A Practical Guide to Game Writing

By Darby McDevitt

[In this detailed Gamasutra feature, veteran game writer McDevitt (Assassin's Creed: Bloodlines, Where the Wild Things Are) outlines useful processes for collaboration between design, production and writing staff, from pre-production through production of a game.]

Video game writers are a frequently misunderstood sort. Even in the most ideal situations, we are often relegated to the status of mortar to the designers' bricks, slipping between the cracks to paste fun moments of gameplay together with a few lines of snappy, expository dialog.

Writers can be further marginalized by a lingering sense among our team members that we want nothing more than to stuff our games full of melodramatic, Metal Gear-sized cutscenes, burdened by a cast of dozens sputtering flowery lines from our 450 page script.

I'd like to steer us clear of this idea, one likely sustained by the apparent misconception that writing is fundamentally about arranging words into meaningful strings.

Clearly this isn't the case, but somehow a large contingent of the game industry has institutionalized this attitude anyway, and its effects can be found in an upsetting number of games released in the past few decades.

Just count the uneasy puns and strained moralizing spilling from your favorite avatar's mouth -- when a writer is hired to write a game, and is subsequently barred from having input into its pacing, its setting, the motivations of its characters, and its mood and tone, writers resort to the only weapons they have left: wry witticisms and declarative pop-philosophy.

The spirit of collaboration games are supposed to embody often seems well outside the writer's reach.

But the truth is, we don't want to hijack your game with pointless soliloquies, and we don't want to write a posturing Hollywood-style epic. Game writers simply want to help designers craft an immersive, interactive narrative experience. With or without dialog, with or without characters, we simply want the game to start somewhere interesting, climb its way over a few emotional peaks, and end somewhere even more interesting. We're good at that sort of thing too.

Not all games require a narrative arc, of course, but it's a rather common feature of quite a few mainstream console titles, and these days if an actual writer is going to pen the script of one of these games -- as opposed to the lead designer or the producer -- some Very Important People probably have a Very High Opinion of the property.

But this doesn't happen as frequently as you might think. Consider yourself blessed if you have actually seen a game writer in the wild, for they remain one of those elusive, added-expense luxuries that many game producers -- their eyes always on their margins -- believe they can do without. And in many cases, it humbles me to say, they're right.

The average game-playing public will suffer a deluge of poor storytelling if a game is knock-down, drag-out fun. But a great story with terrible gameplay will die a fast and lonely death on the shelf. I respect and support this pecking order. Gameplay must come first -- this is the golden rule.

However, if some form of narrative happens to play a design-critical role in your proposed game, it is vitally important to treat it exactly as you would any other design element, not as a separate discipline. So if your team has taken that bold extra step to build a narrative-driven game, there are a number of precautions you can take to accommodate the writer and prevent the story (and your writer) from getting buried beneath endless revisions of your GDD.

First and foremost among these is to make one simple conceptual change: treat your writer as an associate designer. Involve her in the design process from the outset. Even if she is not an experienced technical designer, a good writer can be instrumental in helping inspire unique moment-to-moment experiences that provide gameplay variety while integrating seamlessly into the narrative. Again, writing is not just about clever sentences -- it can also be about narrative shape, motivation, and pacing, i.e. what you do, why you do, and when you do.

Most of my favorite narrative-driven games contain very little dialog in them at all -- Ico, Shadow of the Colossus, Flashback, Out of this World -- but even these titles are "written" in the sense that they have a clear set of emotional shifts, tonal changes, and meaningful moment-to-moment events that compound into emotional pay-offs.

When writers and designers band together and discuss a game's story, characters, dramatic set-pieces, and settings in parallel with ideas about the game mechanics and levels, the team will begin to find exciting and creative ways of conjoining the two disciplines into a more unified experience.

Unfortunately, this synergy can be difficult to find, especially in the trenches of third-party development where the average dev cycle is less than a year. When schedules are tight, producers and designers often maintain a slight distance from writers, imagining we are off "doing our thing" while they do theirs.

But our thing is their thing too. Writing is design. We are both building a world from scratch, after all. So if you empower a writer to absorb and occasionally contribute design ideas, she will carry on with a solid understanding of how the narrative elements contribute to (or detract from) the overall game experience.


