Archive for November, 2006

Towards a theory of game narrative

Thursday, November 30th, 2006

How can we tell more compelling and moving stories in games? This is a question that I guess many of us in the industry are thinking a lot about. There are a number of books out there that tell us all about story arcs, characterization and so forth. Most of these ideas come from the theater and movie writing and there has been attempts to transfer this knowledge into the domain of games. And whereas the skills of characterization, dialog and story structure are as crucial in games as in movies or theater, there are problems and differences that need to be addressed. There are technical problems that arise with the addition to player interaction, multiple choices and so forth and there are content problems that arise with branching story lines. Some of those issues, I have written about in the past, trying to point to means of keeping consistency, controlling damage and hiding the seams in a game. This article tries a more basic take on the fundamentals of game narrative.

Storytellers and processors

Humans are exceptional storytellers and story processors. If you reflect a minute on how much story you are exposed to during a single day you will probably end up on a quite high number. We read books, we watch TV-shows and movies. And we play games. In addition to that we tell a lot of stories, both to each other and to ourselves. Each time we tell a piece of gossip or an anecdote we create that information into a story, and we do that without really thinking about it. Also, our lives turn into stories in our heads as we live along (this is called the narrative paradigm). So everything we do and experience turn into stories, most of which never leave our head. Many clinicians even claim that creating a coherent story of a traumatic event and incorporating this into the self-representation is crucial for the successful treatment of post-traumatic stress disorder. So storytelling is not only fundamental in our social behavior but probably also very important for our well-being.

So we are all a grand pack of storytellers that consume huge amounts of stories. Now, my knowledge in cognitive and neuropsychology is extremely limited, but I wonder what the differences are between the stories we consume and those we produce, and the reason I wonder is related to game narrative. It seems reasonable to believe that one function of our consumption of stories is to enhance our understanding of the world and the people around us. We can call those “external” stories since they are about other people. We also produce stories to project that understanding to other people (by retelling gossip and such), but we also produce stories as a means of self-reflection, to bring structure and sense to our lives. Those we can call “internal” stories because they are about ourselves. Interestingly, some people turn their internal stories into biographies and those turn into external stories when they reach the reader.

Internal game narrative

I have quite distinct memories of the game The Star of Africa that me and my brother and sisters played when we were kids. In this simple board game, the players are explorers who search the continent of Africa for precious gems; the ultimate goal is to find the huge African Star diamond and take it back to Cairo or Tangiers. I remember being sucked into the world of that game and it was thrilling to envision oneself traveling between countries. So as we played the game, I created a story about myself traveling in Africa.

“The Star of Africa” is by no means the only board game that have triggered my imagination and caused me to start making stories as I played along. It might even be that children playing such simple games as Ludo create stories as they go along – I really don’t know. What I do know though, is that when you play at least certain board games, stories are created in the minds of the players. Now, it seems reasonable to believe that playing a video game is not very different. When we run around in “Ratchet & Clank”, “Halo” or “Zelda”, I believe we produce internal stories out of the setting, the game play events and our actions.

External game narrative

When we think of game stories, we probably like to think of them quite similarly to stories in other media. Game stories contain the same elements as movies and books: a protagonist, a conflict and hopefully a few good characters that develop during the game. But it’s not just similarities. The most prominent difference is usually the scale of the story, which in games differ quite much from the ones in a typical movie. Game stories are longer and in some respects more complex than a two hour flick, which can be attributed to game play time.

The usual narrative tools available to tell stories in games are almost the same as in movies. Cinematic cutscenes are probably the most common, but games frequently bring the story into in-game events and conversations with NPC characters as well. Sometimes NPC communication becomes part of the game mechanics (see “Fahrenheit” for example, where dialog is time limited). But either way, the narrative in a game is almost always written by a writer and can thus be seen as a purely external story.

Now the big questions: what are the differences and similarities between internal and external stories? Can the two of them co-exist within the same frame? And is it possible for a player to merge the internal story she creates while exploring “Gears of War” and the one that is told to her in the cinematic cut-scenes?

