50% drama, 20% game
> would be (.3, .5, .2).
This is exactly the same revision I mentioned, without the mentioning of the
third dimension.
http://x71.deja.com/threadmsg_ct.xp?AN=697070473.1&mhitnum=27&CONTEXT=977847
604.76939280
>
> The threefold then represents a projection of "style" points onto the
plane
> defined by w+s+g=1.0. I think this plane can be thought of as pretending
each
> decision is made purely according to one viewpoint, so that all "averages"
> must have w+s+g = 1.0 (any increase in "story" principles requires a
decrease
> in one of the other areas).
>
> However, it's not necessary to presume the three are mutually exclusive.
> A decision made according to principles common to all three views would be
> (1,1,1) for example. This point projects to the centre of the usual
> triangle. As does, unfortunately, (0,0,0) -- the impossibly chaotic
campaign
> where all decisions contradict all 3 sets of principles.
Both of these two previous paragraphs seem mutually exclusive. A "1" should
represent 100% as you stated in the first paragraph. A decision made
according to all three views couldn't be 100%,100%,100% it would be,
instead, 33.3%,33.3%,33.3%.
And the 3rd dimension isn't needed, btw. As long as you're talking about
percentages all you need is a triangle and a point somewhere *within* that
triangle (unless one or two of the views of the model is completely ignored;
in which case the point will appear on an edge or vertex of the triangle,
respectively). The 3rd dimension would be needed if you used a 4-fold model,
though, since you'd be in need of another vertex to represent a 4th view.
Not necessarily. I think he was saying that a decision made which would be
the same decision prescribed by all three viewpoints (a decision that is
harmonized with all three) would be 1;1;1.
I wonder what would qualify in that respect?
Angela
No .sig
No .damn copies of my missing volumes of "Story of Civilization" for
Christmas, either
> Not necessarily. I think he was saying that a decision made which would
be
> the same decision prescribed by all three viewpoints (a decision that is
> harmonized with all three) would be 1;1;1.
I know that's what he was saying but I think that .33, .33, .33 is the
correct way to write it.
Yeah, that's what I was thinking initially, but are there any we can think
of that would be completely in accordance with all three views? I mean, are
there any truly common ideas held, such that a decision made in accordance
with that idea would get a 1 no matter what axis you plotted it on?
Angela
No .sig
No .breakfast yet
You're right - 0 should mean "ignores". I suppose (1,-1,-1) might make sense
for the example above, but I haven't thought through the implications of
allowing negative coordinates.
>> However, it's not necessary to presume the three are mutually exclusive.
>> A decision made according to principles common to all three views would be
>> (1,1,1) for example. This point projects to the centre of the usual
>> triangle. As does, unfortunately, (0,0,0) -- the impossibly chaotic
>campaign
>> where all decisions contradict all 3 sets of principles.
>
>Both of these two previous paragraphs seem mutually exclusive. A "1" should
>represent 100% as you stated in the first paragraph. A decision made
>according to all three views couldn't be 100%,100%,100% it would be,
>instead, 33.3%,33.3%,33.3%.
>
>And the 3rd dimension isn't needed, btw. As long as you're talking about
>percentages all you need is a triangle and a point somewhere *within* that
>triangle (unless one or two of the views of the model is completely ignored;
>in which case the point will appear on an edge or vertex of the triangle,
>respectively). The 3rd dimension would be needed if you used a 4-fold model,
>though, since you'd be in need of another vertex to represent a 4th view.
>
This issue is the entire point behind the 3d idea -- my assertion that 2d
loses information that leads to arguments where people talk past each other.
Every point on the 3D line from (0,0,0) to (1,1,1) gets mapped to the middle
of the triangle, but they mean completely different things. Ignoring negative
coordinates for the moment (the "contradiction" idea) (0,0,0) would mean a
decision made where all 3 concerns were irrelevant, while (1,1,1) would mean
consistent with all 3.
The point was: the triangle midpoint (and, by comparison, most other points on
the triangle) conflates several different things, leading to confused
discussions.
Furthermore, if we can ensure that negative coordinates make sense, then the
"pure" simulationist extreme is a requirement that no decision ever go
negative on the "world" axis. Thus a decision made according to some "drama"
rule that happened to be irrelevant to world-based concerns (e.g. some choice
about which events to look at in detail, where all events were proper
"simulation" events), would be allowed.
My feeling is that playing styles are best described as a positive
vector in an N-space, where N is the number of styles in the model. For
a "perfect" model that encompasses all possible styles, the length of
the vector is exactly 1 (ie., a unit vector), but for an "incomplete"
model like the threefold, it will be between zero and one.
For the threefold, N is 3, and the "style space" is the positive octant
of a unit sphere. With three axes, S D G, you can graph threefold
preferences with the following:
0 <= S <= 1
0 <= D <= 1
0 <= G <= 1
S^2 + D^2 + G^2 <= 1
This shows that a very strong preference along one axis implies that
other styles must have less consideration, but moderate preferences
limit the other styles very little. This seems to fit with the evidence
(that strong simulationism must ignore the metagame, and that strong
gamism precludes most story and world concerns, but that mid-triangle
people *don't* recognize that there are tradeoffs).
Sorry about the math-speak, but I'm not sure how to describe this in a
way that doesn't use it.
--
Bradd W. Szonye Work: br...@cup.hp.com
Software Design Engineer Home: bra...@concentric.net
Hewlett-Packard Cupertino Site, iFL Phone: 408-447-4832
> My feeling is that playing styles are best described as a positive
> vector in an N-space, where N is the number of styles in the model. For
> a "perfect" model that encompasses all possible styles, the length of
> the vector is exactly 1 (ie., a unit vector), but for an "incomplete"
> model like the threefold, it will be between zero and one.
I understand the math-speak, but I'm just laughing my head off that people
think that this somehow makes talking about gaming /simpler/ and /more
communicative/. For god's sake. :D
> For the threefold, N is 3, and the "style space" is the positive octant
> of a unit sphere. With three axes, S D G, you can graph threefold
> preferences with the following:
> 0 <= S <= 1
> 0 <= D <= 1
> 0 <= G <= 1
> S^2 + D^2 + G^2 <= 1
I mean, just look at this. LOOK AT IT!
