Vital Game Narrative: A Conversation With Rhianna Pratchett

British-based narrative designer and scriptwriter Rhianna Pratchett has quickly become one of the few widely-known names in game writing. The Overlord series, for which she provides the dialogue and directs the voice work, has become known as one of the few genuinely humorous series in contemporary gaming. She has also written for the more adventurous and thematically challenging Mirror's Edge and Heavenly Sword.

Game writing is frequently maligned -- and frequently, understandably so. But there's a high level of complexity to the issue of getting good story into games. Here, Pratchett  -- who is the daughter of famed Discworld fantasy novelist Terry Pratchett -- delves into that complexity and examines the barriers that are currently erected; she also offers her thoughts on improving the craft throughout all levels of the game creation process.

On the projects you've worked on so far, how much of a driving force for the overall project vision and metafiction (i.e. backstory, setting, etc.) have you been? How important is that to you?

Rhianna Pratchett: It really varies from project to project, depending on when I'm brought on board and what state the game is in at the time. The Overlord series is probably where I've had most influence in this regard, as I've written, voice directed and helped craft the stories for four and a half titles over the last three years.

These include Overlord, the expansion Raising Hell and Overlord II with Triumph Studios in the Netherlands and Overlord: Dark Legend and Overlord: Minions with Climax Studios in the UK. It's hard to remember a time when I wasn't working on an Overlord game! It's an opportunity that not too many writers get in this industry. I've been very lucky.

I also worked on a great deal of the background and metafiction of Mirror's Edge. However, because it took DICE a while to find a writer, by the time I joined the team all the levels had been designed without any specific narrative in mind. This meant that there wasn't a huge amount of room for delivering that metafiction within the core game.

Certainly working on the Mirror's Edge comic series with DC (which was a prequel story to the game) helped me flesh out the world a lot more. During the series I could focus on the background of the characters and how the situation that the player finds themselves in actually came about. It was a great creative outlet and I'm sure we'll see more games venturing into the comic space.

It's very odd to hear that the story in Mirror's Edge was decided after the levels were built, given how strongly creative the game is and how the story is a big part of its overall image. Can you explain the process more? How much did DICE have when you were brought in, and how much freedom did you get?

RP: It's actually not that uncommon for writers (particularly freelance ones) to be brought on to a project where that is the case, or partly so. If you're not in near the start, then what often happens is that you're presented with is a task akin to writing a script for a film for which all the sets have already been built. It's a tricky to deal with, but certainly not impossible. In fact it makes the art of writing far more like an actual craft; where you're weaving, chiseling, sculpting and molding the story around and through the levels.

Sometimes you find yourself in the rather odd position of writing a story backwards. You're presented with the place the character needs to be and you have to create the path which took them there. It can be quite a fun challenge in a sort of masochistic way.

In Mirror's Edge, I think that nailing down the core gameplay mechanics (particularly as they were so innovative) was the primary focus, as was creating the right kind of levels to show them off. As I mentioned, it had taken DICE a while to find a writer (which in itself isn't always an easy business) and the team couldn't exactly sit around and wait for one.

I believe the game had been in development for a couple of years before I joined the project. At that point all the levels had been designed with no narrative through-line (again, understandable and not uncommon) and there were visuals for a few of the characters, such as Faith, Miller and Ropeburn.

So my task was really finding that through-line, filling in the gaps, building flesh and bones and breathing life. Games writing and narrative design is all a bit Frankenstein-esque. You gradually source and pull together the limbs of your story until it forms a whole body. Then you sew it up and send it out there... feverishly hoping it doesn't savage any small children along the way.

As far as freedom goes, there was a fair amount in terms of helping define the narrative vision that supported the world (which was done mainly by the Senior Producer, Owen O'Brien and myself) but it was within the tight boundaries of the level design. I'm still proud of the world we created¸ even if some of it got rather lost in translation.

How much background info is it essential that the writer create that doesn't (necessarily) fit into the final game -- things like character back stories and biographies, descriptions of organizations, cities, whatever. What and how much of that kind of stuff is essential?

RP: It's vital for a writer to establish all aspects of a game's narrative setup. Even if the audience only gets to see the tip of the iceberg, you still need to have a full understanding of everything under the surface that supports it. This gives your stories and characters much more of that all important truth and less of a feeling that they've been plucked out of the air and exist in a vacuum.

Biographies help you get to know your character; peel back the layers and find out who they really are. Defining things like background, goals, relationships, attitudes, flaws, traits and even language is not only essential for the writer, but also extremely helpful for animators, casting directors and voice actors.