Before Writing Begins

For a clearer picture of this process, let's pinpoint a game writer's most critical early-milestone tasks, beginning with a few pre-production goals. In these early weeks, it's easy to get drunk on a thousand and one ephemeral ideas -- by all means do -- but you'll need to conclude this reverie with a few tangible results:

High Level Narrative Summary. During preproduction, the design team should work directly with the writer to concoct a brief (one to four pages) high-level summary of the primary story. Think of it as your elevator pitch: make it succinct and snappy. This short piece is probably the only story document most of the team will ever read, so it should be clear and compelling. Do this early, and get the client to sign off on it as soon as humanly possible. Read that previous sentence again. Get quick client sign-off every step of the way. Failure to guide your client to a swift agreement on the story may result in endless misery for the remainder of the project.

Major Locations / Levels. This is one area where writers can really get sore if they are left out of the conception process. Designers frequently forge ahead with level concepts and designs without consulting the writer, not taking into account the huge role that setting plays in crafting an interesting narrative.

In video games, place is often more important than character, so this is doubly important. If the writer, designers, and artists band together to nail down the scope of the game's environments, and get a rough idea of how much is needed and how much is feasible, everyone will walk away happy.

This cuts all ways: writers need to know that they'll have the locations they need to tell a good story, while the artists and designers will want to make sure the writer is asking for content that is relevant to gameplay.

Obviously this "relevance threshold" varies with the size of the project, but on small projects with short schedules getting this right can mean the difference between environment artists going home at 6 pm or 6 am the next morning.

Once production begins, the writers work ramps up. This is the point where the entire design team needs to function as a single unstoppable force (for good):

A Detailed Story Outline. With the narrative arc complete, it's time to produce an exquisitely detailed story document, complete with scene descriptions and gameplay objectives. The amount of detail in this doc will vary according to how much the story influences the design, but it should be as thorough as possible. In any case, generating a detailed outline will give you an early understanding of just what sort of game you're making, and how reliant on the writer you will be for design iteration down the road.

In the case of heavily plot-driven games, the design challenges will stem directly from the story -- e.g. rescue a prisoner, assassinate a guard, courier a package. For non-linear games like RPGs this document should be incredibly dense and detailed. For less structured games, the writer's direct impact on the design may be minimal. Understanding this balance ahead of time is critical.

Story Presentation Plan. How, exactly, is the game's story being told, and who is responsible for telling it? Do you have pre-rendered cutscenes or in-engine cutscenes? Who will be putting these scenes together? Perhaps you have no cutscenes whatsoever, and would like to tell your story on-the-fly. Is this feasible? Possible?

Figure it out early. Nothing is more frustrating for a writer than seeing a project scoot forward without anyone having a firm understanding about how the story will be told, since this will affect what she intends to write.

Estimated Cut-Scene Breakdown. If your game does contain cutscenes or animated in-game sequences of any kind, it is crucial to estimate their number very early on to get a good sense of the work to come. If you have a detailed story outline, this should be easy. On tight projects it also helps to determine ahead of time what the expected intricacy and quality of each scene is so your teams can allocate their resources appropriately.

Characters. As you generate your detailed story arc, you'll need to make a clear list of the number of characters needed. Who are these people, and what roles do they play in both the narrative and the gameplay? Which are simple NPCs? Which are robust, interactive characters? Which are bosses? Mission givers? Shop keepers? Tutorial mentors? Et cetera.

The artists will be generating all character models and animations, and they'll want to know the scope as soon as possible. If you spring 15 new NPCs on your artists halfway through the project, they will shank you in the break room -- believe it. Getting the character scope nailed down early will also help you determine how much "incidental dialog" the game will require, for these throwaway lines frequently take up as much space in the script as the main story dialog. This is no trivial amount, so keep close track of it.

Sort Out Your Text Database. This can be a tedious task, but it is crucial to sort out your text pipeline very early, and get your tools up and running. The longer you wait, the more you will hate yourself. Some games have complex or esoteric text requirements -- non-linear conversation systems, for instance -- so it is critical that you organize your data cleanly and clearly.

Also, take a moment to decide how the script will be delivered. Not all writers are familiar with the esoteric architecture of your text database, so if your writer is delivering the script in Word or Final Draft, you're going to need a pipeline to handle its transfer.


When Writing Begins

Once your game's foundation has been laid and the team is ready to start production, the actual writing can begin. This is the fun part. Writers love to write, but without constant contact with the design team, they run the risk of giving you more script than you need, or a script you don't need at all. This wastes everyone's time and makes the writer sad when you have to tell him, "as beautiful as they are, your 100 part limerick-cycle has no place in Chaz Dastard's Intergalactic Star Safari 2: Misremembered Legacy".