The gap between internal and external narrative

I believe that there is a gap between the internal stories we make and the external ones we consume, and I believe that the boundaries between them are sensitive. I have a distinct feeling that the Link I’m controlling in “Zelda” is not quite the same Link as the one who appear in the cinematic sequences. I know in my head that they are the same character. He has the same name, the same clothes and moves the same – and he probably even smells the same. But I intuitively know that Link’s emotions are not the same when I control him as when the game controls him. And that feeling is quite close at hand, even though we try ignore it. When I control Link, his feelings and experiences are synonymous with mine. He is just a “shell” that I step into. I am the entity that takes possession of his body. In the narrative sequences something or someone else is taking my place and instantly the boundaries between the internal and external stories are exposed.

Now, even if the boundaries are exposed (as they are in most narrative games), perhaps there are means of eliminating the problem. Perhaps humans are perfectly capable of dealing with multiple “modes” of narrative within the same frame and they only need the right cues so they can join their experiences into one powerful narrative. It is perfectly possible that the problem is just a matter of learning – that players are becoming increasingly better at merging internal and external stories – and that it is just me who sucks at it and feels bad when my gaming experience is disrupted. But until someone finds out how these things really work in the minds of players, we can only guess. It would be possible to do experimental research and gather evidence that could help us draw better conclusions. Perhaps this information and research already exists, but I haven’t seen it.

To expose the gap – or not to?

During the production of “Riddick”, we tried to follow a narrative design where the Riddick who appeared in cutscenes did not act by himself. We failed at a few places in the game, but nevertheless I believe that the idea was sound: the game should not mess with the character on its own since the player is in control over him. At the time, I did not think of this in terms of internal or external stories, but the idea made sense anyhow. However, where the philosophy we followed in “Riddick” (and continues in “The Darkness”) is about player choices, the more generic idea of internal and external stories has even deeper consequences since it is about player narrative experience. Even if the design of “Riddick” tries to allow the player to stay “in character” mechanically, I believe the boundaries between the internal and external stories are exposed constantly. For example, we put Riddick on the screen in situations where the player was not in control (i.e. he figured in cinematics) and we put words in Riddick’s mouth in dialog sequences.

Could we then avoid exposing the gap altogether? I think so. In “Ratchet & Clank”, there are some cinematic sequences that begin with a sign saying “Meanwhile…” In those cinematics, the player character (Ratchet) is not present. We could have redesigned the narrative flow in “Riddick” in the same way by showing cinematics that explain the upcoming challenges. However, there is a problem with this approach. We deliberately wanted to keep Riddick in the camera view at all times in the game. The reason was that the player and the on-screen Riddick character should share the same knowledge. A “meanwhile…” sequence would give the player information that the on-screen character would not have. There are ways around this, of course.

Chances are that by now you do understand the problems I am talking about, but either think that I am wrong or that I am nitpicking. Who cares about all this phycho nonsense? Shouldn’t game stories be all about fun and exitement and first class cinematic experiences?

Yes, they should! I think it is important not to get lost in philosophical arguments about the player’s role and experience of narrative – at least if you’re into creating games. Also, I would trade a perfect example of problem free, but dull narrative against a vivid, fun and exciting game full of narrative flaws any day! But we are in the process of discovering narrative in a medium that we do not fully understand. Neither do we fully understand how the medium affects us and how we affect its affection to us! Just as other media has developed conventions and rules (the idea of the fourth wall, for example) over the years, we need rules and conventions that are efficient. Tailor-made tools that allow us to focus on the fun and thrilling aspects of narrative: story, structure, characterization and dialog. If we know that the stories we create land in one whole package in the minds and hearts of our gamers – our stories will rock them!

Bears of war

Monday, November 20th, 2006

Starbreeze concept artist and friend Mattias Snygg has written a review of Gears of War that really gets to the primal core of the game – sort of. I’m sure he’s gonna piss the heck out of a number of readers, but that’s what usually happens when you rip the veil from the eyes of the innocent and they get to squeeze the muscles of truth.

Now I’m eagerly waiting for “Dead or Alive: Arnold in Lace”. Thanks Mattias!

The split personality of Fahrenheit (Indigo Prophecy)

Wednesday, November 15th, 2006

The game Fahrenheit is old news, I know. I played through the demo when the game was released and never got to playing through the full game. However, the last few days I have been home with the flu, lying in my couch unable to work. “What an excellent opportunity to get through one of those unplayed games,” I thought and dropped the Fahrenheit disc into my Playstation, took a sip of hot tea and grabbed the controller.