:)
--
***************************************************************************
"I was pleased to be able to answer promptly, and I did. I said I didn't
know."----- Mark Twain, _Life on the Mississippi_
Jason Corley | le...@aeonsociety.org | ICQ 41199011
> You're right - 0 should mean "ignores". I suppose (1,-1,-1) might make
sense
> for the example above, but I haven't thought through the implications of
> allowing negative coordinates.
The implications would be that you wouldn't be able to use any pretty shape
to visualize - you would be forced to use the whole cartesian graph (x,y,z).
Well, at least until 1 unit away from (0,0,0).
> This issue is the entire point behind the 3d idea -- my assertion that 2d
> loses information that leads to arguments where people talk past each
other.
> Every point on the 3D line from (0,0,0) to (1,1,1) gets mapped to the
middle
> of the triangle, but they mean completely different things. Ignoring
negative
> coordinates for the moment (the "contradiction" idea) (0,0,0) would mean a
> decision made where all 3 concerns were irrelevant, while (1,1,1) would
mean
> consistent with all 3.
OK, I see what you mean. Unfortunately it would be difficult to describe
percentages then. For example, if your (w,s,g) equalled (.4, .4,.6), you
couldn't describe it as 40% world-centric.
>
> The point was: the triangle midpoint (and, by comparison, most other
points on
> the triangle) conflates several different things, leading to confused
> discussions.
Agreed.
> Furthermore, if we can ensure that negative coordinates make sense, then
the
> "pure" simulationist extreme is a requirement that no decision ever go
> negative on the "world" axis.
That's possible, the shape would be a cube two units on a side with the zero
point dead center. Each two opposite faces (the 1 and the 6 or the 2 and the
5 or the 3 and the 4 on a six-sided die) represent "completely contradicts
with ..." and "completely agrees with ...". The vertex where the 1, 2, and 3
meet could represent "completely contradicts all three principles"
(the -1,-1,-1 point) while the opposite vertex (meeting point of 4, 5, and 6
(the 1,1,1 point on the graph)) represents "completely agrees with all three
principles".
I hope this helps.
Bradd W. Szonye <bra...@concentric.net> wrote:
>David Alex Lamb <dal...@qucis.queensu.ca> wrote:
>> I've seen some discussions of the 3fold founder on interpreting what
>> the triangle means. It seems to me to be easier to understand in terms
>> of a "projection" of a 3-dimensional space.
>
>My feeling is that playing styles are best described as a positive
>vector in an N-space, where N is the number of styles in the model.
Frankly, this and others of the recent spate of threads has
moved me pretty solidly over to the camp that rgfa is a forum for
esoteric discussion of little to no practical value. While I still
consider the threefold a decent kernel of an idea, I really wish
discussion of it would go away.
If the threefold is of any use at all, then it should come up
and be useful in "regular" discussion of actual game situations and
problems.
On the other hand, I realize that if I want discussion to
go a different way, I should start new threads. I am currently just
getting back from holidays -- I will try to think of some new threads
to start soon.
--
John H. Kim | Whatever else is true you
jh...@fnal.gov | Trust your little finger
www.ps.uci.edu/~jhkim | Just a single little finger can
UC Irvine, Cal, USA | Save the world. - Steven Sondheim, "Assassins"
> This issue is the entire point behind the 3d idea -- my assertion that 2d
> loses information that leads to arguments where people talk past each other.
> Every point on the 3D line from (0,0,0) to (1,1,1) gets mapped to the middle
> of the triangle, but they mean completely different things. Ignoring negative
> coordinates for the moment (the "contradiction" idea) (0,0,0) would mean a
> decision made where all 3 concerns were irrelevant, while (1,1,1) would mean
> consistent with all 3.
*I* don't like this idea (I am for the 2D geometry only).
If a decision is coherent with all 3 directions, then:
**you can't tell where is the decision in the triangle**
because the whole triangle is only a vertex in this case.
(the dramatic vertex, the world vertex and the metegame
vertex can not be distinguished in this situation,
*SO* they are in the same location in the space of decisions)
Eric Lestrade
> Frankly, this and others of the recent spate of threads has
> moved me pretty solidly over to the camp that rgfa is a forum for
> esoteric discussion of little to no practical value. While I still
> consider the threefold a decent kernel of an idea, I really wish
> discussion of it would go away.
I agree. Compare the last month or two of threefold arguments to some of the
discussions we had for months before, and the waste of time becomes almost
sickening.
I'd rather remove the threefold from the FAQ completely instead of going
through periods like this. It does have the positive point of identifying
people one is better off not dealing with in the first place (as they launch
old and flame baited attacks on it), but most here don't use that advantage.
--
Brian Gleichman
glei...@mindspring.com
Age of Heroes: http://gleichman.home.mindspring.com/
Free RPG Reviews: http://gleichman.home.mindspring.com/Reviews.htm
>"John Kim" <jh...@cosmic.ps.uci.edu> wrote in message
>news:92e196$ass$1...@news.service.uci.edu...
>
>
>> Frankly, this and others of the recent spate of threads has
>> moved me pretty solidly over to the camp that rgfa is a forum for
>> esoteric discussion of little to no practical value. While I still
>> consider the threefold a decent kernel of an idea, I really wish
>> discussion of it would go away.
>
>I agree. Compare the last month or two of threefold arguments to some of the
>discussions we had for months before, and the waste of time becomes almost
>sickening.
>
Given that both of these guys KFed me, someone else can answer. If
this place isn't for discussions about things like the threefold, what
is it here for.
If all it's here for is telling of old "war stories", then personally,
I prefer to masturbate in private.
Alain
"Brian Gleichman" <glei...@mindspring.com> wrote in message
news:92gvck$4sg$1...@slb7.atl.mindspring.net...
> Hey, chill out guys! If you don't like a thread, just ignore it.
> I personally also find these theory discussions to be rather
> uninteresting. I just ignore them.
All nice and well, but that only solves part of the problem.
Many people here have limited time available for posting. If they're
spend all or even most of it defending or attacking the threefold like
they've done over the last couple of months there is no one time left
for other threads.
