Re: Game Design in General.


Fri, 25 Aug 1995 11:22:31 -0400

whizzard@uclink.berkeley.edu (Gerry Kevin Wilson) writes:
> Well, as interesting as the last week or so of Inform questions have
> been, I find myself bored enough to start a new thread. It occurs to me
> that probably every one of you knows exactly how I go about writing a
> game, but I've yet to hear the slightest on how the rest of you go about
> it.

First I come up with some nice scenes and images. (For "Weather" it
was... mm... I don't want to get into spoilers, but it was the idea of
seeing the same scenery in different weather and times of day.)

Then I decide to write a game.

Then I come up with the story. Theme, plot, characters, ending (not
that I'm any good at characters.)

Now I need puzzles. I already have the first: figuring out what the
plot is and what the ending must be. (This is always a puzzle. What's
the point of IF, else?) I start dumping out more scenes, writing them
down on a big piece of paper, considering each one in terms of
possible puzzle content. Maybe half of these scenes will make it into
the final draft, most likely heavily altered. However, since I'm
trying to maintain the theme and atmosphere of my story as I create
all of these images, I know the final result will be on the right
track.

As I'm doing this, I start to pick out scenes (and puzzles) that I
like, to be in the final version. Later scenes are shaped with
attention to fitting in with the ones I've chosen. At this point I'm
concentrating on making things weave together, so that nobody notices
how patchwork the whole process is. :)

Eventually I have this huge structure carefully balanced on top of my
head, a little machine with every part working correctly. Then I write
it down, generally as a map or plot diagram with lots of
incomprehensible little boxes.

Then I start implementing.

There is a non-zero chance I've made an error in the design process.
This is likely to be a trivial alternate solution. (It's easy to avoid
the mistake of making an impossible game -- just play it through on
paper -- but it's hard to see every possible combination of actions,
when I just went to all the effort to make everything intertwine.) If
I find an error, usually halfway through programming, I scream and
rend somebody limb from limb, and then start patching. Patching
usually involves throwing away some good stuff in order to fix the bad
stuff. This sucks.

Then I finish implementing.

"Weather" did in fact turn up an error halfway through programming --
in fact, it was the stupid kind -- the game was unsolvable. I did some
hasty scratching on my plot diagram, and nobody's complained that the
seams show, so it must have worked. Heh. (I don't have my notes here,
and I don't remember exactly what the error was. Sorry.)

--Z

"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the borogoves..."