How I make my Choice of Games interactive fiction
I get asked this a lot, so here's my usual routine for writing my interactive fiction for Choice of Games. I've been making these types of games for nearly 9 years now, and developed these techniques over several projects.
For other kinds of game writing and narrative design I work differently depending on the project, though I tend to work from top down and use code as an outlining tool in a similar way.
Concept
First comes the idea, vibes, general sense of what I want from a game. These have often been rattling around my head for a while. For Creme de la Creme, I'd been poking around at the idea of a school story during the final months of making Blood Money. For The Earth Has Teeth, the seed of the idea came from five years before and had been forgotten about until I picked it up again and polished it up.
From this I build a an elevator-pitch paragraph which I send to my editor. That includes the PC's role, the choices they're facing, a few different ways in which they might handle problems, some overarching goals, and a few major NPCs.
Outline
Once the concept is picked, I develop the concept into a chapter-by-chapter outline, including the PC's stats, mutually-exclusive goals, NPC relationships, major branches, and how the PC is engaging with the world at each stage of the game.
As part of this, I include a choice per chapter to show how the mechanics will work.
The outline will go through at least one revision in response to my editor's thoughts. For example, in The Earth Has Teeth I was asked to include more concrete goals for the PC to strive for. In Honor Bound, I was asked to specify more details about why the PC's been assigned to be Catarina's bodyguard, which led to a much richer background for the PC.
I'll often tweak the stats as well: I tend to err on the side of more complexity and then realise - from editorial feedback, once I start coding, or both - that streamlining would be wise.
Chapter Planning
Once a project's outline is greenlit, I can start working on the game itself. I usually plan chapters by hand, doing a broad-strokes outline scene-by-scene. Before I start planning I like to know how and where the chapter will end, in what ways it will be branched at that ending, and what is required from the chapter.So I lay that out. For a shorter chapter, or something where I want to be especially precise with use of stats, I might plan individual choices - I did this with Noblesse Oblige, where I was working to a tight wordcount. Usually, I do it more broadly.
At this stage, I make sure that I'm aware of where the branches are. For example, in Chapter 3 of Honor Bound there are many different characters with which the PC can spend time during a camping trip, and I planned that carefully before starting to code to make sure that it was evenly spread and that I had a sense of what everyone was up to.
Coding
Some ChoiceScript writers draft and code at the same time, which I used to do (and tend to do for smaller games). But I find it a lot better to code the whole of a chapter - sometimes even two - as a way of doing further planning and get a sense of how large and complex a chapter will be.This is where a lot of the creativity happens: it's when I think of interactions that feel fun, ways in which the PC can engage with the world, and friction that can occur. For me, I like separating that kind of creativity with the creativity of drafting, because it's a different kind of brain power.
When I code at this stage, it's all playable and able to be tested either manually or with ChoiceScript's QuickTest and RandomTest features. I include all the stats changes and tests that I want, and at the end of coding I use a variation of Emily Short's method described here to test how difficult the stat tests are. I also use RandomTest to check how many times a stat is tested, on average, so I can tweak things if one stat is disproportionately powerful compared to another.
Whenever I don't check my stats and rush onto drafting I always regret it. Still, there's always at least one point during a project where I'm so eager to race ahead that I do exactly that. We're all human.
Drafting
For my workflow, I give myself targets for a day or week's work. I used to use wordcount for this - but since sometime during writing Royal Affairs, my targets are based on how far through my code skeleton I am. This gives me a better sense of scope than wordcount, and is far less demoralising if I happen to have a low-wordcount, but high-productivity, few days.I tend to work from top to bottom of my text file - which broadly translates to beginning to end of the chapter, though not exactly - though in some cases I'll start at the end or with particularly wordy, complicated, or thematically-repetitive scenes. Examples are intimate scenes, which are always very wordy, and things like the date or proposal scenes in Creme de la Creme and Royal Affairs, which are all individually written but have some shared beats and can get tiring if they're all written at the end of the chapter.
When drafting I'll often shift things around that I coded - this is part of why I code first, so that I've got more flexibility and less sunk-cost-fallacy about that kind of thing.
Once I've drafted a chapter, I playtest it and do whatever revisions I think are needed before sending it to my editor.
Revising
And once I get my notes from my editor, I do the same again.Rinse and repeat Chapter Planning, Coding, Drafting, and Revising until the full draft is done, including WIP-sharing and revisions in response to player feedback along the way. Then...
Review, Playtesting, Draft Revision
Another Choice of Games editor who's fresh to the project plays the full game and gives detailed feedback. Along with this, CoG will invite testers to play and give feedback, as well as paying continuity readers to go through the text of the game to spot errors or anything confusing.This is one of the more intense stages of the writing process. It involves a lot of feedback triage, prioritisation, attention to detail, clear-sightedness, and humility. It's usually a month or two of reading feedback and making changes in response.
Each time I've gone through this stage, I've added a significant amount of writing to the game, to the order of tens of thousands of words.
Once everything is finalised, it goes to CoG's copyeditor. And at that point, I'm pretty much done with the project - except for social media marketing, an author interview with CoG, and anything else around it that I might need to do. The timeframe very much depends on the size of the game and copyeditor availability, but it's usually two to three months between it being sent to the copyeditor and the game being released.
At this point, depending on what other work I'm up to, I'll start writing up elevator pitches. I'll have been thinking about my next ideas during the mid-to-late stages of the draft anyway, and I'll often have come up with some possibilities during playtesting when I don't have writing to do. So, then I start it all again!