Brian
Sent via Deja.com
http://www.deja.com/
What a fine contribution _you_ seem to be making to the spirit of the
group. Your maturity is inspiring to us all.
Biff
--
-------------------------------------------------------------------
"Me? Lady, I'm your worst nightmare - a pumpkin with a gun.
[...] Euminides this! " - Mervyn, the Sandman #66
-------------------------------------------------------------------
>Robert Scott Clark wrote:
>>
>> If this place isn't for discussions about things like the threefold,
>> what is it here for.
>>
>> If all it's here for is telling of old "war stories", then personally,
>> I prefer to masturbate in private.
>
> What a fine contribution _you_ seem to be making to the spirit of the
>group. Your maturity is inspiring to us all.
>
I have taken part in the threefold discussion, and I have not butted
in to those discussions that I care nothing about. Now a couple of
asses (sure I'll insult them, it's not like they can see it) come
along and complain about the only thread going on around here that I
give a damn about. Pardon me if I get a little defensive.
I'll give you the chance too... what should we be talking about here?
> > My feeling is that playing styles are best described as a positive
> > vector in an N-space, where N is the number of styles in the model. For
> > a "perfect" model that encompasses all possible styles, the length of
> > the vector is exactly 1 (ie., a unit vector), but for an "incomplete"
> > model like the threefold, it will be between zero and one.
>
> I understand the math-speak, but I'm just laughing my head off that people
> think that this somehow makes talking about gaming /simpler/ and /more
> communicative/. For god's sake. :D
which is why i prefer to ditch the math, push the limits out way past
anything found in the real world, and eliminate any (absolute)
mathematical relationship between the axes. it becomes just a labeling
system. i notice there aren't endless flamefests about the campaign axes
around here, and i suspect that's in part because they don't use a
mathematical/geometric model that implies either a zero-sum rating, or a
complete set of categories.
--
woodelf <*>
woo...@rpg.net
http://members.home.net/woodelph/
I did not realize that similarity was required for the exercise of
compassion. --Delenn
> OK, I see what you mean. Unfortunately it would be difficult to describe
> percentages then. For example, if your (w,s,g) equalled (.4, .4,.6), you
> couldn't describe it as 40% world-centric.
no, but you can say "about equally concerned with world and story, and a
bit more concerned with game; not obsessed over or disinterested in any of
the elements". now, given that there is no objective rating system, nor
any way to apply it, it seems to me that that's as accurate as we can get
in the first place. after all, who's to say that your .4 isn't my .8.
i'm equally distrustful of percentages in this context--i don't think most
of us could accurately determine the relative importance of the 3-fold
axes (or any other set of criteria) to a decision beyond the fineness of
the Fudge scale.
well, i've been advocating roughly such a visualization model ever since i
discovered r.g.f.advocacy a few years ago. the differences are slight,
but possibly significant. i prefer to think of it as a 3-space with each
axis corresponding to one of the "points" of the 3-fold. there we agree.
however, i've always envisioned it as a 0 meaning no regard (positive or
negative) for the style in question--it hadn't occurred to me that anyone
would make a decision deliberately contrary to any of the styles in
question [i'll get back to that in a bit]. anyway, i think of the space
being the triply-positive octant of a sphere, because i find the sphere
surface useful for the model. the surface of the sphere is the
limit--*if* you are making a decision that is so extreme in one or more
ways as to be sitting on the surface of the sphere, then (and only then)
you can not alter one value without altering at least one of the others.
however, i envision this space as huge and the surface distant because,
IME, it is rare that a decision point could not be altered to better
comply with one of the styles of the 3-fold while not affecting its degree
of compliance with the other 2. if we can successfully rate the
game-ness, world-ness, or story-ness of most (99+%) GM's decisions on a
scale of 0 to 5, then the sphere surface is out around 10.
i prefer not to use the triangle at all, because in a triangle you can't
alter your distance from any vertex without altering your distance from at
least one of the other vertices. to me, this is contrary to the actual
experience roleplaying. IME, you can often wiggle your, say, story-ness,
by a substantial bit, without helping or hindering the game-ness or the
world-ness measurably (if at all). i think a further advantage of the 3D
model is that there's more room to play, geometrically. a given rules
system or GM could easily define some sort of space or surface within the
larger 3-fold space. due to the GM's preferences, or the nature of the
system, all decisions would be limited to that space or surface--and the
outer limits of that space could be much closer. so, (still using a
standard range of 0-5 and absolute of 0-10), perhaps Burros & Banditos has
a very detailed setting that, for whatever reason, is noticably at odds
with its mechanics over the impact of magic on the setting. you might be
in a situation where, as written, you can't exceed a world-ness of 2
before it starts to inpinge on your game-ness, and vice versa.
now, as to those who find this sort of theory uninteresting: if i
understand correctly, you find it uninteresting because it's
non-productive and/or not applicable in a practical way to RPing. i
suggest that a visual model more like i have above, rather than a
triangle, would sidestep most of the endless arguments over the 3-fold.
IME on this newsgroup, they mostly boil down to 2 sorts: disagreements
about to what degree one style trades off with another, and disagreements
about the validity of the axes chosen. this model explicitly visualizes
the styles as only trading off as much as you want them too--i.e., the
absolute limits of playability are way out there where they are not a
concern to the real world, but a particular situation could easily impose
tighter limits. and by trying to be less of a restrictive mapping of
preferences (the triangle implies that you select the values--and thus
your preferences--in some sort of mathematical way, and are thus limited;
this implies that you may categorize yourself as you see fit in all the
values--5,2,0; 1,0,0; or even 5,5,5), i hope it would lessen people's
feelings of being left out by the model, and thus fighting it.
furthermore, it accepts the addition of as many axes as you care to name,
without altering the model's functionality. all axes are considered,
mathematically, to be orthogonal. if, for you, there is a relationship
between 2 or more of them, that's fine, but there is no need to alter the
entire model to reflect this (which seems to be another source of
argument). basically, i've always considered the 3-fold a labeling tool,
not a categorizing tool. it adds useful (to me) terminology, with (in
theory) agreed-upon definitions, and thus cuts down on a lot of
miscommunication. it's when it starts trying to map real-world decisions
to a zero-sum model that i think it falls down. remove the zero-sum, and
i think it will work better.
for those who think it's got the wrong axes: perhaps to be useful, it
needs to be expanded to look more like the campaign axes. everyone can
throw in their favorite axis, and we'll see which ones strike a chord with
more than one or 2 people, and prove to be useful for discussion. for
those who think models are useless theory: sorry, can't help; they help my
understanding, so i see them as useful. it's like the impasse i had with
me intro chem prof: he said that one only learns theory by working the
math repeatedly, whereas i only learn theory by being taught theory.