Nip your writer's graphomania in the bud by establishing clear boundaries. This should be simple if the writer has been involved in the design from the beginning, since all parties involved will understand the extent of the game's writing needs. Keep track of everything before it needs to be written, as it is being written, and after it has been written. A game writer without defined boundaries or direction -- especially an off-site, contracted writer -- runs the risk of writing something as sensible and useful to your game as Andre Breton's Soluble Fish.

Script, First Draft. Between the greenlight and first milestone, the writer should be busy as hell. On short projects, ideally she should have a finished first draft of the script by the first production milestone, as this will help the level design process move smoothly.

On longer projects, the writer and level designers will be working back and forth quite a bit to make sure neither one lets a detail slip, edging ever closer to a first draft.

Demand Story and Script Sign-Off, Again. Be crystal clear with your client: the script needs to be read and comments forthcoming as soon as possible. Of all the client-side headaches I have ever encountered, this is the most painful.

Many clients make the mistake of believing the script is the single most important aspect of their game, and therefore spend months and months poring over details that contribute very little to the final game experience. Delays of this sort can hold up level designers and cutscene artists in the most asinine ways imaginable, wasting time that cannot be easily recovered.

One little discussed benefit to hiring an experienced writer is the fact that, relative to coders and artists, good writers work incredibly fast. Text is cheap and takes very little time to edit and revise. But this advantage is of no use to anyone if writers aren't aware that anything needs revising.

I have lost count off the number of times a seemingly innocuous level design change or map layout has rendered a chunk of my dialog obsolete. When I have not been made aware of this chance, the resulting headache cannot be cured by earthly medicine.

Darby: Listen to this gem, guys: "Sally forth to yonder Black Forest, stalwart Wayfarer, for there you shall find a crystal dagger of such rare-"

Producer Person: Ah, Darby, sorry... the Black Forest was scrapped and replaced by a Walmart. We should have told you.

Darby: Ah... okay, hold on. Where's my pen?

Woe betide the team that discovers this incongruity only after the actors have recorded all of their dialog. Again, keep the writers and designers partnered at all times.

Into Production

Now you're well into production, and the heavy lifting has begun. If you have nailed all the earlier tasks, the rest of production should proceed smoothly, barring any client interference. This is supposed to happen only if you've been naughty, but the unfortunate truth is not so black and white. There are more than a few imposing clients out there who, for understandable if not always sensible reasons, believe the story can be endlessly revised up until Beta. So be wary, keep calm, and carry on.

At some point during production, the script will be finished and the writer will feel like she is nearing the finish line far before the rest of the team. Don't let this illusion persist. There is still a bit of work your writer can help you with:

Casting. If you are recording with actors (and who isn't, these days?) now is the time to figure out who will be making your characters speak. On small projects that don't have an official story director, the writer can be of immense help. It's crucial to get your casting done well in advance of your recording date. Actors have hectic schedules and you'll find all the best ones rather busy if you try to snag a few the week of your recording session.

Final Script. As difficult as it is, the writer will have to stop tweaking her dialog and settle on something. Of course, it's a good idea to encourage the writer to streamline what she can. The script may be laden with timely wit and wisdom, but it is still, above all else, a game script and if it tests a player's patience, that can be a problem. More to the point: the longer the script, the more time it will take the cinematics team to craft the cutscenes or scripted sequences. So when the writer buckles down and kills her darlings early, it keeps everyone from doing superfluous work.

Voice-Over Recording Sessions. Some writers make great VO directors; some don't. But all good ones should be able to re-write their dialog on-the-fly, so make sure your scrivener is available for the recording sessions. When she hears her dialog spoken aloud for the first time, she's probably going to want to change it. Allow some leeway, but don't let her get carried away. Try to limit changes only to what is egregious or erroneous.

Once you hit Alpha, the writer's job gets a lot easier. But there are still a number of good reasons to keep one around, locked in a cabinet somewhere, just in case.

Proofreading. Writers should never copyedit and proofread their own work, it's true. This is a fact that holds doubly true in the game industry where the volume of text written is often comparable to that of a novel. On the other hand, it's rare to find excellent proofreaders hiding in the QA department, so make sure as many eyes are on the text as possible, including the writer's.

Non-Dialog Text Revisions. It can take a long time to nail down all that tutorial, database, and menu text your game has accrued slowly but consistently over the span of the production. Lucky for you, text is cheap to implement and fix, and is quite safe to alter even up to the last minute (provided you're still proofreading).

And with that, your writer's job is finished and your game is nearly complete. Well done, folks. Take a breath and clean your white board. The whole process starts again in five... four... three... two... one...

Return to the full version of this article
Copyright © UBM Tech, All rights reserved