Some people who are starting up the process of writing a ChoiceScript game have asked how to plan/outline/storyboard/etc. a game before writing.  I don’t presume that we know the best way, let alone the one true right way to do things, but I thought people would be interested in how we plan our games.  This is a monster length post, so I’m going to put it beneath a cut.

I’m going to start with the end of the planning process, with the intermediate products that we like to have in hand before we start actively writing vignettes, and then I’ll talk about how we get from an initial idea (“we want to write a Napoleonic naval adventure” or “we want to write a space opera adventure”) to these intermediate products.  When I start writing a vignette, I want to have:

• A one-sentence capsule summary of what the game is about–what are the themes, what are the points, what is this game about not just in terms of genre but in terms of dramatic tensions and so forth.
• An outline–a list of the vignettes, roughly in order, with about a sentence or two describing each vignette.  If the game will branch at some point, so that some vignettes only happen based on certain choices, that would also go in the outline.  For example, our current outline has a couple of vignettes that are listed as “if X is true, then a vignette about blah” and “if they get to this point and make XYZ choice, the game ends; otherwise, they move on to the next vignette.”  The internal structure of the vignettes is typically not present at this stage, or it’s only alluded to–“an evil wizard sends a demon to try to kill you; you play cat-and-mouse with the wizard as she tries to trap you and you try to force her into a battle” might be the description of the Evil Wizard vignette from Choice of the Dragon, which ended up having a huge number of internal choices and complexity.  For games on the scale of Choice of the Dragon or Choice of Broadsides, there should be roughly 8 to 12 vignettes in the outline.  For a really long game, with maybe 30 or 40 vignettes, I would organize it into acts or some other form of structure above the vignette level.  That said, I would recommend against starting with a game that’s much longer than 8 or 10 vignettes.
• A dramatis personae–a list of all of the major characters besides the protagonist that we expect to show up.  That should include the “villains” or antagonists, but it also can include friends and allies, romantic interests, and rivals.  Most effective stories are heavily character driven.  We perceived the relative lack of well-developed characters in Choice of the Dragon as one of its greatest weaknesses.  By identifying some major characters upfront in Broadsides (Jones, Villeneuve), we were able to weave them through the story and make for a more satisfying set of conflicts and interactions.  I include references to characters from the character list in the outline, so the draft dramatis personae might list “your obnoxious older sister” and “the girl next door” as characters and one of the vignette descriptions in the outline might read “Your older sister tries to humiliate you and ruin your budding romance with the girl next door.”  Major characters should show up at least 3-4 times as important parts of vignettes.  As that implies, there really isn’t room for more than 5 or 6 major characters at the most in a 10ish vignette game, although there can be many more minor characters.
• A tentative list of global variables–the goal here is to make sure you have some handle on what sorts of things the player is going to be trading off.  Honor versus cleverness?  Intelligence versus strength?  Pragmatism versus idealism?  Romance versus ambition?  Variables can come in pairs, but they don’t need to, and it’s often better if they don’t.  If you define things strongly as pragmatism versus idealism, the temptation is to write lots of choices that are basically the same:  do you do the pragmatic thing or the idealistic thing?  If instead you have pragmatism, idealism, ambition, and honor as independent variables, then it becomes easier to see all of the different combinations:  sure, you can have the pragmatism versus idealism choice, but you can also have the idealism versus ambition choice, or the honor versus idealism (do you do what your duty ostensibly is, or do you do the thing you know is right?).  Also, don’t forget relationship variables–how well do you get along with the major characters?  Are they all still alive?  In our development, the initial variable list has never been totally accurate–we’ve always added variables as we write, and we’ve always designed in variables from the beginning that were either underutilized or cut entirely in the actual game.  But having a rough list up front makes things easier.
• A couple of vivid scene ideas for some of the vignettes in the outline to get the writing process rolling.  I find that having “that scene that I really want to write” helps me get started.  Once I’m started, it’s much easier to keep going, or to find the awesome in another vignette idea.
• Any worldbuilding stuff that I need for this game.  How does magic work?  What are the relevant nations in the political drama?  What is religion like?  For the most part, we’ve taken a very light approach to worldbuilding.  Dragon started with “a fantasy medieval world with dragons and goblins and humans and magic,” and that was pretty much it.  Broadsides started with “the Royal Navy of the United Kingdom circa 1800, in a war with a France analog, with the serial numbers filed off and archaic names in use (Albion for England, Gaul for France) and the option of gender flipping.”  But sometimes you need more.  For one of the games that’s currently in development, we knew that we wanted magic, but we had to spend some time figuring out what our magic system looks like.  We knew that we needed something analogous to the Reformation, but we didn’t want to use that directly, so we needed to come up with the social/religious conflicts that would fill that role without being the Catholic Church and the Protestants.  In general, I encourage people to do only the worldbuilding that is necessary to producing their games, or at least to be aware that while worldbuilding is fun, a lot of worldbuilding can be something that’s done for its own sake, not because it’s actually contributing to making the game.