----
your comment about making decisions deliberately contrary to a given style
got me thinking: does anybody do this? that is, does anybody say "these
are the precepts and values of style X, and in my game i strive to do just
the opposite"? as opposed to "these are the precepts and values of style
X, which conflict with those of style Y, and in my game i strive to uphold
style Y". and as opposed to "i strive to make my game like foo", which
is exactly opposite to style X, unbeknownst to the speaker. if so, i
guess i'd be inclined to use negative space for that. which would be a
further advantage to a 3D model, IMHO: you could make a distinction
between playing more like X and playing less like Y, whereas the triangle
just tells you you moved further away from Y and closer to X, without
telling you why.
I agree with you, but there are two counter arguments to this. I'm
going to roughly state what some people have said to me, so if I
misrepresent your position, sorry.
1. Warren wouldn't necessisarily agree with your "wiggle" room idea.
He has said that the world produces a result, and any deviation from
this result is a "trade-off".
I have responded to this with the idea that the more detailed (and
strict) your model is (this works for world model, story model, or set
of game rules), the less "wiggle" room you have, and consequently the
more the model represents a trade-off. Conversly, the less detailed
the model is, the more possible/acceptable results there are, the more
wiggle room there is.
2. Justin Bacon has said that the threefold is about conflicting
decisions, and does not apply to decisions where the GM doesn't have
to choose between ways of making the decision.
>----
>your comment about making decisions deliberately contrary to a given style
>got me thinking: does anybody do this? that is, does anybody say "these
>are the precepts and values of style X, and in my game i strive to do just
>the opposite"? as opposed to "these are the precepts and values of style
>X, which conflict with those of style Y, and in my game i strive to uphold
>style Y". and as opposed to "i strive to make my game like foo", which
>is exactly opposite to style X, unbeknownst to the speaker.
As one of the people who has suggested possible negative values, I can
say that the first of these is definately not what I was talking about
(and I don't really get what you are trying to say with the second
one)
Let me use the example of Gamism, since that is where I would be most
on the negative axis. Given your example, a +G player would want more
highly skilled players to succeed, and a -G player would want the
worst player to succeed. On this scale, I would place myself very
close to 0. By this scale, I find both +G and -G undesirable. What I
was attempting to do by labling myself negative on the game scale was
to differentiate my style (where I actively want player skill to have
no effect - even at the cost of a better story, simulation, whatever)
from the other style where the player wants +G, but must sacrifice it
to attain some other goal.
Let's take two decisions that look like (rated from 0 to 1)
D/G/S
(0.75,0.2,0)
and
(0.75,0,0)
When I say anti-gamist (-g) what I am refering to is the person who
picks the second choice.
> >i prefer not to use the triangle at all, because in a triangle you can't
> >alter your distance from any vertex without altering your distance from at
> >least one of the other vertices. to me, this is contrary to the actual
> >experience roleplaying. IME, you can often wiggle your, say, story-ness,
> >by a substantial bit, without helping or hindering the game-ness or the
> >world-ness measurably (if at all).
>
> I agree with you, but there are two counter arguments to this. I'm
> going to roughly state what some people have said to me, so if I
> misrepresent your position, sorry.
>
> 1. Warren wouldn't necessisarily agree with your "wiggle" room idea.
> He has said that the world produces a result, and any deviation from
> this result is a "trade-off".
>
> 2. Justin Bacon has said that the threefold is about conflicting
> decisions, and does not apply to decisions where the GM doesn't have
> to choose between ways of making the decision.
it is exactly these two ideas that i'm trying to soften by re-modeling the
3-fold, as it is exactly these ideas that i think lead to most of the
arguments over the validity of the 3-fold, and prevent us from using it as
a better communicative tool. i would guess that case one is someone who
is essentially on the surface of the space, out at teh limits, where you
must trade one value for another--but i maintain that this is an unusual
position, and probably one that only simulationists often even approach
(based on the fact that i have rarely, if ever, heard someone who
considered themselves strongly gamist or dramatist claim that any
deviation from the ideal would of necessity be a lessening). case two
sounds like someone's personal decision space, within the larger 3-fold
space. IME, most GMs seem to play at the limit of the personal space in
3-fold terms. for that GM, any change in the degree of one quality of the
3-fold affects the others. but a different GM might be able to start from
exactly the same position, and alter one of the values without altering
the others at all, due to different precepts of how the game should be
run.
> >----
> >your comment about making decisions deliberately contrary to a given style
> >got me thinking: does anybody do this? that is, does anybody say "these
> >are the precepts and values of style X, and in my game i strive to do just
> >the opposite"? as opposed to "these are the precepts and values of style
> >X, which conflict with those of style Y, and in my game i strive to uphold
> >style Y". and as opposed to "i strive to make my game like foo", which
> >is exactly opposite to style X, unbeknownst to the speaker.
>
> As one of the people who has suggested possible negative values, I can
> say that the first of these is definately not what I was talking about
> (and I don't really get what you are trying to say with the second
> one)
ok, since math seems to be the preferred means of getting these ideas across:
we start with a scale from 0 to 1 for each axis, with 0 meaning no
consideration for the style in question, and 1 being very strong
consideration. now let's further assume that we have 3 people who hcan
all be described as (-1,.5,.5). in that case, quote one would refer to
somebody who has the negative one precisely because they detest the style
of that axis for some reason, to the point where they go out of their way
to work against its precepts; taking away this dislike, without changing
anything else, would leave them at (0,.5,.5). quote two is somebody who
simply likes the aother two axes better, and in order to get them up to a
score of .5 each she's had to sacrific first-axis-ness, bringing it down
to -1. the third quote is from someone who's never heard of the axes.
she would describe her own game as being a +1 vector along a single axis,
which translates into the 3-fold space as (-1,?,?).
remove the zero-sum, and i think it will work better.
