1st Place in the Interactive Fiction Competition

This entry was posted by on Wednesday, 18 November, 2009 at

Jack and I are giddy with glee, as our recently released text adventure just won 1st place in the 2009 Interactive Fiction Competition. As someone who’s been tracking the community of IF authors for 15 years, this is a bit of a fantasy come true. Most of the really accomplished, famous authors didn’t enter games this year (or released games outside the competition), thus making room for a new generation of authors. We owe the old-timers big thanks for inspiring us, writing great tools, and giving us a chance to shine!

If you haven’t played the game yet, go do so! Cuddle up with a laptop and cup of cocoa. You can get the game (and the source code too) from the main main website we set up. You can file bugs there too.

Note: My one frustration is that genre seems to have a bad reputation among gamers. The natural-language parser is mocked for being overly primitive and unfriendly to casual players. Paraphrased (from a friend):

The creature approaches!

> swing sword
What do you want to swing the sword at?

> creature
What about the creature?

> attack creature
What do you want to attack it with?

> the swod
I don't understand "swod".

> sword
What do you want to do with the sword?

I'm sorry, the creature has eaten you.

In reality, enthusiasts of text adventures consider the primitive parser to be a feature, not a bug. It expects commands of the form “verb noun” and only understands about 30 verbs. So it’s an easy interface to master; experienced players know them all by heart. If you haven’t played text adventures before, be sure to have this crib sheet with you, as it explains the sort of commands most games understand.

And now the obligatory post-mortem on the experience, taken from a post I made on the rec.arts.int-fiction newsgroup yesterday.

  1. The methodology of “write the transcript first” really works. Emily Short mentioned this technique on her blog, and I’m here to testify. As a programmer, I’m always tempted to start fiddling in I7 in technical ways, wondering if I can implement some clever algorithm — and then later trying to figure out a way to justify its use. This is not the way to write a good game. Instead, come up with a GOOD STORY first (or partner yourself with a great writer like Jack), and write out the entire hypothetical transcript first. Think of it like a screenplay: first conceive the whole experience from the user’s point of view, and decide if it’s a good script. If it is, then worry about the implementation. (For the curious, the original transcript Jack wrote — before a single line of code existed — is over here.)
  2. Keep the player captivated at all times! We goofed by requiring too much repetition of mundane routine for the first half of the game. IF geeks and programmers generally had the patience to muddle through (or noticed the status bar changing, and were intrigued about the double-meaning of things). But at least half of the players out there — including some beta-testers — rightfully had no patience for such a thing. “Just let me do something INTERESTING already!” Many people simply weren’t able to delay gratification (or keep faith) as long as we’d hoped. Especially when you have 20 other games standing by, ready to test. Given the blog reviews, we were convinced we were headed right for the Banana of Discord. Emily’s review and Jim Aikin’s reaction were the canonical example of this. We’ve learned our lesson here.
  3. Avoid linearity. This was my fear all along, when I first read Jack’s transcript… particularly in the second half of the game. I liked the story enough to overlook it, but reviewers correctly called us out. Killing invading bots may be fun, but this still ain’t no Photopia. In the future, we need to really construct some non-linear mid-game plot flow.
  4. Write a hint system.This seems to be the most requested feature, and I was surprised. I grew up playing Infocom games, when games were supposed to take weeks to solve and ‘walkthroughs’ were expensive InvisiClues you had to mail away for. Ordering the walkthrough was a badge of shame, an admission of defeat. These days, the culture seems to have changed quite a bit. People not only expect every game to have a walkthrough, but they check it after being stuck for 10 minutes (!) Maybe that’s just the environment of the Comp (when people are in a rush to “get through” quickly and judge), but clearly an in-game hints would make the game much more accessible to a wider audience. Perhaps fewer people would have run screaming from the repetition. 🙂