Likewise, knowing the background story can be incredibly handy for level designers and artists, especially when they come to define a visual meaning for the world and a sense of place and purpose. For Mirror's Edge I even ended up creating lists of products used in the world and advertising slogans. Games writing can occasionally lead to crazy levels of diversity.

The Overlord series is notable for its humorous approach. There's not a lot of (intentional) humor in games. What do you see as the barrier?

RP: I think the barrier comes about from not realizing that for humor in games to work well it needs to have a multilayered approach. For example, in Overlord the core gameplay and level design are inherently twisted and fun. I mean, you're an evil overlord with an army of sycophantic minions that rampage around the countryside, looting and pillaging at your command. What's not to love about that?

Plus you're also seeing the world from the baddy's point of view, which isn't one represented that often, making it inherently attractive. Look at how well something like Dexter has done (okay so he's sort of a bad guy and sort of isn't.) On top of that, we've always tried to make sure that the script and the gameplay are tightly woven and we use great voice actors who really "get" the spirit of the titles.

That's not to say being funny in the game-space (or in any space) is easy. Especially given that as a writer you don't really have complete control over the usual comedy tools, such as timing and context. Therefore you have to spend a lot of time working in-depth with designers at a micro-level to be able to compensate for that. It's quite an undertaking.

The games that I've personally found funny have all got this multilayered approach at heart. Particularly titles like Psychonauts, Destroy All Humans and Tales of Monkey Island (all the Monkey Island games, in fact.)

Certainly, working on the Overlord titles has made me realize how much gamers appreciate playing a game that makes them smile or laugh. Be it through a funny animation, or a line or a piece of gameplay. I don't want to get all Patch Adams about this, but comedy in general is often undervalued, yet it can be extremely powerful.

Overlord

The genre question: do different genres of game, in your opinion, lend themselves better to storytelling or is it all in the approach the designers take?

RP: Considering that you can tell a story in six words, as Hemingway illustrated with "For sale: baby shoes, never worn", I firmly believe that most genres can become powerful storytelling vehicles. Largely it is down to the narrative sensibilities of the team and how seriously narrative is taken within the game-space.

Obviously RPGs and adventure games lend themselves well to traditional and more linear story telling methods, as do the slightly slower paced action-adventure games. But it's important to keep in mind that there's no "one size fits all" here. Not even close to it. The way you tell a story in an FPS needs to be very different to the way you tell a story in an RTS or a platformer.

I've come to the conclusion (through trial and error, mainly) that the faster-paced the action is in the game the more you have to plan the narrative in advance and properly structure it into the level design. If you don't then it's much more likely that the pace and delivery of your story will fall completely out of line with your gameplay.

Unfortunately, it's often these types of games where the narrative is the last element to be addressed. But the longer it's left on the back burner, the narrower the options for story delivery become. Often you can be left with some very linear (and usually very expensive) options.

What do you think of the necessity of overlap between design and writing -- i.e. do you feel a tight integration for true gameplay-based storytelling is required, BioWare approaches it?

RP: Writers definitely need to be more integrated into the development process, whether that's working in-house or on a freelance basis. I've always had the best results when I've worked closely with the designers. I think the industry has generally accepted that having professional storytellers onboard is a good thing.

But working out what to do with them once you've got them is proving to be much more of an ongoing challenge. There are still a lot of misconceptions about writing, particularly in regards to how long it takes. I've certainly come across the underlying assumption that crafting a story should require the length of time it takes to flap your hands at a keyboard and just write words. Any words.

I don't believe that it's essential for writers (particularly if they're freelance) to be present in a studio all the time in order to produce good results. Technology has made the world a much smaller place and things like instant messengers, Skype, SVN, Fogbugz and wikis are a godsend. It's mainly a question of developing clear and consistent communications and resource sharing channels.

Besides, most of us need to dive back into our own personal writing burrow and roll around in muse juice, from time to time.

How closely do you work with teams -- do you turn in scripts, or do you go to studios and work in-house? Can you contrast the different approaches you've taken on different projects?

RP: I actually loathe the whole "first draft, second draft, third draft and you're done" approach. It seems to be a Hollywood hangover. Writing is rewriting, and nowhere has that been more the case than with games. It's a hugely iterative process. You have to be flexible and roll with the rest of the development cycle.

From personal experience, I do believe that it's also extremely beneficial for the writer to be included in the recording process. Not only do they know the context of every single line (or should) but they can provide an essential level of context and character depth that actors thrive on.

