There are several challenges associated with delivering a compelling
story using a game as a the medium. While gaming companies have, over
the years, developed various coping mechanisms for these, at a low level
they are still a barrier to delivering literature as a game.
1. Player as Character
Well, yes, the strongest selling point for delivering a truly immersive
experience is that the player essentially takes the role of one of the
characters. This, however, presents a problem. The more strongly we
characterize the protagonist, the more likely we are to drive a wedge
between the player and the game. Obviously, personality drives actions.
If the author controls the main character totally, there's no
interactivity. If the author offers a tabula rasa main character, the
author no longer has access to all the "interesting main character"
2. Choice as an Obstacle to Plot
Here's the "well, duh" one. Simply put, an author needs control of the
plot in order to drive the story. Sure, we can simply throw a lot of
towns, random monsters, and quirky NPCs at the player and call whatever
happens a "story." But I think, qualitatively, we realize that there is
a difference between a well crafted story and a laundry list of events
that happen. However, of course, the more we shift control from the
player to the author, the more we make the player an observer.
3. Game Genre Limitations
Game characters seem to spend most of their lives killing hordes of
wandering monsters, collecting enough plot knick-knacks to open a
curiosity shop, having strained and linear conversations with NPCs, and
playing odd little logic puzzles (which seems to be the de facto method
for locking doors). In short, the characters participate in a lot of
highly-interactive mini-games, presumably in order to make up for the
fact that they're getting railroaded in the big picture. While this no
doubt makes for a fun game, it's tricky to have only a half-dozen tools
for expressing authorial intent.
4. NPC Limitations
Here's another "well, duh" one. Ever try to become friends with one of
the bit-part NPCs in a game? Unless the game is scripted such that NPC
X is actually a key character, NPCs have all the character and
background of a plank of wood. IMO, this is one of the reasons why the
"survival horror" genre works. Zombies that can only say "graaauuughh"
are a bit easier to buy than villagers that can only say "Welcome to
To be fair, genre and npc problems are technology limitations. Just as
Diablo built dungeons on the fly, I fully expect future games to build
new areas and characters on the fly. Like the Diablo dungeons, there will probably
be X set templates with Y opportunities for variance and Z constraint to
insure the new people/places still drive the plot. In fact, in my free
time, I've been working on a database-driven engine to do just that.
I'm sure I'm not the only one.
Characterization and choice issues, on the other hand, are much more
difficult. Choice could probably
be fixed by an engine that allowed some open-ended event progression,
while still nudging the plot to follow certain archetypal patterns.
Wayne C. Booth, in "A Rhetoric of Fiction," identifies three things
which drive readers:
1. Interest in the welfare of the protagonist (Will Jane go back to
2. Desire to learn new things (How many cities are in the game?)
3. Desire to see certain patterns completed (Good triumphs over evil.
Scrappy young outcast works his way up to overcome decadent social elite
An archetype-based story engine could simply be told that the author
wants to use pattern X and Y. The engine could then script events to
drive them toward X and Y. However, such an engine would be a technical
marvel. I'm betting on Bullfrog or Square being the first to pull such
a thing off.
Regardless, IMO, the characterization problem is a genre
limitation. As long as there is "you" in the game, stories will
necessarily be constrained by "your" personal morality, playing style,
Kuro5hin is full of mostly freaks and hostile lunatics - KTB