It may work better for you, but it would work worse for me.
I would also note that the 'zero sum' aspect is strictly a representational
issue; you can stick two axes on the triangle and make it non zero sum, and you
can add a label to the origin and make your positive octant of a sphere zero
sum.
your comment about making decisions deliberately contrary
to a given style got me thinking: does anybody do this?
It's not clear to me how this would even be possible. In the only example I
can think of, if one specifically tries to avoid making a good story, one ends
up being very concerned with story. An existentialist story is still a story.
Warren J. Dew
Powderhouse Software
> Regarding the threefold, 'woodelf' posts, in part:
>
> remove the zero-sum, and i think it will work better.
>
> It may work better for you, but it would work worse for me.
which is why i'm suggesting that for many (most? all?) GMs, a personal
zero-sum surface/space would be used, which falls somewhere within the
octant that describes all possible decisions. i'm trying to find a way
that the model can support both zero-sum and non-zero-sum interpretations,
in such a way that it still is a useful labeling and communicative device.
> I would also note that the 'zero sum' aspect is strictly a representational
> issue; you can stick two axes on the triangle and make it non zero sum,
and you
> can add a label to the origin and make your positive octant of a sphere zero
> sum.
only if all decisions take place on the surface of the sphere (not in its
interior) [unless i misunderstand what you mean].
> your comment about making decisions deliberately contrary
> to a given style got me thinking: does anybody do this?
>
> It's not clear to me how this would even be possible. In the only example I
> can think of, if one specifically tries to avoid making a good story, one ends
> up being very concerned with story. An existentialist story is still a story.
which is sorta why i'm wondering if anybody ever does this. personally, i
think unconcerned is as slight as you can get, and if you go beyond that
you're once again raising your "score" in that axis (just as you say).
> (based on the fact that i have rarely, if ever, heard someone who
> considered themselves strongly gamist or dramatist claim that any
> deviation from the ideal would of necessity be a lessening).
As a someone considers themselves strongly gamist, I'd make that claim.
> > (based on the fact that i have rarely, if ever, heard someone who
> > considered themselves strongly gamist or dramatist claim that any
> > deviation from the ideal would of necessity be a lessening).
>
> As a someone considers themselves strongly gamist, I'd make that claim.
so, you're saying that, given two ways to resolve a decision, one of which
has (story,world,game) values of (0,0,.8) and the other of which has
(.1,0,.8), you would find the latter decision less gamist because of the
slight added consideration of story, even though the amount of game
consideration hadn't been reduced? that is, even if decisions would not
compromise the game-ness, and there was a reason for them (one of your
players would be happier, say), you would not make them?
> that is, even if decisions would not
> compromise the game-ness, and there was a reason for them (one of your
> players would be happier, say), you would not make them?
The example is flawed.
Any time that I consider story under conditions that require a gamist
decision, I must compromise game-ness.
>so, you're saying that, given two ways to resolve a decision, one of which
>has (story,world,game) values of (0,0,.8) and the other of which has
>(.1,0,.8), you would find the latter decision less gamist because of the
>slight added consideration of story, even though the amount of game
>consideration hadn't been reduced? that is, even if decisions would not
>compromise the game-ness, and there was a reason for them (one of your
>players would be happier, say), you would not make them?
I cannot speak for Brian, of course, but this makes no sense to me.
The Threefold doesn't deal with outcomes, and breaks immediately
when you ask it to do so: it deals with intentions.
A gamist player/GM can be just as concerned with purity of intentions as
a simulationist player/GM. For example, if we have agreed that a
situation will be resolved solely on the basis of player skill, rules,
randomizers, and the established situation, then I'm violating that
agreement if I allow story to influence the outcome. It's still a
violation even if the outcome is one that could have been produced
by skill, rules, randomizers and situation, because in fact it *wasn't*.
If someone bribes me to throw a game of chess, that game is ruined as
a sporting contest. It's still ruined even if having me lose was the
expected outcome of a real, unthrown game (say, the opponent was
much better than me)--because the outcome depended on money instead of
skill. I didn't have the appropriate intent.
To me your reasoning sounds like saying "It really doesn't matter about
the bribery, because the outcome was the same as it probably would have
been anyway." But most chessplayers would agree that it *does* matter
about the bribery; and gamist roleplayers tend to think it does matter
when you manipulate outcomes for story purposes.
It's worth noting that the story manipulation will change the results
away from what game considerations would produce; otherwise why would
you bother doing it? Manipulation that never changes the outcome
is no manipulation at all.
I don't see the Threefold as symmetrical on this point; I have never
met a dramatist player who had a class of resolution methods he was
disbarred from using, whereas both gamist and simulationist players
generally do. We've talked a lot about what simulationist players are
disbarred from. For a specific example of what a gamist player is
disbarred from: I am accustomed to say to my (fairly simulationist)
GM "I have no idea what to do here, but my PC would: can we fudge
it?" I could never say that in a serious chess game, and I don't think I
could say it in a serious gamist RPG: it misses the whole point of
player skill as a determining factor. But it makes sense in a sim
game, and is not necessarily disbarred there.
It would be tidier to have a symmetrical figure, but that's not what
we have. I think, therefore, that the whole idea of mapping the thing
in spaces of whatever dimension is useless. The points are not
interchangable; we can't come to a conclusion "Being X means you
don't want Y and Z" that is symmetrically true across the triangle.
It's just not what roleplayers seem to be like; the dramatist point
is very fundamentally different from the other two.
The only use of thinking of it as a triangle, as far as I can see, is
to express the point that if you are closer to one thing you are
further from the others.
Mary Kuhner mkku...@eskimo.com
Then it has even less of a requirement to be a trade-off model. You
can have the intent to maximize everything, even if you know you will
not succeed.
If this is not what you are saying, please explain.
> the dramatist point
>is very fundamentally different from the other two.
>
Well here is something I can absolutely agree with.
