The Fifth Annual Game Design Think Tank Project Horseshoe 2010 |
![]() |
Group Report: In Search of Better Narrative in Games |
Participants: A.K.A. "The Mechanical Turks" | |
Adrien Banet-Rivet, Ubisoft |
Squirrel Eiserloh, TrueThought LLC |
Link Hughes, CCP Games | Katherine Isbister, NYU Polytech |
Facilitator: Linda Law, Project Horseshoe | |
![]() |
|
A Foolhardy Fishing Trip
We ignored them, and struck out from the shoreline anyway. In many ways, our predecessors were right. The waters of our discussions were choppy and the turbulent winds of our brainstorming threatened more than once to overturn our craft. In the end, however, we did draw forth a few insights... which, unsurprisingly, led to further questions. That said, these new questions were interesting and informative. To the best of our knowledge, they may not have been fished up previously. So, in the spirit of providing stepping stones for further inquiry, and with the hope that you might wish to embark on your own foolhardy fishing trip, we will share our catch with you. Three Fishing HolesAt first the going was hard and navigation was confusing. In seeking better definitions and direction for our quest, we identified three “fishing holes” – broad, categorical, contrasting approaches to narrative in games. These provided a foundational language by whose light we were able to finally make some headway. 1. The “Find Meaning” Approach – A “Closed” Narrative StructureA closed narrative structure is a detailed and meticulously crafted story featuring strongly defined characters and, quite often, a linear plot. Concentrated bursts of narrative such as lengthy cut-scenes featuring dramatic encounters and extensive dialogue are commonplace. The game mechanics may be entirely focused around pacing the delivery of the narrative, and may or may not contain much in the way of inherent fun when divorced from the story they serve. The player consuming a closed narrative is encouraged to immerse fully in the experience and wholly suspend her disbelief. Primary characters and the setting around them undergo transformations intended to create a bond with the player, facilitating the experience of altered emotional states. Plot twists, cliffhangers, and other surprising developments draw the player on to the game’s climax and resolution – and possibly sequels – in much the same fashion as a well-written novel or a cherished television series. The designer creating a closed narrative crafts the characters and a complete narrative for the player to experience and identify with. Setting and gameplay activities are used to reinforce this. All aspects of the game must invest the player and get her closer to the story. In the end, however, it is the player’s perspective on the experience and her interpretation of that story which bring forth meaning. Archetypal examples of closed narrative games: Silent Hill 2, Final Fantasy X, Alan Wake 2. The “Choose Meaning” Approach – A “Filtered” Narrative StructureA filtered narrative structure provides a framework to players that they can use and make their own, to some degree—this kind of narrative structure typically involves an intriguing world and evocative narrative framework and plot, but without providing detailed and explicitly ordered storylines among well-defined characters with elaborate backstories. Sometimes these sorts of games spring from the narrative worlds of other media, and make use of the evocative and enjoyable aspects of the particular narrative world to create an engaging play space of opportunities for players. In these kinds of games, the narrative is really a kind of referee for what sorts of things are acceptable/plausible/viable for players to do, framing up potential actions and likely consequences. In this sense, there is more structure in the game born from the narrative than in the ‘Create Meaning’ approach. A player consuming a game of this type is encouraged to use the narrative framework to generate stories with friends and/or for herself, through action in play. It can be fun and challenging for players (and sometimes the stuff of legend) when they craft narrative grounded in quests or activities for themselves, above and beyond what the game makes obvious. In so doing, they deepen their own narrative experience and, when they share their stories, enhance the experience of other players. The designer creating a filtered narrative tries to offer enough structure, story, and setting to allow for immersion, with plenty of room for embellishment and player ownership – and even evolution – of the narrative. It’s a balance between providing enough structure to generate player loyalty and emotion, while providing enough space so that people want to put their hearts into playing out and reading in new aspects of the story. Archetypal examples of filtered narrative games: EverQuest, Echo Bazaar, Fable 2 3. The “Create Meaning” Approach – An “Open” Narrative StructureAn open narrative structure contains very little in the way of authored narrative. Rather, the story rises from the complex and often surprising behaviors created by the player’s interaction with dynamic systems and game mechanics. Learning the game and experimenting with its rules leads to unforeseen discoveries, which can lend a sense of progression and story arc to a game experience where none has been explicitly mapped out. Usually open narrative games are themed on top of familiar motifs – such as classic fantasy monsters or modern family living – which provide a context for players crafting their own story around the game mechanics, while simultaneously making the mechanics intuitive and accessible (e.g. everyone knows that you need two people to make a baby, or that wood is flammable). For a sense of story to develop in an open narrative game, the player must often directly engage with the system/mechanics for an extended period of time – usually over several play sessions. The player is frequently given the ability to create structures or other lasting works within the game. The changes wrought on these creations as they persist within the environment of the game mechanics can be one of the most affecting aspects of an open narrative.
