This is justified in the context of this game, because the PC is being told
(or helping to tell) a made up story.
But how do people feel about this kind of thing being done in a game without
such a frame story ? Would it feel too contrived for the geography to be
generated to fit the plot, so that whichever way the player goes, they find
room A before room B ?
Static fiction authors get away with coincidences like this (certain events
which happen to happen before others) all the time, but interactivity makes
it easy to wreck the plot by doing things in the wrong order - wrong from
the author's point of view, anyway ...
I think this is a reasonable thing to do, to a point. I might start to
question the effect if it felt as though the player's interaction had
no purpose whatsoever, and there was nothing to do but step through
places and events laid out in order. But as long as there is some
feeling that my participation as player is important, I don't mind too
much. (For that matter, the player's participation can be important to
the experience, as I think it is in "Rameses", even if it's not
possible to change anything about the plot at all.)
Jan ! I would love to know what comes after the comma, or if it was really
intended to be a full stop ...
Just a full stop.
I think that it depends entirely on how it's pulled off and in what kind
of game it appears. Certainly, if you give some indication to the
player that this is going on, there wouldn't be any problem with it.
This does not mean that you _have_ to give such indication to the
player, but I think more care has to be taken in that case.
I disagree partially with Jon's statement:
If you could guarantee that the player would only play through the game
linearly, never restoring to an earlier stage, never dying and
restoring/restarting, and never using information from a previous
play-through on a repeat play-through, then I would agree completely.
Of course, this is impossible.
Still, in some games, it would just be fine to not mention the dynamic
generation of apparently static content. But when the player is trying
to accomplish something (such as solving a puzzle, or attempting to see
what happens if he didn't just happen to walk into the one coference
room (of many) in which his wife was cheating on him with his best
friend. Such a player would be surprised to find that they can't solve
the puzzle by trying to go a different way, or that no matter which room
they chose, they couldn't avoid the discovery.
In many cases, this surprise would be short-lived and the player would
figure things out quickly, which sort of defeats the purpose of hiding
the dynamic generation from the player in the first place. This
wouldn't be a huge problem, but it's something I would try to avoid
provided a) I didn't _want_ the player to be surprised and b) the
work-around wasn't too difficult, time-consuming, or literarily awkward.
On the other hand, I can imagine that there might be situations where
the discovery that some of the apparently static content was generated
on the fly would be postponed until quite a bit after a potential
decision based on that "static" content. Example: the player makes an
decision early in the game whose effect is based on the layout of an
area that the player won't get to until the endgame. Then the player
might feel cheated out of a solution that they thought was particulalry
clever (even though it used knowledge of a prior run-through).
Another problem that might arise (this is primarily important for games
with puzzles) is that even after the player learns on his own that the
map is generated dynamically, he might be plagued with doubt about other
seemingly static aspects of the game. "The other security guard hates
me because I decided not to take the bribe. Have I just stuck the game
into an unwinnable situation, or is he supposed to find a reason to hate
me no matter what I do?" Granted, a good game should never put the
player in such an unwinnable situation, but as was discussed in the
Bathrooms thread earlier, I think that the game writer should never
assume that the player knows his game is well-written.
If either of these types of situations were possible for a game I was
writing, I would try to minimize the damage. One way would be to
include in the readme and about-message, something like "None of the
puzzles here are supposed to require advance knowledge of the plot or
layout of the game, and the game should never get into an unwinnable
situation without it being obvious." But mostly, I'd just try to avoid
this sort of situation.
Of course there are dozens of artistic reasons why one would want to
generate exactly this type of confusion, and in any of those cases,
these considerations don't necessarily apply.
One of the many game ideas I'd been kicking around had as part of its
charm a very cheesy "twist"-ending that would be completely different
for different run-throughs (depending on the way the game was played,
not randomly, because I hate twist-endings that aren't properly
foreshadowed). I wanted to generate a lot of the "static" content on
the fly, depending on past actions of the player. So players that have
been poking around under cabinets and behind paintings would actually
find lost keys and hidden safes, while those who went around
interrogating everyone would find all sorts of gossip. It was meant to
be a very small game with enormous scope. (I find this happens to most
of my ideas. I come up with an extremely tiny game idea with only one
room, or no NPCs, or a single very simple gimmick, and soon enough, I
find that without actually breaking any of my original constraints, I've
got a game idea the size of Curses.)