First of all I want to recommend the game to those of you who haven’t played it. It’s a well made, well told adventure story game (the developers prefer to call is “interactive film”). I really enjoyed it and was truly impressed with many parts and aspects. However, most good things have flaws – especially games. First of all, I was a bit put off with how the story unfolded. From the beginning, the storyline is tight and interesting. It moves on a personal level bringing me as a player close to the characters. Later on, however, it explodes into grandeur and “way too many elements and twists”. I think the turning point is somewhere around the awkward “Silence of the Lambs” reference. I still wonder why game stories seems to be destined to become overloaded with elements and themes. I mean, serial killers, the Mayan culture, secret orders, strange artifacts, secret military research, Armageddon and AI. Come on! Furthermore, I was not very keen on the “self awareness” game references. One of the characters repeatedly states that “it feels like I’m in a videogame” and it just kills the mood for me. This is a shame because Fahrenheit has shitloads of great mood and excellent story moments.

Story from Multiple Views

The game is told in a non-linear fashion with story branches here and there. Most of them seem to lead back onto the main track, but there is enough back-references to the choices you make to keep the illusion alive. However, one aspect of the non-linearity puzzled me.

Through the game, you alter between the various main characters of the game. The method of character switching is nothing new (see “Eternal Darkness”, for instance) but Fahrenheit does a thing that seemed very strange at first.

I first noticed it during a scene where Tyler Miles (one of the main characters) is questioning the waitress from the first murder scene. I moved through the dialog choices and was suddenly presented with an option to give her answer as well. Through the game, this happens a few times: the player is playing both sides of a conversation at the same time.

I thought about this and wondered what the design intention could have been.

Either the game is just moving its point of view rapidly, or something far more elaborate is going on. If the former is the intention, the designer might not have thought it was a difference between shifting character between chapters (as is the normal case in the game) and shifting characters arbitrarily. There are a few other sequences where the player is able to shift character in the middle of an action sequence, but in those cases it is by player choice. It is also clear in those sequences that you are playing two characters “at the same time”, so to speak.

The Tyler-Waitress scene it is different though. The player is given no hint that the game is shifting focus and is just presented with a “Yes” and “No” dialog option. It is up to the player to realize that the option does not belong with Tyler but with the waitress. This, as I though about it, was a quite serious design flaw.

However, as I continued to think about it, a more interesting conclusion appeared. I figured that perhaps the game is doing something much more intricate than just shifting focus?

Propelling the Storyline

“What if the one driving the story of the game is me?”

It might not seem that this question is getting us anywhere. Isn’t it so that the player in any game is driving the story? Well, in a way that is true. However, most games seem to give the player a somewhat different role. The fuction of the player is partially to drive the story forwards, but the role is different. In Fahrenheit, I initially felt that my role was that of Lucas Kane. Then perspective shifted and my role was that of Lucas Kane, Carla Valenti and Tyler Miles – alternating. Then more views were added and suddenly I lost my sense of having any role in the game – other than driving the story forwards. And there’s the key! Being the story-propellor can in itself be a player role. Quite an interesting realization. Suddenly I was sitting in the lap of the narrator, turning the pages. I was not part of the story – I was playing it like a toy.

Now, I don’t really believe that the designers of Fahrenheit intended that the player should have this role. If it indeed was a deliberate design choice I’d have to say it was a clever one, although the implementation and realization of the idea haltered. It could have been much clearer and stronger. If the player is given a story to toy with, I reckon there should be more options and buttons to push – but that’s just me. I’d also love to try out a scenario and then be able to backtrack and investigate a different route without disrupting the flow of the game (I can do this by selecting to replay chapters from the main menu, but it’s too clumsy).

All in all I think that Quantic Dream did a fantastic job with Fahrenheit. It is a great game that provides a fresh and bold take on narrative in games. It is not without flaws but contains enough interesting gameplay and storytelling to be a source of inspiration and learning for game designers and writers alike.

Death rewinds the time

Thursday, November 9th, 2006

Imagine that you are reading a book. You are a few dozen pages into the story when suddenly the book flips back a couple of pages and you are forced to re-read the last chapter.

Imagine watching a movie in a theatre. It is a tense thriller and you are totally immersed in the story. Suddenly the movie stops, rewinds and you have to see the last fifteen minutes again.