In some open narrative games, the designer will impose typically simple constraints or goals onto the system to provide a core tension to motivate the player’s free action within the space (e.g. “build an amusement park that earns at least $X a month”). In others, the game system provides natural tensions which serve to drive the player’s actions (e.g. beware of the zombies that come out at night). Some completely open play systems, such as Lego-type building blocks, impose no tensions whatsoever and leave the player to invent her own goals. Archetypal examples of open narrative games: Minecraft, The Sims, Burnout Paradise A Complicated Terrain of Technique and ReceptionPlumbing the depths of these three fishing holes, we noticed that different members of our intrepid band responded differently to each narrative category. The implication was that none of these categories were inherently more valuable than the others. Being a fairly small group, however, we couldn’t reasonably draw sweeping conclusions based solely on our own small cross-section of reactions. So we conducted a survey. We asked a single multi-part essay question: “What’s your favorite story in a game ever (either authored OR emergent)? What was it like to experience that story? Why was it so great?” Querying twenty-one of the industry’s brightest and most collaborative design thinkers, the range of answers we received was astonishing. Some excerpts include:
Within the diverse pool of responses, there were anecdotes that described favorite stories which lay within each of the three narrative structures we had identified. From a respondent who had appreciated a closed narrative: “…it was subtle. It trusted me to figure it out, and it allowed me to interpret its images rather than beat me over the head with them. The power came from quiet creepiness, and it made me disturbed for the [whole of] play. The model of the creatures, the solitude of the town, the strange reactions of the NPC, and the unexplained horror were all part of it.” Another, describing the merits of an enjoyed filtered narrative: “Having certain parts of the narrative defined, and other parts remain undefined, afforded me the ideal opportunity to inject my own meaning into the story, giving me a sense of personal authorship. Had they defined everything for me, there would have been no room for me to inject myself into [the world and narrative].” And another, which expressed the virtues of an open narrative: “The reason it was amazing was that the story was 100% systemic; it grew unintentionally from the Sims’ systemic behavior… It was, in fact, the way that the system actions (without designer intention) triggered meaning and understanding in my mind. The “a ha” clicks in my head that led to an “understanding” of these “characters” were marvelous.” Our conclusion was confirmed: players do not respond uniformly to narrative techniques or structures in games. One insight, then, that we fished from the deep was this: there is no grand unified theory of game narrative. Attempting to forward one really only answers a kind of Rorschach indicating a theorist’s own game narrative preferences. What is needed, rather, is research which begins to map out the complicated terrain that lies between narrative techniques and player reception. With a better informed cartography of this broad topic, future seekers will have a far better basis for making grounded and useful conclusions. Avatar Suits and Imaginary FriendsWhile we had neither the time nor the resources to really begin exploring the connections between technique and reception, we did hit upon another insight that provided an example of the discoveries lying in wait for future investigators. Specifically, we discovered something peculiar about the mixed reception of players to the great whipping boy of game narrative: the cut-scene. Though they can be rich with narrative content, some players loath them; and while they interrupt the flow of normal gameplay, some players love them. Why is this? In discussing cut-scenes, we noticed that our four reactions to them were quite mixed. One of our number loved them, considering the thirty-minute cut scenes of Metal Gear Solid 4 and Xenosaga Episode 1 to be high points of those game experiences. Another member of our party, however, reviled them, denouncing their usurpation of control from the player. The other two of us were somewhat negatively disposed towards them, but had enjoyed them on occasion. Going back and forth over this disparity in experience, we eventually identified that we perceived our connection to the characters in the games we played very differently. Three of us (and presumably many other players) viewed strongly defined, external characters – of the sort that populate games rife with cut-scenes – as a kind of avatar “suit.” The player’s character comprised a worldview, context, and set of abilities that the player got to inhabit and walk around in. When a cut-scene interrupted this experience, it felt to these inhabiting players as though the suit had suddenly taken action without the player’s ability to participate, which was quite frustrating. When the player is inhabiting Lara Croft, running full tilt toward a train that is about to depart, it is aggravating when, having reached the edge of the railway platform, the game suddenly shifts to a cut-scene in which Lara makes the jump – just barely – and manages to pull herself up onto the train. The player wearing the “Lara suit” could have done that himself and it would have been great! Why did the game take that away from him?
