Chris Crawford was kind enough to send over a review copy of "Chris Crawford On Interactive Storytelling." Between State of Play 2, Accelerating Change, my own writing tasks, and Second Life, reading his book kept getting pushed down the stack, but I've finally had a chance to read it. I disagree with Chris on several points, but I still quite enjoyed it. I also appreciate his zeal in blazing his own path.
Read on for the ups and downs of "On Interactive Storytelling."
First and foremost, I am very much from the school of "Story? Who gives a rat's ass about the story??" school of game design. If it isn't fun when hacked up with blue squares, grafting buzzword-enabled graphics and having Claudia Christian doing the voice acting isn't going to make it fun. Neither Magic: the Gathering -- Armageddon, Road Rash 64, nor Second Life have any story to speak of. So, at first blush, I seem like the wrong market for his book. As such, I was pleasantly reminded that Chris is way too good of a game designer to avoid teaching you something every time he writes.
His book opens with a discussions of interactivity, verb thinking, and the interrelation of storytelling and culture. Small quibbles aside, these opening chapters are full of useful information, especially about the storytelling focus ("It's about people, stupid!"), place (stories occur on stages, not maps), time, choices, spectacle and a nice rant against the "tyranny of the visual." He moves on to discuss interactivity ("A cyclic process between two or more active agents in which each agent alternately listens, thinks, and speaks.") and how it factors into human cognition. This section in particular should be read and understood by all game designers.
At this point, the book finally introduces interactive storytelling as a distant cousin of games, books and movies. He goes to great lengths to explain why interactive storytelling won't (can't) evolve from any of these other art forms, focusing in on the tensions between plot and interactivity. In particular, he correctly points out that many apparently interactive forms don't actually allow the player any choices and that true interactive storytelling needs to allow the player many closely balanced decisions.
His next sections attempt to explain why interactive storytelling attempts have failed, focusing first on the fact that programmers are lousy storytellers and most storytellers are lousy programmers. I think that he spends too much time here -- and that there are simpler explanations related to content creation and the inability to handle realistic grammars -- and we all know exceptions to his rule. Fortunately, he follows with a wonderful section on simple strategies that fail. I found this section especially interesting since many of these approaches are also responsible for skyrocketing game development costs.
He then discussed several strategies that might work, including data and language driven models. Again, these discussions apply equally well to any game design and are quite informative. Like the rest of his book, he does an excellent job of providing citations and resources so that the reader is able to explore these concepts more fully. His focus on visual languages, and the inverse parsing he uses to make an effective user interface, are extremely thought provoking. Finally, his personality model section is an excellent introduction to building a personality UI. Great breakdown and discussions.
The last third of the book details his experiences and techniques in building the Erasmatron, his interactive storytelling engine. While technically interesting, I found these portions to be less useful.
Where I disagree with Chris is the overall approach to building interactive storytelling. Given experiences with simulation and user creation, it seems to me that you might be more able to create "interactive storytelling" by expanding something like Grand Theft Auto to allow a wider array of choices. Alternately, users could be recruited (a la The Diamond Age) to create a truly interactive experience without having to build believable AIs. I also think that place and display is an important component and that a truly interactive story would be better in a 3D environment that as text.
However, I hope that Chris continues to push the boundaries of interactive storytelling. Much like the relationship between MUDs and modern online games, text-based interactive stories can explore design spaces and conduct experiments that aren't happening in mainstream games, which is very good for all games!
I enjoyed the book, and would recommend it to anyone who thinks that NPCs can never be anything more than vending machines and monsters...
My take on the book's thesis is: Stories are not games. Stories are about people interacting. Therefore, to make interactive storytelling, build a simple world with well-designed NPCs that can be interacted with on as deep a level as possible. Throw in a director AI to ensure there's a plot.
An even more boiled down summary is, "Interactive storytelling is about NPC AI."
A boiled-down summary of MMORPG design books is, "Virtual worlds are about interacting with other players."
A boiled-down summary of adventure games is, "Adventure games are about interacting with things, adding a backstory."
Personally, I think of interactive storytelling, virtual worlds, and adventure games as sub-parts of a larger entity.
Posted by: Mike Rozak | Jan 28, 2005 at 19:18