Writing for Games by Allen Stroud

Know the Task

“Be like a reed and bend in the wind.”

Some games companies prioritise the fiction writing and organise the other creative elements – like art, or level design, or rules – to follow on from the initial drafts of the story. Others begin with the art and ask writers to contextualise. It is important you are comfortable with the role you’ve been asked to perform and that you recognise you might not get a say, or your ideas might not be taken on board in the final release. One piece of advice I would suggest to anyone who wants to write for games is to make sure you know the brief you’ve been hired to work to.

Often, this type of work means the company you are working for owns the content you produce.

I gave the opening quote to another writer who was working on a project with me. He spent far too much time arguing for his ideas, not realising that he had been heard and rejected. He couldn’t move on because he had come to the project as a fan. The excitement of a fan being asked to contribute to a new version of a game can be really useful, but when the fannish qualities get in the way of doing a professional job, then you have a problem.

Starting Out

In winter 2012, I was browsing Kickstarter for an example project to show my filmmaking students. At the time, I was trying to introduce them to crowdfunding as a possible option for their work, raising funds for short films and small projects. I was also making the point to them that a large number of individuals were launching crowdfunding projects and might need the support of people who could shoot good footage and edit it.

While searching for a project to discuss in class, I found the Kickstarter for Elite Dangerous, which was eventually released in 2014.

This project was trying to raise money for the launch of a sequel to a game I had loved as a child. I cast my mind back to my experiences of Elite (1984) and Frontier: Elite 2 (1993). I had played them for hours. They had been an escape into another world that had allowed me to imagine what it might be like out there in space. I followed the crowdfunding campaign through its last days, pledging my support and finding I was not alone. Thousands of fans had come aboard and were sharing their experiences of the previous games. The last days were halcyon as we could all see the project would be successful.

One of the offered ‘rewards’ from the project was to write a piece of official fiction set in the game universe. A diverse collection of writers, both experienced and inexperienced, had backed sufficiently to achieve these rewards, myself amongst them, with a plan to write and publish Elite: Lave Revolution (2014). When the dust had settled, I contacted Frontier Developments and offered my services. My masters degree had involved the design of worlds in fantasy and science fiction. I thought I might be able to help the company sketch-out information for the writers so they could create fictions that would be consistent with the game environment. Producer Michael Brookes (who recently passed away in Summer 2023) replied and invited me onboard to help and so my journey into the stars began.

Worldbuilding for Games

In Deconstructing the Starships (1999), Gwyneth Jones wrote that, ‘one thing science fiction and fantasy certainly have in common is the imaginary world, a world that must be furnished with landscape, climate, cosmology, flora and fauna, human or otherwise self-aware population, culture and dialogue’.

Elite Dangerous is a sandbox game – which means you as the player get places in a self-contained environment to do as you want, within the game’s playing parameters. In this case, you start with a little ship to fly and can make your way in the galaxy, trading, or fighting, or mining. Frontier Developments intended to simulate a colonised Milky Way Galaxy thirteen hundred years in the future. The background to that idea includes a technique known as procedural generation.

Procedural generation is a programming technique that allows small teams to create huge playing environments. Instead of designing the locations, programmers create libraries of content. The computer then selects from those libraries, either as a piece of (mostly) random generation or to a defined pattern.

The difficulty for the designers lies in making each generated space station, colony and individual come across as part of an environment that has depth and character. This is a real challenge. If the libraries of a procedurally generated game are detailed enough, they can create incredible things, far beyond the imagination of those who designed them. They can also create rubbish, if the parameters aren’t specified in enough detail (the procedure), or a repetitive set of environments if there are not enough alternatives to choose from in the library of content. This latter issue is where a writer has to work hard, creating individual character for a space station, an asteroid base, or a planet. The environment of Elite Dangerous is procedurally generated. My task was to give it character – to make objects and people appear unique and have a story. So, my first work in the games industry was to create a history, mapping out the timeline from AD 2013 to 3300. There was a lot to draw from (the previously published material from the previous three games) and also a lot to invent.

The games imagine three political factions, the Federation (a commercial bureaucracy), the Empire (a dictatorship) and the Alliance (a vast democracy of star systems). They co-exist uneasily and compete for resources/overlap historically/have little contact.

After organising the sequence of events, I wrote them up, like a historical record, but with a bias to favour the game’s oldest faction – the Federation. I then wrote them up again, this time from the perspective of the Empire, and then again, from the perspective of the Alliance. In total, I wrote six guidebooks of 10-20,000 words in less than three months.

These books and a set of additional content were edited and tweaked by Frontier Developments so that they fitted the studio’s vision for the game. Once they were happy, the documents were then shared with official fiction writers and the other departments of the studio so they could be used to develop the frontline content – the writing players would see in game or in the tie-in fiction.

There are a few terms for this kind of document. Some people call it a ‘plot bible’, others a ‘world canon’. I call it a macrotext, to remove any religious implications and distinguish it from experimental writer and literary critic, Cristina Rose Brooke’s megatext, which describes the general assumptive archetypal iconography of a genre. Writing in the second half of the twentieth century, Rose identified themes in different types of writing that build on each other. So, for example, readers know what vampires are when they are mentioned in in a book, the writer can shortcut a little if they want, relying on the reader’s knowledge (megatext) of other stories, etc.

But this is different to a macrotext. A good macrotext goes beyond plot, is often bigger than a world and flexes with each publication, allowing itself to be updated and altered as needed, rather than restricting the imagination of the writer.

Once you have the main structure, what can also be important is what you don’t include. That applies to all the sections of this article, and to writing in general. But we’ll come back to that particular point at the end.