Indeed, the sense of the character as an external, autonomous entity went very deep for this player. Even though, fundamentally, a strongly defined character would not move about the world without the player’s direct intervention, he didn’t perceive it that way when playing games like these. The character, Alan Wake, for example, was a separate entity and so, when this player pressed up on the joystick, he saw it as “helping” Alan navigate rather than doing it for him. This type of relating player was his own mechanical Turk in these types of games. He perceived Alan as having his own sophisticated AI, but was effectively his own midget inside the box. Speaking with other players in the wake of our discussions, it is clear that there are many other players who share this perspective. Indeed, it may even be a cultural bias in the East. During his presentation, The Crystal Mythos and Final Fantasy XIII at GDC 2010, the Game Director and Scenario Writer of Final Fantasy XIII,Motomu Toriyama, identified a difference he saw between the presentation of player characters in western RPGs and Japanese RPGs. He explained that in western games, the player’s perspective was essentially the same as the main character’s perspective. The camera was extremely close to these characters and always slightly behind them, hiding their expression. In Japanese RPGs, on the other hand, the player was presented with a bird’s eye view of events, which positioned their perspective firmly outside of the main character. Toriyama felt that Japanese players were more comfortable with this external perspective and preferred to be bystanders to the events presented rather than explicitly driving the action themselves. While it isn’t clear what the ratio of inhabiting players is to relating players, there seems to be a large number of both. We were three-to-one in our group, but the strong sales and passionate fan bases of games like Final Fantasy and Metal Gear Solid seems to suggest that our representation was skewed. Whatever the truth, this example serves as a signpost indicating that there is a lot more to learn about how players interact with digital storytelling. Golems are Insufficient StorytellersIt is important to note the unique role of human creativity when it comes to the creation of good narrative in games. For games in each of our three “fishing holes” (closed, filtered, and open narrative), the essence and quality of the story always relies on humans. In each case, the burden of making the play experience fun and meaningful rests with some combination of the designer and the player(s), without whom, we believe, there can be no meaningful experience. In a closed-narrative game, the designer/writer authors a hand-crafted experience for the player. The designer’s uniquely human intelligence and creativity allows for subtle and interesting plots, threads, characters, and settings to be intricately woven as a masterwork. The player then is left with the human responsibility of experiencing the story in her own way, bringing to it her own participation, interpretation, and meaning. In a filtered-narrative game, the player expects to be more active in co-creating the narrative experience as lived during play, but wants a strong framing narrative and world within which to do this. The designer/writer’s challenge is to author evocative framing and sprinkle interesting story fragments in the game, providing a refereeing structure that facilitates the players’ ability to improvise and perform. In an open-narrative game, the designer servers more as a systems/mechanics/dynamics author and less as an overt storyteller. The designer creates an interesting, consistent, and well-defined possibility space, keeping in mind what sort of experience(s) he might want to be able to afford the player within that space. It is, then, the player who provides the crucial human creative element in producing a great game narrative – with the game system facilitating a positive feedback loop between the player’s creativity and the mechanics and emergent dynamics of the game system. In all cases it seems to us that it is chiefly through uniquely human creative input that good gameplay / narrative experiences arise, with the creative burden being divided between the designer and the player. This begs a looming question: can, in fact, this creative input only come from a human? Could an AI serve to replace a designer in composing a gripping story? Could an AI replace a designer in composing a compelling system in which mechanics combine into interesting emergent dynamic gameplay? Can any amount of programming fundamentally take the place of human creative input? Opinions on this topic are mixed, even within our small group. Clearly, well-programmed systems can aid the designer by affording interesting settings – in the case of a procedurally-generated world – or even goals (e.g., “You must bring the <ABC item> to <XYZ person>”). And complex code logic can provide texture or variance within a possibility space. Randomly generated items or NPCs with randomly generated names, features, and behavior characteristics can, to a limited extent, afford greater longevity or replayability. However, at the end of the day, no AI or algorithm is likely to create a brilliant closed narrative, nor create a set of game mechanics that will combine to form really interesting and well-balanced dynamics, nor impart meaning onto a gameplay experience. We therefore believe that, for the time being, the burden of creating great narrative in games continues to rest solely on the shoulders of humans, with both the designer and the player being fundamental to that process. Wisps to Light Your Way