I've often provided studio support on my games and I was the lead voice director on Overlord I and Overlord II (assisted by the fantastic Dan Gardner and Tim Bartlett of the Audio Guys.) It was great to be involved at that level and it's something that more writers should push for -- at the very least in a support capacity.

I think as a director I'm probably pretty demanding because I've got so much invested in the lines and characters. Plus, I've usually been up until the wee small hours prepping the scripts and casting every last minor character... so picky and over-caffeinated!

These days I find myself collecting voice actors. Marc Silk (who plays Gnarl in the Overlord games) has been a fantastic find as well as a good friend, whilst Jules de Jongh not only voiced Faith in Mirror's Edge but also put her vocal talents behind Lil' Red and Doris in Overlord: Dark Legend as well as Mistress Juno in Overlord II.

Ultimately, when you're recording something like the Overlord games it can be tremendously good fun. Marc, Dan and I have even developed our own directing language. "Gravel Police" = "Rasp your voice more", "Less windmill" = "Make it less cute", "More Scooby" = "Vary the pitch", and "They've got a loaf on" = "That person isn't happy" (mainly based on Dan's fondness for bread-making.) When you've been working together for three years, this is just the kind of stuff that happens.

How collaborative have your scripts been with the teams, and how collaborative do you want the process to be?

RP: The Overlord games, both with Triumph and with Climax, have been pretty collaborative and I worked very closely with the level and audio designers. This meant that the needs of the narrative and the needs of the gameplay could coexist in relative harmony. It was a similar deal on Heavenly Sword, although as it was an extremely narrative-led game, the levels were mainly shaped around the story, rather than the other way around.

This hasn't always been the case on other projects. Sometimes I haven't even got to see the game I'm working on and I'm pretty sure that most members of the team couldn't pick me out of a line-up. I think this is symptomatic of the industry-wide problem I mentioned earlier; namely that companies don't often know how to fit a writer or narrative designer into a team and therefore have a tendency to keep them at arm's length.

However, I think that narrative professionals have to take the lead in breaking themselves out of that bubble and pushing for better integration. Everyone is learning, here. We're all on the same side.

At what point in production do you come into the picture typically? At which point do you wish to become involved?

RP: The short answer is "the sooner the better." But it doesn't always happen that way. A few times I've been lucky enough to be brought in between one year to 14 months before the project ships. But even then, it can still be not early enough, especially if all the levels have been designed. If I can't be there right at the start, then I like to get involved about six months into a project. That's usually when the core mechanics are in place and there are a few level ideas to work from, but the structure isn't set in stone.

As importantly, at what point in production do you exit the process, and again, at which point would you prefer to do so?

RP: It's not quite a "from my cold dead hands" deal, but I try to make myself available right up to the last full text lock and often beyond. After the main script is written, recorded and in place I'll usually move on to addressing any non-VO quest text, system text, chapter names, additions to the manual, marketing copy and basically pretty much anything that involves words.

It's also important for the writer to be involved if any narrative surgery has to be performed on a project. No one relishes chopping fingers or even whole arms off of their narrative babies, but I firmly believe that the writer is the best person to wield the scalpel and stitch up the wounds. No one needs a story bleeding all over the place, least of all the person who created it. Regrettably I haven't always been given the opportunity; sometimes the industry can be pretty brutal towards narrative. You learn from it and move on.

A lot of complaint is made about cinematic experiences -- i.e. cutscenes -- in storytelling but some developers continuously show them to be effective. What's your take?

RP: There's no denying that given the fan-base of games like the Metal Gear Solid and Final Fantasy series, many gamers enjoy cutscenes, even incredibly loquacious and lengthy ones.

Whilst, personally, I'd rather a game wasn't turned into a wannabe movie, I believe there's still a place for artfully crafted, well timed and smartly paced cutscenes. Granted, the games that manage to do all three are fairly rare. Putting interactivity aside for a moment, there's still a lot we can do to improve our linear storytelling. There are exceptions (there always are) but our strength in this regard is by no means across the board. It is improving though, title by title.

Cutscenes are still an important tool in our narrative toolbox, and we shouldn't throw out the hammer just because we keep hitting our thumb with it. We just have to learn how to wield it a little better.

Cutscenes vs. direct during-gameplay storytelling -- is one more or less effective in your eyes? Can they coexist in the same game?