Characters for Games

As with many other writing projects, when writing the background and invented history of a game, a writer creates a framework for what will come after his/her involvement. If the game is released with other material, then often that material will support the story of the game itself. You may find yourself writing a set of stories leading up to the beginning of the game, or to flesh out of specific in-game scenes.

Sometimes, while you are putting together the background, things leap out at you. When you have big questions, like what caused magic to return to the world? Or how did the war between nations start? As writers, we write about people and sometimes, characters step off the page and into your imagination. The best non-fiction books about history always gave me that impression, so I seek to emulate them. Great histories bring to life the long dead: people like the Earl of Warwick – known as the ‘Kingmaker’ in the Wars of the Roses, or Hino Kunimitsu of 14th century Japan.

You can find individuals in every nation of the world who, according to the records, have seized control of events. Their actions and choices have had consequences that reverberate through the ages.

In a computer game, these kind of characters exist as part of the story. But for the most part, stories in a game revolve around the player character – the personification of the player in the game. That might be someone they create at the beginning, or a predesigned character. Designers often make a choice on how much they engage with the player character.

For example, in Elite Dangerous, there is very little initial engagement with the player character, but the game is designed to react to the player’s actions. Performing illegal acts, like piracy gives the player character a criminal status. This in turn, changes the way space stations and other spaceships interact with them. All of those lines delivered to the player character have to be written in a way that feels authentic.

Other games take very different approaches. In Baldur’s Gate 3, the character generation start by giving the player the option of playing with a pre-generated character. This character has a detailed backstory that is already incorporated in the game. Or the player can generate their own character. If they generate their own, some of the cut scenes still have lines delivered by the player character and a writer has to think about the tone of those lines, so they can work for any type of player character.

In Phoenix Point, I created the historical character, Randolph Symes III, as a fictional archivist who had compiled information about the Phoenix Project and the dilemma facing the player – the Pandoravirus crisis (yes, from 2016-2019 I spent my time writing fiction for a game about a global pandemic). Symes is an example of a character who stepped off the page and ended up as a narrator of a collection of video and audio scenes that the player collects to discover the truth about the situation the player character has found themselves in. The player character finds Symes’ work and research into finding a solution for the Pandoravirus and alien incursion.

However, there is no specific backstory for the player character, other than that he is the hero of the moment, tasked with trying to save humanity. They could be anyone, which is part of the point. This allows the player to think about what may have happened to them before beginning the game.

Working this way, the player character is engaged with discovering what happened through Symes’s diaries and recollections.

In Chaos Reborn, I wrote a series of short stories that introduced the player character and that describe events that happen just before the game starts. That character remains nameless and is written about in the first person. This approach created a hand over of the character to the player, but the stories give the initial context.

This is similar to a visual introduction cut scene, like you see in Baldur’s Gate 3, or games like Bioshock (which is itself worth viewing as it’s one of the most well written games ever made).

Plot and Narrative for Games

Within any game, there will be a method by which the story is imparted. The narrative process goes through a number of iterations. Sometimes you it might get it be given to you as a writer, as a brief summary, which you then turn into a specific script. Sometimes it might be a monologue or written with visuals, the information being communicated in either the dialogue or images. When you get involved, depends upon the specifics of your role in the project.

There are two key differences with writing plot and narrative for games as opposed to other media. The first is about screen real estate. Games need writers to do more with less. Compression of information is essential. Getting across all the necessary information in as few words, or in a specific number of words is important. Words take up the screen, or if delivered through audio, take up time. In both instances, those words might be delaying the player from playing the game as they absorb the information.

The second difference about writing plot for games is redundancy. Often games are about choices, empowering players to go left or go right. When you bring interactivity into a story, you begin to create redundancy. You are writing content that might remain unseen by the player. A writer has to accept that in this medium and still try to make everything as good as it can be.

Both of these factors change how you write. It is important to be aware of them and use them as tools in the toolbox when you need to. Plot and narrative can be anywhere in a game that has words and images. In fact, the more narrative that is imparted through a variety of different systems within the game, the better. It makes the story feel part of the experience, which is essential if it is going to be integral. Research trees are a great place to sprinkle a little plot, cards in a boardgame can be a good place, as are special secret briefs for players in a roleplaying game. Every image and piece of text are an opportunity to tell and enhance the story and draw the player into the intriguing twists and turns of the plot.

Alternatively, you may find yourself writing for a game that doesn’t take the opportunity to think like that as it follows a puzzle – cut scene process. That format is frustrating and ironic. Players are being rewarded for their participation with a non-participatory briefing, so beat the level, watch a cutscene, rinse and repeat. In this kind of game it is better to think of other ways to empower the player’s participation.

Epsen J. Aarseth published Cybertexts: Perspectives on Ergodic Literature in 1997. The book explains the concept of ergodic literature – a text that requires more than trivial effort to read or understand. Mirror writing is an example of this, or code, or any type of writing that means the reader has to undertake a specific action to understand the text. This provides the experience of the text with a sense of achievement and worth when it is read.

For those unfamiliar with writing interactive stories, Twine is a good place to start (http://www.twinery.org).. Twine allows you to start creating interactive stories. It publishes to a single HTML page, so you can play them through and get used to the elements of redundancy and compression.

Last Thoughts

My experience over may years is that games are a great medium to work in and whilst most of the focus in this article has been on computer games, many of the techniques translate into other forms. Writing for games makes you aware of what you are doing in a different way and provides you with some ways to experiment.

This, in turn, can apply to your other fiction as well.

I also mentioned what you choose not to include in your writing for games. This is an advanced skill. Knowing to leave questions and mysteries in a text that intrigues but doesn’t frustrate. The ultimate goal is to spark the reader’s imagination so that the experience stays with them after they’ve put away the source books, board and dice, or packed up the games console. Achieving that is what can bring them back, again and again. In fact, one amazing way of doing this is to…

That’s all for now.