>The only use of thinking of it as a triangle, as far as I can see, is
>to express the point that if you are closer to one thing you are
>further from the others.
In a general sense, I will agree, but not specifically. If I looked
at 10000 games, and each one increased dramatist concerns, then I
would expect a trend in lowering in the others, BUT, for any one of
these games, I think it is entirely possible that the story could be
increased without harming anything else.
> > that is, even if decisions would not
> > compromise the game-ness, and there was a reason for them (one of your
> > players would be happier, say), you would not make them?
>
> The example is flawed.
perhaps. the more i think about it, the more i think it's a logical
abstraction that couldn't occur in actuality.
> Any time that I consider story under conditions that require a gamist
> decision, I must compromise game-ness.
but that pretty much answers my question, nonetheless: to the degree that
my example situation can be approximated in reality, the very act of
considering non-game elements lessens game-ness for you (as a
representative gamist). unless i misread you.
part of this stems from unclear writing on my part: i *was* talking about
intentions, not outcomes. or at least i intended to be. so, the question
still stands: is a decision less gamist because non-gamist elements were
included in the decision, even if the degree to which gamist elements were
included was not at all diminished? [i'll grant you that, while i can
state this as a mathematical construct, i can't actually come up with an
example of it. nonetheless, i would love to hear how the gamist feels in
this hypothetical situation, even if it's one that could never come up in
reality.]
> I don't see the Threefold as symmetrical on this point; I have never
> met a dramatist player who had a class of resolution methods he was
> disbarred from using, whereas both gamist and simulationist players
> generally do. We've talked a lot about what simulationist players are
> disbarred from. For a specific example of what a gamist player is
> disbarred from: I am accustomed to say to my (fairly simulationist)
> GM "I have no idea what to do here, but my PC would: can we fudge
> it?" I could never say that in a serious chess game, and I don't think I
> could say it in a serious gamist RPG: it misses the whole point of
> player skill as a determining factor. But it makes sense in a sim
> game, and is not necessarily disbarred there.
>
> It would be tidier to have a symmetrical figure, but that's not what
> we have. I think, therefore, that the whole idea of mapping the thing
> in spaces of whatever dimension is useless. The points are not
> interchangable; we can't come to a conclusion "Being X means you
> don't want Y and Z" that is symmetrically true across the triangle.
> It's just not what roleplayers seem to be like; the dramatist point
> is very fundamentally different from the other two.
which is precisely why i propose model where the axes are all orthogonal:
it enables you to represent eitehr the trade-off model or the
rate-each-one-separately model, and since it doesn't favor either
interpretation, it (IMHO) makes it easier to discuss which of those is the
case. frex, your point about non-symmetricity is very interesting to me.
i'll be thinking about that one all the way through tomorrow's game
session. and i think part of why i never thought about it before was the
triangle that implicitly puts all 3 axes equally in opposition.
> The only use of thinking of it as a triangle, as far as I can see, is
> to express the point that if you are closer to one thing you are
> further from the others.
which is the part i'm contesting. so far, i'm not sure my argument is
sound, but i have this gut feeling that you are *not* inherently further
from the others as you get closer to one. i'll keep triyng to track this
down, or better articulate it, so that i or another can ascertain whether
i'm full of it. ;-)
> but that pretty much answers my question, nonetheless: to the degree that
> my example situation can be approximated in reality, the very act of
> considering non-game elements lessens game-ness for you (as a
> representative gamist). unless i misread you.
Although I have a more complex view of gamist than the FAQ, this is true for
those cases where there's overlap.
I don't see the Threefold as symmetrical on this point; I
have never met a dramatist player who had a class of
resolution methods he was disbarred from using, whereas
both gamist and simulationist players generally do.
I got the impression that David Berkman felt that any methods outside of what
he recommended were, at a minimum, 'bad gamesmastering'.
Alain avoids using dice for his own campaigns.
It's not clear to me that this is a threefold concern, but it's not clear that
it isn't, either. I do get the sense from some story advocates that "a really
good gamesmaster/storyteller can always come up with a better story than
soulless randomizers or mechanics", which implies that there are some
tradeoffs, even if they aren't with the other vertices of the triangle.
> It's not clear to me that this is a threefold concern, but it's not clear that
> it isn't, either. I do get the sense from some story advocates that "a really
> good gamesmaster/storyteller can always come up with a better story than
> soulless randomizers or mechanics", which implies that there are some
> tradeoffs, even if they aren't with the other vertices of the triangle.
"A really good GM can always come up with a better challenge for player
skill than leaving it to something randomized or mechanistic, because
players themselves are neither random nor mechanical."
"A really good GM can always stay truer to the game world by coming up
with results himself rather than leaving it to something randomized or
mechanistic, because then the GM's vision of the world is unclouded by
things occurring in the metagame like dice rolling across the table."
You're right, there are trade-offs, you're right, there is a preference
here, you're right, it doesn't fit the Threefold.
--
***************************************************************************
"I was pleased to be able to answer promptly, and I did. I said I didn't
know."----- Mark Twain, _Life on the Mississippi_
Jason Corley | le...@aeonsociety.org | ICQ 41199011
Uh, oops. *sheepish grin* It's even worse than I remember it being when
I wrote it!
>Frankly, this and others of the recent spate of threads has moved me
>pretty solidly over to the camp that rgfa is a forum for esoteric
>discussion of little to no practical value.
I've gone from vocal "newbie" back to lurker to mostly ignoring the
group, myself. I agree that it's gotten out of hand.
>While I still consider the threefold a decent kernel of an idea, I
>really wish discussion of it would go away.
I don't think the threefold is the problem as much as the hair-splitting
discussions that go along with it. There seems to be a *lot* of
intolerance for differing opinions on it, despite that the whole point
of the thing is to dispel intolerance. Two major things seem to provoke
the most hostility: using a term in an intuitive way or as the FAQ
describes it rather than in the arcane way an old-timer understands it,
or saying something even vaguely reminiscent of David Berkman or other
past "enemies" of rgfa. Despite the fact that I find Berkman's beliefs
regrettable and even offensive (as they've been explained to me), I've
been killfiled or threatened with killfiling repeatedly because I'm
"just like Berkman."