RP: They can coexist in the same game just fine. Most of the titles I've worked on have used a blend of narrative delivery techniques. The Overlord games use a lot of on-the-fly ambient and directional dialogue, as well as cutscenes.

However, more interactive cutscenes or, what I'd personally like to see, more context for limited/locked view points (like being frozen in ice in BioShock's Fort Frolic, or being held on the metal Citadel transport pods in Half-Life 2) is eminently desirable. But there are two problems inherent in that (actually there are probably loads, but these two shout the loudest to me.)

The first is, as I mentioned earlier, that interactive narrative has to be supported by a game's core level design structure. It can't just be slotted in. Developers need to adopt the mindset of thinking about narrative right at the start of a project. I think we're still a little way off from that.

The second issue is that there's simply no one-size-fits-all solution to this. Whilst BioShock, Portal and Half-Life 2 made undeniable progress in game storytelling, the interactive elements were composed in relatively small, closed-off and controlled spaces (again, with level design playing a large part.) This would certainly be hard to replicate in something like a large, open world RPG with lots of exterior locations, or a traditional strategy/adventure game.

I'm not denying that these are important steps, but they're still quite small ones, and not an instant and all-encompassing solution to the interactive versus non-interactive debate. However, I do think the ways in which the aforementioned games showcased the power of visual storytelling, in particular, has something to teach the industry as a whole.

What techniques have you taken from storytelling in other media? What can't you take?

RP: Well I guess we're back to the Frankenstein metaphor again. It really is a little bit of this and a little bit of that. The central components of good, structure, plotting and characterization are essential, but given the nature of the medium you need to constantly revise your techniques to fit with the titles you're dealing with.

I'd say that games share more in common with writing for TV than they do with writing for film. Each involves turning scenes and bringing out characters in a relatively small space with limited resources. At the same time I think there's a kinship between writing for games and writing for the stage. Particularly in the way that audiences engage with characters. The difference being that in games you're pulled out of your seat and straight through the fourth wall.

For me one of the most useful things to keep in mind when creating stories (and particularly characters) is the F. Scott Fitzgerald quote "action is character." Whilst this is important for all storytelling mediums, I think it takes on a unique significance with games where action is such a tangible and often very visceral component.

Often games seem to have an odd disconnect between the player character and the action which they're performing. So I often try to start with the central premises of the gameplay and work out what kind of character would be engaging in such activities, where would they have come from, what might have happened to them and what impact would that have on their mental state.

Mirror's Edge

For example, in Heavenly Sword, how would being a skilled (and somewhat bloodthirsty) warrior impact on Nariko's ability to be a normal, emotional human being? What would push Faith (Mirror's Edge) to throw her body through so much physical exertion and yet appear to have disengaging her empathy for what's going on around her?

Do you feel there's an inexorable march towards in-game storytelling or more tightly integrated storytelling in the medium?

RP: I don't think we'll be able to do the former until we really master the latter. And believe me, there'll be a lot of hard graft in doing that. But it has to happen. We have to get our hands dirty.

Apart from anything else, there's so much that can be done to just make the stories themselves more engaging and original, before we attempt to remove our training wheels completely. I love this industry with a fiery passion, but even I wish from time to time that developers would just step away from Hollywood's '80s action movie scrap heap and be a little bit braver with direction, theme and content.

I think using the player as a storytelling vehicle is something that will become very important to games narrative in the future. Right now it's unique to our medium, very powerful, and yet we've only just begun to scratch the surface.

The best game stories have always been created with a sense of partnership within the "white space" of the game world; the writer and the player(s). As the ever-eloquent Margaret Robertson puts it "It's not about putting more emotion in games; it's about putting more emotion in players."

Controlling and creating effective game/storytelling pacing is extremely difficult. What can you contribute from your end when working with the development team? What have you learned about the process?

RP: Games writers can be a complainy-pants bunch, sometimes, we really can, but there's a lot that we can do to make the process smoother from our side. Regular communication is at the heart of it and understanding that it's a two way street and everyone is learning from each other.

I know I've banged on about it, but integration really is the key. If you can work one-on-one with a level designer then together you can create a narrative that works across the board. This is much more desirable than forcing them to deal with a script that they've had no input into. If a designer feels that the script is working for the gameplay rather than against it, they are much more likely to protect it in the instances when a writer may not be able to. Like a sort of Narrative Deputy.

Games are collaborative and writers are often external contractors rather than team members. How does that affect the writer's control over the story's overall vision and the ability to craft compelling stories, given those constraints?