There are so many more questions to be answered and insights to be fished forth. It is our sincere hope that this record of our explorations will spur further questions within bright, curious readers. Only with your own focused forays and explorations can we evolve our understanding of narrative in games. We hope that the insights and paired quandaries contained herein can serve as wisps to light your own investigations. Should you strike forth to chart these murky waters, we wish you good speed and a bountiful catch. The following is the entire collection of responses we received from our survey. Twenty-one of the thirty-some 2010 Project Horseshoe attendees responded to the following question: “What’s your favorite story in a game ever (either authored OR emergent)? What was it like to experience that story? Why was it so great?” Respondent: Adrien
Note: Adrien verbally related these ideas to several stories he told from the original EverQuest, but we failed to capture them on paper. Respondent: BryanEverQuest – my first and only character. A wood elf druid. It was the story of that character’s life. Starting off [with navigating a] difficult to understand city. Growing up, leveling, learning the world and the game. Making self-imposed pilgrimages back to my home city to buy new spells, [even though nothing in the game mandated it]. Eventually getting access to the plane of his goddess. Camping out at her feet when I quite the game for good. [It was an] incredibly personal experience. Online gaming was so new to me, and I loved imprinting [my own meaning] on things that didn’t have much in-game meaning. For example:
[The experience] was great because I felt a sense of personal connection and authorship. Other thoughts: Having certain parts of the narrative defined, and other parts remain undefined, afforded me the ideal opportunity to inject my own meaning into the story, giving me a sense of personal authorship. Had they defined everything for me, there would have been no room for me to inject myself into [the world and narrative].” Respondent: DanMy brother & I atop a hill in Battletech, with our custom designed 100-ton mechs. A horde of medium mechs with [Long Range Missiles] attacked. We blasted them back. In the end, our reactors were redlined, legs destroyed, and ammo almost expended. The last enemy crested the hill. My brother fired his last medium laser and scored a headshot, killing [the enemy mech]. He then promptly overheated... and exploded. The [resulting explosion] destroyed my mech as well. We broke out laughing. To this day, this game experience is a foundation of my relationship with my brother. Static narrative pre-baked by designers is a bullshit waste of time that appeals to a dying customer. Stop gussying up the poor worn-out mechanics you are prostituting. People and systems – not plot. Respondent: DevinTitle: Red Dead Redemption Reason: The story let me experience the Wild West – in the shoes of the characters. What really made it great was:
The best story evar was part of the conclusion of a [Dungeons & Dragons] campaign where my halfling barbarian “Biggen Burble” died saving the entire party. It was a tremendously emotional moment. A sheer joy combined with sadness because the campaign had no [resurrection] or raise dead spells, [so my character’s death was permanent]. I really felt like an important part of the story. The conclusion was fitting, and the other players sang a song in my memory. My character became part of the tapestry of the world and, as such, lived forever. Respondent: JasonIt was a game of The Sims I played. The story was about two sisters who shared the same house; one was an actress, one was an adventurer. The “one line” tale is that the actress sister always had the advantages, luck, and ease, while her sister did all the work. Until the actress sister had a huge cocaine binge and had to earn her way back from zero. The reason it was amazing was that the story was 100% systemic; it grew unintentionally from the Sims’ systemic behavior. The “cocaine binge” was a moment when I left my PC running unpaused for three hours, and returned to find that the actress had lost all her progress while the adventurer had been functioning at 100% the whole time. It was, in fact, the way that the system actions (without designer intention) triggered meaning and understanding in my mind. The “a ha” clicks in my head that led to an “understanding” of these “characters” were marvelous. Respondent: JeremyFinal Fantasy VII taught me that long-form narrative could fit very well into games. I also really enjoyed the optional branches of the narrative which merged back in very effectively