Some people really need to stop painting big red targets on the newbies,
and actually listen to what we *say*.
>If the threefold is of any use at all, then it should come up and be
>useful in "regular" discussion of actual game situations and problems.
Agreed. There is some value in meta-discussion and discussion of the
theories and models, but I too am tired of the hair-splitting and
condescension and hostility.
--
Bradd W. Szonye Work: br...@cup.hp.com
Software Design Engineer Home: bra...@concentric.net
Hewlett-Packard Cupertino Site, iFL Phone: 408-447-4832
That's actually a problem, however. (It's a problem for mid-triangle
people anyway.) Why? As David Alex Lamb said, it makes it impossible to
distinguish between the GM like Scott, who considers most all of the
threefold styles irrelevant, and one like me, who prefers Gamism but has
strong elements of Dramatism and Simulationism too. Even though there's
a good chance that we wouldn't like each others' games (I might think
his is pointless and meandering, while he'd think I'm a rules lawyer
obsessed with details), the threefold puts us at the same "point" in the
triangle. It treats the mid-triangle people as "all you folks" -- we're
second-class citizens, not worthy of distinguishing between us, unlike
the "corners" which all have their own sense of community.
This is *part* of the reason why mid-triangle people find the threefold
less useful. Another reason is that moderate, mid-triangle folks
encounter tradeoffs much less freqeuently than corner styles; the whole
insistence of the threefold as a "tradeoffs model" makes it seem out of
touch with reality to somebody who just doesn't encounter tradeoffs,
even when mixing the styles.
woodelf <woo...@rpg.net> wrote:
>it is exactly these two ideas that i'm trying to soften by re-modeling the
>3-fold, as it is exactly these ideas that i think lead to most of the
>arguments over the validity of the 3-fold, and prevent us from using it as
>a better communicative tool.
I don't really have anything to say about the second counter-argument,
but I agree with you on the first. I think that the insistence on
recognizing trade-offs is a flaw, not a virtue, because it alienates
people who never encounter a need to trade off one style to accommodate
another. I too believe that there are a wide range of styles where it's
possible to change the mix of 3fold components without encountering a
trade-off situation.
However, I also recognize that *some* preferences do require tradeoffs,
and it's precisely those styles at the "extremes" where there is little
room for deviation. Thus, the recognition of trade-offs is important,
but the insistence that there is *always* a trade-off, in all styles, is
a major flaw.
I agree that the "bubble" picture of the threefold is much better than
the "triangle" picture; near the surface of the bubble, you need to make
trade-offs or you break the bubble, but more toward the center, there is
a lot more wiggle room in what works for you.
This is not the only source of controversy re: the 3fold, but it is one
of the major ones.
This is, of course, one of the major points of contention regarding the
threefold. Some styles make trade-offs very obvious, whereas others are
very tolerant of them. It seems like the trade-offs are important only
to the strongest styles, like extreme gamism and strict simulationism
and the sort of mid-triangle people who try their darnedest to satisfy
all three axes.
For the more stylistically moderate, there's a lot more wiggle room, to
the point that there isn't really a trade-off at all. A flat, trade-off
model suits such people extremely poorly, and they're some of the more
vocal opponents of the threefold.
I agree with woodelf that a "bubble" model describes the general gaming
population much better than a "triangle" model does.
> your comment about making decisions deliberately contrary
> to a given style got me thinking: does anybody do this?
>
>It's not clear to me how this would even be possible. In the only
>example I can think of, if one specifically tries to avoid making a
>good story, one ends up being very concerned with story. An
>existentialist story is still a story.
I agree with Warren here. Being contrary to a style is merely a perverse
way of taking that styles concerns into consideration. You may have a
very different result, but you end up obsessing about the same things.
While a "negative style" may make sense theoretically, just as
existentialism is a distinct style, it's so unusual in practice that I
wouldn't worry about it beyond a footnote (just so that people know that
you've considered it).
> This is *part* of the reason why mid-triangle people find the
threefold
> less useful. Another reason is that moderate, mid-triangle folks
> encounter tradeoffs much less freqeuently than corner styles; the
whole
> insistence of the threefold as a "tradeoffs model" makes it seem out
of
> touch with reality to somebody who just doesn't encounter tradeoffs,
> even when mixing the styles.
As someone who considers his own style "mid-triangle" (halfway between
simulationist & gamist, but on the far side of the dramatist corner from
the center) I have to say that I encounter tradeoffs all the time, and I
find myself... confused ... by your insistance that you don't.
Maybe you only count something as a "tradeoff" if high stakes are
involved, while I consider even minor tradeoffs to still *be* tradeoffs?
In the part of your post I didn't quote, you talk about Dave Lamb's and
your games being different, despite both being "mid-triangle" and about
how you probably wouldn't like to play in each other's games. From my
pov, you've each made a different set of trade-offs. Your style has both
advantages and disadvantages compared to his, and vice versa. If his mix
of winging-it and mechanics didn't have advantages over yours (e.g. not
being "obssessed with details") no one would prefer his to yours. If it
didn't also have disadvantages compared to yours ("pointless and
meandering") then no one would prefer your style to his. And a
"tradeoff", in my view, is where one accepts that one's position has
- and has to have - both advantages and disadvantages. Like yours. Like
mine. Like everyones.
Erol K. Bayburt
Ero...@aol.com (mail drop)
Er...@ix.netcom.com (surfboard)
Add me in, I don't have trade-offs either.
>
>Maybe you only count something as a "tradeoff" if high stakes are
>involved, while I consider even minor tradeoffs to still *be* tradeoffs?
Nope. Maybe a mild tradeoff, once in a blue moon, but under normal
circumstances, no tradeoffs whatsoever. In fact, it is only in "high
stakes" situations that I would ever see any tradeoffs at all. For
minor things, there is always enough wiggle room to fit everything in.
>
> If his mix
>of winging-it and mechanics didn't have advantages over yours (e.g. not
>being "obssessed with details") no one would prefer his to yours. If it
>didn't also have disadvantages compared to yours ("pointless and
>meandering") then no one would prefer your style to his. And a
>"tradeoff", in my view, is where one accepts that one's position has
>- and has to have - both advantages and disadvantages. Like yours. Like
>mine. Like everyones.