Rewinding Stories

Now imagine this happening all the time, with all the books you read and films you see. A couple of times an hour, the story rewinds by itself and you will have to get through that part again. A bummer, right? Bigtime. When you sit down to enjoy a story, you are expecting it to be presented continuously. Sure, there is a difference between books and films. I sometimes have to step back in a book and re-read passages that I kind of “slumped” through. So book reading is perhaps not that a continuous process. Still, we would probably be extremely annoyed if we tied the “rewinding” of books and films to a set of rules which caused them to happen at somewhat regular intervals. Say, for example, that someone installed a loudspeaker in your home. Whenever that loudspeaker sounded (with a beep, for instance), you would have to clap your hands ten times before the next sound beeps three seconds later. The beep may go off at anytime when you are reading or watching a movie, and the penalty for not clapping your hands ten times within that three second timeframe is a rewind of the current story. Five pages if it’s a book, ten minutes if it is a film you are enjoying.

Does it sound like much fun? Nah, not really. But now let’s look at how “rewinding storylines” relates to dying in videogames.

Dying in Videogames

Games that tell stories come in many different flavors and I have previously written (here and here) about different approaches to game storytelling. So, let’s look at how dying in a game affects the storytelling. The extreme book and film analogy given in the beginning of this article may be contrived, but it shows an interesting property of stories: When we experience a story, we internally fit that story to a timeline; and whatever drives the narrative (a game or a DVD for instance) cannot tamper with that timeline without consequences. For example, we expect movies to tell the events that will fit our internal timeline sequencially (in some cases we get it in pieces that we along the way fit together in our heads. “Memento” is a good example). If we push rewind on the DVD remote control, the timeline is instantly disrupted and the experience is ruined. Try for yourself. You will most certainly lose immersion and it will take some time to get back into the story.

Dying in many games disrupts the timeline in just this way. The reason for having the player character vulnerable and lethal is obvious – it makes the game challenging which is a good thing. However, when designing a game that kills players that are not good enough, you have to realize that the storyline is in danger. One common design is that upon death, the player is thrown back a bit on the timeline and he/she has to progress through the same part of the story timeline again. This has effect on the immersion and the experience of the story content. Suddenly an in-game event, a dialog with an NPC or cinematic becomes a mechanical obstacle, something you most probably want to get past as soon as possible. Personally, I tend to ignore the content of the storytelling if I am exposed to it a second, third or fourth time.

A game that follows this design is “Tomb Raider: Legends”. Before some of the boss fights there is a cinematic sequence that is part of the story. However, if you die during the battle, you have to watch that sequence again. In that particular game, it would probably be possible to make those cinematics cleanly separated from the gameplay and add checkpoints around them to prevent the problem. However, many games suffer from this problem.

Another approach is to alter the save system so that instead of rewinding the story timeline, the game sends the player back to a previous location, but the story progress is still intact. “Jak & Daxter” is such a game. This solution is not without problems though. The player is sent back physically, but the timeline is still moving on so the player has to run through a level that is “cleaned” up to the point of the latest death. This is confusing since it says two different things at the same time. First, the timeline is intact, secondly it cannot be since you are moved back in the world instantly.

Checkpoint-wrapping the Story

It might seem that the optimal solution is to do the “Tomb Raider: Legends” thing and just add checkpoints that solves the repeating cinematics problem. But I strongly believe that games should try to tell the story in-game. Without cinematics you tend to resolve to in-game events, scenarios and NPC conversations for storytelling. One can wrap all such elements in checkpoints, but the problem is still there – at least partially. In a well designed in-game event or scenario, the player can die in the middle of the event. You can of course bring back the player after a death to the very spot and time where he died (to prevent repeating story), but that is in effect the same as saying that the player is unable to die. The challenge of staying alive is removed since there is no penalty for dying. To solve that, you could introduce another kind of penalty. You could, for example, remove pieces of the inventory or revive the player with limited health, but that is a truly horrifying way to go. Suddenly you will run into a scenario where players can get truly stuck. The QA department will kill you for it and real life is not wrapped in checkpoints, remember?

So, do we have any other options?

You cannot Die!