RP: It's probably very telling that the words "writer's control over the story's overall vision" seem a little strange and unfamiliar to me. On most projects writers are pretty far down the ladder, whether they're internal or external. The same could be said of other entertainment mediums, as well (aside from the more successful US TV writers) although sometimes it feels like the games industry has added on a few more bottom rungs, just for us. See? I said we were a complainy-pants bunch!

Ultimately, you are working as part of a team, so as a contracted writer or narrative designer you're not going to be the one making the final decisions. Having specific people responsible for the narrative of the game (and nothing else) is still fairly newish.

So unsurprisingly, the power that writers get is usually pretty limited (and often less than wider assumptions may suggest) simply because we're still feeling our way through the narrative-versus-gameplay minefield. So really it boils down to a question of input, rather than control.

Once again this really harks back to the way in which writers are integrated into the team and how seriously narrative is regarded. Things are getting better, but writers are still often underused and poorly managed. When it comes to games writing, it's not just how good you are, but how good you're allowed to be.

Moving to specifics a bit, the character of Kai in Heavenly Sword is unconventional and eccentric -- a lot of that is conveyed through her animation and character design as well as her dialogue. Can you talk about that interplay?

RP: Kai's visuals came first and then it was all about breathing life and motion into her avatar. We wanted her to be quite feline and playful in her movements as a contrast to the heavy brutality going on around her and her subsequent detachment from it all. That's also a sense that she may be cavorting with, or even speaking directly to, something that can only be seen by her.

It's always tricky when you're dealing with a younger character (although Kai's mental state is a fair bit younger than her physical state) that they don't become annoying. There's a fine line between cute and weird and just plain irritating. I think it actually helped us that we didn't use a child actor to play her. Given that she's actually pretty violent, it could have been... complicated.

Kai was probably the most challenging role in the game and consequently I spent quite a while talking to Lydia Basksh (the actress who both voiced and acted Kai) about the character, her past and her journey during the game.

Lydia was able to capture Kai's layers brilliantly; her resilience, determination to hold onto lost innocence and her sheer devotion to her adopted sister, Nariko. I've always maintained that in some ways Heavenly Sword is a love story. It's just not a love story about a boy and girl, but one about sibling love.

Heavenly Sword

And can you talk about the storytelling functions of characters that are left-of center in game stories?

RP: What I think worked well for Heavenly Sword was that from a narrative point of view, we didn't waste characters. We had a small cast but they were all tightly wound into each other's lives. One of the themes of the game was about the sometimes screwed-up nature of familial relationships. Initially it was demonstrated through Nariko's relationship with Shen, as both daughter/father and student/teacher, and her bond with Kai.

It's then reflected and distorted in Bohan's volatile (and equally problematic) relationship with his son Roach and the childish machinations of his generals, Whiptail and Flying Fox. In Whiptail's case she is instrumental in sending the relationship between Nariko and Shen spinning out of control, tearing the two characters apart, whilst Flying Fox is a predominant player in Kai's story.

I think NPCs (although Kai was a lot more than that) can be vitally important for highlighting story themes and important traits in both protagonists and antagonists. They really are the narrative pillars of a game world.

You talk about having a small but relevant cast in Heavenly Sword. Ken Levine spoke about chopping out many NPCs in BioShock and consolidating the roles of several into one character. But many games have sprawling casts of NPCs -- many disposable or essentially interchangeable. What do you think about approaching the issue of character in games?

RP: It's a difficult one. Primarily because the function an NPC provides can vary from game to game, genre to genre. Nevertheless, good characters linger. Players latch onto them much more than developers perhaps realize, and therefore much more attention needs to be paid to their creation.

In all honesty I haven't been able to do everything I wanted with every single character I've worked on. You don't always get that kind of freedom as a contractor, and quite often you still have to fight tooth and nail for it when you're on the inside. However, the ones I think that have been most successful (in terms of player feedback) have had strong themes resonating right through to their personality traits and quirks. Almost to the extent of being somewhat larger than life -- there's not always room for much character subtlety in the game-space.

In the case of Heavenly Sword, I think that allowing the player to actually participate in several of the characters' emotional journeys helped create that elusive player-to-character bond. Too often characters can become, as you say, interchangeable and nothing more than talking parts of the level design.

Really, the solution is to properly anchor your characters in the game world and elevate them from whatever role they may be performing as a gameplay signpost or walking tutorial etc. Be it to illustrate a facet of the world, a relationship dynamic, a mirror on the protagonist or whatever. Your characters are the pillars of your story, so make each one count.