Here is where you are missing what we are talking about. We are not
talking about the concept of trade-offs in general. We are talking
about the concept of trade-offs within the threefold model,
s vs g vs d.
Basically, here is an example of what I am saying...
Situation 1: I make the decision as "in world" as I would ever care
to make it.
Situation 2: I tell the best story that either I am capable of
telling, or that I care to take the time and effort to tell.
Situation 3: I make a decision that meshes just as well with the
world as 1, and tells as good a story as 2.
No tradeoff for me. I get to have my cake and eat it, too.
S1 - a: Pickier GM makes in world decision that is much more limiting
than the one I made
S2 - a: PGM tells the best story he can. It is more detailed than
mine.
S3 - a: PGM tries to do both, but they don't fit. He has to pick one
over the other.
Tradeoff for him.
The big point is, that even if I decided to include no story concerns
at all, I still would not put more effort into making a more complex
world model. Yes there is a tradeoff, but it is with something
outside of the model - my desired complexity/effort.
I *do* encounter threefold tradeoffs sometimes, but it's usually only
when my concern for one aspect of the threefold is higher than normal
(for me). For example, sometimes I get strong inclinations toward
simulationism; that conflicts with my usual gamism, and usually I rule
in favor of the gamism.
However, most of the time, my dramatist and simulationist leanings don't
interfere with my overall gamist slant, nor with each other. For the
most part, it's because I have moderate preferences (regarding S/D
anyway), and a little "wiggle" in any particular style doesn't
perceptibly affect the other elements of my style.
Another way you can view this is to say that I trade off simulationism
and dramatism against arbitrariness and whim, *not* against gamism or
each other. My style isn't strong enough to require balancing different
aspects of the threefold.
>>If his mix of winging-it and mechanics didn't have advantages over
>>yours (e.g. not being "obssessed with details") no one would prefer
>>his to yours. If it didn't also have disadvantages compared to yours
>>("pointless and meandering") then no one would prefer your style to
>>his. And a "tradeoff", in my view, is where one accepts that one's
>>position has - and has to have - both advantages and disadvantages.
>>Like yours. Like mine. Like everyones.
>
>Here is where you are missing what we are talking about. We are not
>talking about the concept of trade-offs in general. We are talking
>about the concept of trade-offs within the threefold model, s vs g vs
>d.
Right. The less "threefold-oriented" your style is, the less you need to
balance *threefold* concerns against each other. A GM with a high level
of arbitrariness in his style might trade off whim against gamism; that
GM can honestly say that the increased gamism does not affect the level
of dramatism whatsoever, because it only affects non-threefold concerns.
>The big point is, that even if I decided to include no story concerns
>at all, I still would not put more effort into making a more complex
>world model. Yes there is a tradeoff, but it is with something outside
>of the model - my desired complexity/effort.
Righto. And to such a person, it does no good to tell him "you must
trade off between story and world," because in his experience, that's
not where the trade-offs lie. In my experience, the most common
trade-off is detail vs effort, and the next most common is probably game
vs world. There are plenty enough people, however, where story vs world
vs game trade-offs are so rare that the threefold seems pointless or
useless, especially when you insist that trade-offs between the three
are inevitable.
That's how somebody like Warren can believe that the story vs world
trade-off is an invaluable part of the model while other folks don't see
a trade-off at all.
What's important is that we don't err in either direction: the
traditional problem is people who don't think that story *ever* trades
off with world, but it's no more useful to insist that story *always*
trades off with world.
1) By the nature of the threefold model there are *bound* to be trade-
offs, because the three corners are different
2) Those who do not experience trade-offs do not do so because they are
not as "purist" or "picky" as those who do - i.e. they simply don't care
about decision intent issues as much
3) Problems with this come about, ISTM, due to an expectation that the
model will describe all aspects of a style. I don't believe this is so.
It's a bit like describing a car and saying that it's red. The
specification of a colour cannot be said to completely describe the car,
and for someone who really doesn't care what colour car they drive it
could be seen as useless information. But for those who have a definite
preference for a specific colour (or even more than one, or a mixture)
it still conveys useful information.
>I agree that the "bubble" picture of the threefold is much better than
>the "triangle" picture; near the surface of the bubble, you need to make
>trade-offs or you break the bubble, but more toward the center, there is
>a lot more wiggle room in what works for you.
>
I disagree, not because I don't think that the idea that some people are
less "connected with" the triangle is valid, but because I think the
"bubble graph" should have a myriad "origins". The fact that decision
intent issues are not important to you probably means that *something*
else *is* important, but it will not necessarily be the same "something"
for all "decision intent unconcerned" people. Better, I think, in this
case to leave the triangle as a 2D system and clarify that it is only
one aspect of a description of a style of play - one dealing with
decision intent.
Cheers,
Andy
Sounds about right to me, too.
> There are plenty enough people, however, where story vs world
>vs game trade-offs are so rare that the threefold seems pointless or
>useless, especially when you insist that trade-offs between the three
>are inevitable.
>
>That's how somebody like Warren can believe that the story vs world
>trade-off is an invaluable part of the model while other folks don't see
>a trade-off at all.
My favourite "story" over "world" example is the number of destroyed
ST:Voyager shuttles. I imagine the writers saying "but we *have* to blow up
another shuttle to set up *this* situation required by today's plot". My wife
and I lost count midway thru Season 1; I suspect by Season 5 (where we're
viewing now: repeats from the Space channel in Canada), enough mass of
shuttles have been destroyed to rebuild Voyager twice over. One of these days
they're going to have to show us the giant shuttle replicator tucked away in a
corner of Cargo Bay 1.
I forget exactly where, but Marion Zimmer Bradley was once criticised for
having the journey between two particular cities take several times longer in
one book than another, for no apparent in-world reason. She apparently
responded huffily that in her world, journeys took the time demanded by
dramatic circumstances.
--
"Yo' ideas need to be thinked befo' they are say'd" - Ian Lamb, age 3.5
http://www.cs.queensu.ca/~dalamb/