Okay, so that’s what I want to end up with as we start writing.  How do we get there?

We typically start with a pretty strong genre choice–a concept, often embodied in a title.  At this point, we don’t have the elaborated one-sentence description of the game (which is really more of Dan’s thing than mine, anyway).  So we might say, “okay, this is a game about werewolves in the modern day” or “this is a James Bond/Mission: Impossible superspy adventure.”

Our next step is typically brainstorming a list of possible vignettes.  This isn’t an outline, yet.  This is just “here’s a bunch of ideas for awesome vignettes within that genre.”  So the werewolf vignette list might include:

• waking up in a strange place with no memory of how you got there and blood all over your tattered clothes
• battling another werewolf for dominance in the local pack
• dealing with a werewolf hunter who’s trying to track you down and shoot you with a high-powered rifle with silver bullets

The superspy vignette list might include:

• Infiltrating a foreign military base to find the secret files
• Assaulting the master villain’s underwater lair with a team of Navy SEALs
• Skiing down the Swiss Alps in a running gun battle on skis
• Seducing the villain’s daughter to get her to betray her mother (and maybe falling in love in the process)
• A totally awesome car chase in a tricked out sports cars through the narrow and winding streets of Rome
• Escaping from a horrible death trap

Some of the vignettes might be incompatible, and they might be in entirely the wrong order.  That’s fine–my goal at this stage is to come up with cool ideas, with vignettes that will be fun to write and play.  Also, note that while there are other characters in these vignette descriptions, those characters are totally undefined archetypes at this point–I haven’t decided that the villain has a daughter, I’ve just brainstormed the idea that a seduction scene could be cool and in genre.

One part of the brainstorming process is to make sure you have all of the tropes on the list.  So I might look over the superspy list and realize, “hey, the highstakes gambling game in a tuxedo is totally a trope that we’re missing,” so I add another vignette for that.  Anything that you would feel like “there totally needs to be X” belongs on the list.

At this point, we have an unwieldy mess of vignettes in no particular order–and typically way too many, like 15 or 20 vignettes for a 10ish vignette game.  We then start doing the work to make it coherent and more satisfying.  Dan likes to develop the one-sentence summary at this point, to pick the themes and core conflicts/issues in the game; that lets you make choices about what stays and what gets cut based on the overall design you’re aiming for.  I lean more towards starting to identify recurring characters and starting to build a dramatis personae, and starting to order and string together the possible vignettes into a coherent whole.  In any event, we start sorting the vignettes into a coherent plotline–ordering them and linking them together, while adjusting them so the characters match up nicely.  As the list becomes closer to the outline, you start asking questions like “do we like the pattern of dramatic tension that’s forming?  Is there enough space between when your CIA contact is introduced, when you begin relying on him like a brother, and when he betrays you (or, depending on stats, reveals that he was supposed to betray you but couldn’t bring himself to do it)?”  We rejigger the order to make the whole satisfying.  And we cut like mad.  This is the point where we start saying, “that vignette would be cool, but it doesn’t fit” and so we cut it.  In Broadsides, the Fleet Action and a vignette with fireships were both in the original brainstormed list of ideas.  They’re staples of the genre and would probably have been awesome vignettes.  But they didn’t fit the outline we were developing, so we cut them.

So now we’re most of the way there:  we have the one-sentence summary, the outline of vignettes, and a proto-dramatis personae.  So we fill in the rest of the dramatis personae, now drawing from a basically complete outline.  We then go back through the outline with an eye on the characters, making sure that they show up at the right times and get enough space to develop and become meaningful.  And we start sketching out the variables for the game, now that we have some sense of what the big picture choices are going to be.  This is also the point where we do any real worldbuilding that needs doing.  We’ll already have some implicit worldbuilding done, but here’s where we might say, “hey, we say that this triggers a religious conflict, but we don’t really have religion defined.  Let’s figure it out to some extent.”  Continue some editing, tweaking, and adjusting, and you end up with the prep products I described above:  a coherent summary, an ordered outline of vignettes, a dramatis personae, a tentative list of variables, any worldbuilding that we need, and a couple of vivid scenes ready to spark writing.  I didn’t call out the development of the vivid scenes, but for me that’s a natural part of brainstorming the vignette ideas–“I can totally picture this scene” or “I know I want the player to make a choice between this and that.”  And so then I pick a vignette to start with and begin writing.

Later, I’ll give some thoughts on how to pick a vignette to start writing with, and how to plan the structure within a vignette.  But this post is already long enough, so that will wait for another post.