A confession first: the subject and the mechanics of my Twine story for this Ludum Dare were decided and thought way before the theme voting started. It was going to be a story about the life of an ethically-challenged guy and a game about telling other characters tre truth or some lie and getting away with it.

This was my game after 2.5 hour of work.

My intention was adapting the idea to the theme once it was announced. This was really a crazy bet: some of the most arcade-oriented themes would have forced me out of this Ludum Dare. Fortunately, You are the monster was the best theme for me, since my head was already full of storylines that allowed the character to become a child psycho, a teen womanizer or an adult corrupt leader.

Since I knew what I was going to do, I had a luxury that many participants lack: I could plan. This plan, that I laid down using Todoist, was the only game-related thing I did before the compo. It was also the most important thing I did.

 

First step: adapt idea to the theme.

I allocated 30 minutes for this, but used more like 30 seconds.

 

Second step: configure environment.

Not that Twine has a lot to configure, but I created a git repository. I also used the Twee2 tool to write Twine from a text editor, which had been released in alpha 3 days before. Twine writers: check it out.

 

Third step: code the game logic.

I allocated 5 hours for that, thinking that it would take 4. This is a hypertext game and the logic is absurdly simple, right? I spent 6.30 hours: 162.5% of my initial estimation.

Got image from http://ghiasi.org/2013/10/hofstadters-law/

First I worked fast enough, but then all the silly details like making sure variables are given a value at the beginning so there’s never a division by zero started raining on me. By the time I finished, I was a bit down. I had no idea what was coming.

 

Fourth step: put a debug method in

Twine has its own debug mode, but apart from that I wrote a footer passage that printed all the variables in a readable way. Twine writers: do this. And don’t forget to remove it.

 

Fifth step: write an awful lot of text

This is where everything derailed. I estimated 10 to 15 minutes to write each scene, averaging 200 words, no revisions, no care about style, just churn the bastards out. With that I expected to hit 50 scenes with ample time to rest. I wrote down the ideas for the storylines and started writing the story.

That night I had 13 scenes. Each one was taking 30 minutes at the very shortest: I was generating content 3 times slower than planned.

22 Twine passages. I'm writing way too slow. I'm not sure what I will be able to get done in time. #ludumdare #ldjam pic.twitter.com/K8AUZ32jVr

— Pseudavid Menti (@pseudavid) August 23, 2015

Why? Well, did I mention that I’m not a native English speaker? But that was a lesser problem than actually making up the story. Not having a tight plot but vignettes scattered across the whole life of the character left lots of room for improvisation, but my writing skills simply didn’t cut it. Hofstadter’s Law again.

Since content was needed for the player to appreciate the gameplay, I was in serious trouble.

 

Sixth step: polish the CSS a bit

Very easy part and a relief from the stajanovist story hell I had written myself into. Twine authors: Google web fonts are your friends.

 

Seventh step: write another awful lot of text

Until I broke down and thought of quitting. It wasn’t going forward.

 

Eighth step: lay down and plan again

This saved my game.

When you’re so behind schedule that the schedule is going to go round the world and get you from the back, is taking a long rest at bed a fine idea?

Yes sir. I did and it saved my game.

All the time I had been writing from a loose and fuzzy list of story ideas. I had no measure of progress, no indicator of completion. That was what took my energy away.

So I lay down on the bed and thought of what I had left. I went through all my story ideas, fleshed them out and divided them into scenes. Then I went back to Twine, created the empty passages, went back to bed, thought of another block.

It took 2 hours but at the end I knew exactly what I was going to do.

 

Ninth step: write according to plan

Now I wasn’t just churning out content: I was filling some empty boxes. When all were filled, I would have a mininum viable game. A completion indicator totally changed my attitude. Soon I was writing the best and fastest text in the game. I knew when and how I was going to make it, and now it was only a matter of physical resistance.

Two hours before the deadline I finished the minimum planned game and could take some time to add extra content.

 

Tenth step: wrap it up and upload

Hint: Hosfstadter’s Law again! IT TAKES LONGER THAN EXPECTED.

The final Twine with its glorious 44 passages.

A conclussion

I wrote some conclussions about content and jams here and here.

Had I not planned in advance, I would have not overcome the insecurity when I saw everything taking twice as long as expected.

I finally couldn’t test the game thoroughly. I barely found one bug, but the gameplay is not quite balanced and according to comments most players like the game but don’t get quite what they were expecting.

Yesterday I removed all the code, counted the words of actual game text and was totally amazed to find out that I had written 8,500 words. Lots more than I believed, quite long by Twine’s standards. I think that the biggest problem with my game is that it’s too short.

Tags: interactive fiction, Twine