[Every day during GDC, Everybody Dies creator Jim Munroe is blogging for GameSetWatch discussing the creative process for the GDC-related text adventure he'll be building for us. Here's part three, following Monday's and Tuesday's entries.]

I ran into Ernest Adams yesterday, top hat and all, on my way home from the Ten Bit party. His thoughts on interactive narrative have strongly influenced me.

I'll rant to anyone who will listen that interactive andnarrative are not this simple thing to combine -- they're not the chocolate and peanut butter that people seem to imagine. The more the player tells the story the less the author tells the story, and the inverse is true.

So I've been trying to come to terms with the idea that as a fiction novelist a lot of the tools in my toolkit aren't that useful to bring to bear on game making. I personally believe that a linear plot doesn't suit the medium.

Despite that, my recent well-received text game has a completely linear plot. While I spared Ernest the role of Father Confessor while we chatted, I have a weird feeling that this linearality is a kind of backsliding on my part. Linear plotting in games is just wrong.

It's kind of like the beginning of music videos -- they often literally acted out the lyrics of the song, modeling themselves as mini-movies.

And they were usually bad, though in retrospect hilariously so. and eventually grew out of this literal phase to become the much less linear and more evocative creatures they are today.

Happily, Margaret Robertson's "Stop Wasting My Time and Your Money: Why Your Game Doesn't Need a Story to Be a Hit" has re-calibrated me.

She's an advocate for great writing in games, but makes the case that linear storytelling is intrusive, expensive and usually unsuitable thing in games.

After highlighting what went right in games like Ico, Eternal Darkness, and Half Life 2, she challenged the audience to explore different methods of creating richer games with creative solutions native to the form.

For this GDC text game, I'm starting to imagine a game that pivots on social interactions. By modeling the conference as a kind of social ecosystem full of people that I can simply drop my player into, I'm hoping for a baseline dynamism that won't have to rely on a linear plot for a compelling experience.