Dispatches, thoughts, and miscellanea from writer Jon Konrath

Interactive fiction versus games

I’ve been thinking a lot about interactive fiction, trying to find good examples online and learn how to turn existing books into games, or write new hybrid game/books, and it’s made me consider the definition of the two.

First, I’ve been playing with this tool called Twine. It reminds me a lot of the old Hypercard, which is sadly gone. Twine essentially lets you create an interactive game by creating a bunch of little boxes or cards in its interface (they call them Passages) and then connecting them together. It uses a wiki-like syntax for creating the links. You can also use a collection of macros to do basic if/then logic and set/get variables, or you can use straight JavaScript to do more. When you’re all done authoring, it spits out the target in HTML, which you can easily host wherever.

I’d previously looked at another tool called Inform, which produces a compiled output that can run on a z-machine. Back in the Infocom days of Zork and other programs, they used the z-machine format for text-based games. Now, you can get a z-machine interpreter for just about any platform (including phones) and can play old games like Zork, or a multitude of other games that have since been authored.  (Although playing a game that involves a lot of typing is not that great on a touchscreen phone.)  For me, Inform was a bit of a dead end, because hosting a z-machine game on the web isn’t that intuitive (there are applets and whatnot, but it’s a huge pain and a slightly clunky end-user experience) and learning how to develop something in Inform has a massive learning curve.

As I thought about this, there were a bunch of different types of games or fictions possible with these tools. And in trying to differentiate them, I started thinking about them along three (or four) different axes.

First, there’s the content-per-page axis. Think of a conventional book: it’s got chapters, which vary in size, but are usually a few or a few dozen pages long. In a paper book, where you’re deeply immersed, that’s an okay chunking of the content. Contrast that with a game like Zork, and you’ve got maybe a sentence of content at once. You aren’t thrown long passages of paragraph after paragraph; you are presented with maybe a sentence or two between commands.  (If you don’t remember, here’s a video.)  On a web-based piece of Interactive Fiction, there’s going to be a sweet spot between those two. You want the person to be immersed into what you’re doing, but you don’t want to present them with ten thousand words of scrolling.

Second, there’s the linearity axis.  A conventional paper novel is completely linear: chapter 1, chapter 2, and so on. A choose-your-own-adventure book is a typically a tree structure – here is a great example of one. There’s no real outer bound on this axis, except that you can get more and more insane with the number of nodes, choices, choices per node, and endings. And you can loop. Go dig up an old C343 computer science book and read up on depth and breadth for more info. But there’s going to be a sweet spot there, too. The old Bantam Books CyoA books were bound by their published length, about 120-140 some pages. On the web, there’s no such limitation, aside from the reader’s patience.

(Also a note on linearity: just because a book isn’t having you make choices, doesn’t mean it’s not further down the linearity axis. Even the most rudimentary plotted books are sometimes jumping between the main story and a B story. Fiction can start at the end and work backwards, or jump around, even within a linear book. And things like footnotes and endnotes give you the ability to “jump” to the side for a moment to give you some side info. And you’ve got stuff like Nabokov’s Pale Fire or Cortzar’s Hopscotch, which make Zork look almost linear.)

Third, there’s the game logic axis. Printed novels have no game logic; there are no variables, no javascript, no programming. A game like Zork has a ton of game logic: you have inventory, there are combat rules, things happen at random times, and so on. A dungeon crawler text adventure could be entirely dynamic, spitting out a new map every time you came.

There’s also potentially a fourth axis, which is the presentation level. Books are text, maybe some images.  You could add in more styling, graphics, sound, video, and so on.  If you want to go whole-hog, consider a printed book versus the presentation in a Grand Theft Auto game.

All of this has me pondering what to do for a book like this. The simplest thing would be to take one of my linear novels, like Summer Rain, and make it web-based; a web page per print page, and maybe add in some pretty pictures. That’s pretty boring, and useless – you could just go download the Kindle version. The next level would be taking something like Rumored and arranging it in a tree-like structure, with wiki links between the nodes. That could be interesting. It also makes me think about going in the opposite direction, writing a book that’s interconnected in a web-based structure, and then flattening it into a linear print book. I kinda-sorta did that with The Necrokonicon, which went from wiki to print.  All of the hotlinked words were bolded in print, indicating you could manually page over to that topic.

A project I started messing with was the idea of a game based on a book, something with game logic built into it. I started writing a Twine mashup of Summer Rain and the Necrokonicon. You wake up in a boarding house room in Bloomington in 1992, and then you wander around the sandbox of campus, almost GTA-like, getting dressed and walking to Lindley hall to log into a VAX computer, find people to hang out with, spend your few dollars getting something to eat on Kirkwood.  This was a fun project to start, but exhausting. I needed a solid set of stories to tree up within this large matrix of the campus topics, like people you would need to meet or tasks you would need to accomplish, and I ran out of steam on that.  I also wasn’t finding the right balance on axis 1, unsure of how much text to put on each page. It was a fun distraction, but within a few days, I barely had my house and the few blocks around it mapped out; I could easily burn a thousand hours trying to world-build the thing, and that wouldn’t even get into the story.

I’ve got to get back to other writing, but I do want to do something with this at some point.