The most drastic solution is to prevent the player from dying altogether. This was done very successfully in “Myst” and there are other games that take a similar approach. However, the true “you cannot die” solution seems to be limited to adventure games where the challenge is not to stay alive, but to solve the next puzzle. So if we are making an action game, we have to come up with something better.

“Grand Theft Auto” has a semi “cannot die” solution where the player respawns outside the hospital. You lose some of your progress and have to start over again with current mission, which means that time is turned back nevertheless and story elements are repeated.

Damage Control

“Prince of Persia” provides an interesting take on the problem. Whenever the player dies, a narrator voice says “But this is not how it happened…” and time is rewound to a previous location. So that game makes the problem part of the narrative which is quite clever. They even go as far as making the rewinding of time a gameplay element. However, the narrator voice quickly becomes tiresome and the storyline is equally disrupted and immersion breaks nevertheless.

There are most certainly other ways of doing damage control to cover the problems and I wouldn’t be surprised if there are examples of games that does it with better results than “Prince of Persia”, but I believe that the problem is rooted deep in a discrepancy between gameplay (interactivity) and how we experience and relate to stories. Perhaps the optimal solution is not to design gameplay mechanics that serves the storyline, but to alter the way the game narrates.

The Game World as the Story

One game with a fundamentally different approach to story is “Metroid Prime”. The game features cinematics that provide the setup for the game, and that provides presentation and closure at certain points in the game (before a boss-battle, for example). But those cinematics have little to do with the narrative. Instead, it is the game world that provides the story.

In “Metroid Prime”, the player can dig deep into the story of the planet the game is set on by investigating the environments and by “scanning” certain objects to collect pieces of story information. This part of the game is purely optional, but it deepens the world and the interested player can use those pieces of information and build a story timeline for him- or herself.

There are other games that take a similar approach (“ICO” shares some of those traits, but has more of a storyline than “Metroid Prime”) and to me, those games make a more profound story experience since I, as a player, is unfolding the story from within the world. But while those games can avoid the rewinding timeline altogether they have problems telling the same kinds of stories as games like “Tomb Raider: Legends”. Whereas most games that contain a story tries hard to be a movie, games such as “Metroid” makes no such effort. In effect, this means that the cinematic game uses the language of the cinema to convey story. Cinematic events and dramatic conflict pile up in the player’s face and it works – sort of. But shifting the perspective and allowing the player to unfold, experience and build the story for him- or herself has profound effects on how the story can be presented. The story of “Metroid Prime”, for example, is “passive” in the sence that you experience the story elements second hand – through the pieces of information you gather. Very little “active” storytelling goes of in that game, whereas in “Tomb Raider: Legends”, almost all the storytelling is active, and told through cinematics.

The Essence of the Medium

So, which is the technique most true to the medium? The question is largely irrelevant since the storytelling potential of games makes anything that people enjoy and get rewarded by true and worthwhile. There are innumerable ways of telling a story – just look at the many ways books and movies tell theirs. Still, I believe that we need to understand our medium in order to be able to go further and evolve. Cinema was nothing more than “filmed theatre” in the beginning and one might argue that most games today are little more than “pieces of cinema stitched together by passages of gameplay”. But as cinema has moved way beyond filmed theatre, I believe that game storytelling will move way beyond cinema

An interesting thought experiment is to look at various media and search for story-containing works that could not ever work in another medium. In literature, a candidate might be “Finnegan’s Wake”, by James Joyce and in cinema “Koyaanisqatsi”, by Godfrey Reggio. (One could argue that the same “essential story” can be told in another medium, but the effect on the audience will probably be very different.)

Now, do you know a game that tells a story that cannot be told in another medium?

Once again on the move…

Sunday, November 5th, 2006

So it was time to move the page to a new server. Unfortunately I was more or less forced to do this (the bastards on my old server charges way to much) and an unpleasant side-effect was that my own content-management system (written in Perl with a SQLite backend) does not work – at all. So it’s either rewrite or charge up a ready-made CMS. I tried MovableType before and absolutely detested it. This time I’m gonna try Textpattern. It looks kind of nice and has Textile built in right from the start (I used it for my system) so I hope I won’t choke…

Still, there is a lot of work to be done. I need to move all the old material over here and do some visual customization so this place will probably be a mess for some time now. We are currently crunching “The Darkness” too which means I don’t have much spare time to spend on such worldly matters as the ol’ homepage. So bear with me.