and enhanced my understanding of the core narrative. I also have several stories that I really liked which happened in D&D games. Most of them involved pushing the experience in ways which weren’t intended by the designer. For example: We were being chased down a tunnel by a Balrog (we were on a flying carpet). I had a Rod of Splendor which could create a tent full of everything needed for a party of 200 people once per week. I used the rod to conjure the tent in front of the Balrog, entangling it and allowing us to escape. Respondent: JuanI can’t pick an example out of the top of my head, but I’m certainly most impressed when the actions I do in a game are telling me the story. Small cut-scenes to reinforce the idea are good (like in Shadow Complex) but the key in my opinion is to plan the player’s actions so (s)he is told a story as (s)he progresses through the game. Respondent: KatherineI really liked playing out the story of PaRappa the Rapper – a cute, optimistic little dog who learned how to rap in order to fulfill his dreams. I really liked watching him/me perform the moves, and seeing how people reacted to what he/I did. I loved celebrating his successes and I knew he’d be indefatigable in his efforts even if my fingers failed him, and would never bicker or bring me down. He made me want to be as cool under pressure as he was. Go PaRappa! You gotta BELIEVE! Respondent: KennyBioshock:
Red Dead Redemption:
Dream Chronicles (okay, this is one of mine):
A Zork story... A long long time ago on a computer far away, there was Zork. I always played with my husband and we would discuss and map together. Drawing the maps was actually fun. And I loved the moment of discovery every time I entered a new “room”. The description of the room and items was something to look forward to. And the fact that it was a [text-based] game created a huge, elaborate landscape in my mind. Perhaps my favorite part was going from room to room and saying, “Hello, sailor” – which would always be followed by the response from the game, “That doesn’t work here.” I don’t know if it was intentional, but there was a rumor that somewhere the phrase “Hello, sailor” would result in something cool happening. It was tons of fun trying to find the room that the phrase worked in. I never saw it work, but I heard that in Zork 3 (or so) near the end of the game a ship sailed by and if you said, “Hello, sailor,” the sailors would say “Hello” back to you. :-) Respondent: LinkPlaying through Silent Hill 2, I found myself enthralled by its air of mystery and foreboding atmosphere. In playing through it, however, I discovered that it was a revolutionary game in two ways. First, it utilized an implicit metric-tracking system that read motivation into the player’s [play style]. The story had very different outcomes based on the way the player approached it, but no explicit choices are ever made. No other game has ever done this and I am flabbergasted that this revolution was, basically, ignored. Second, I had recently lost my mother when I played the game. Silent Hill 2 explores death from multiple perspectives – both how dying changes the afflicted, and how grief can impact one’s outlook. Most media presents death as a discrete event which has an immediate impact and is then moved past. Silent Hill 2, however, shows just how traumatic death can be. Its reflection of my own inner turmoil was a comfort in that difficult time. Respondent: LisaOkami had a great interest curve – tricked me into thinking it was over and then surprised me with “OMG there’s so much more to go!” which delighted me. The story was supported by the mechanics and the aesthetics and it was all tied together in a nice little immersive package. Nobilis!! OMG so many good stories that were generated by the players. It was so great because the GM had a great “yes, and” attitude and made our contributions feel like [they were] part of the plan all along. The game structure supports a greater feeling of freedom than, say, D&D (in my opinion). Respondent: NickMy favorite is still probably Silent Hill 2, leaving out my own roleplaying experience (if you want that, Google “Measure for Marriage”). It was my favorite because it was subtle. It trusted me to figure it out, and it allowed me to interpret its images rather than beat me over the head with them. The power came from quiet creepiness, and it made me disturbed for the [whole of] play. The model of the creatures, the solitude of the town, the strange reactions of the NPC, and the unexplained horror were all part of it. Again, this is without my roleplaying experiences. My feelings from tabletop [roleplaying] kick the ass of any other narrative I have played. But that’s my work and my bias. Respondent: PeterI enjoyed Final Fantasy 6 because, while not always well-written, it was coherent and offered character development and emotional depth I wasn’t expecting and don’t often see in games. It also had minor but enjoyable gameplay links to the narrative (differing character abilities), though it was completely linear. Respondent: RonRingworld. The adaptation of a favored novel, early on... 1990? The excitement of being able to enter into that world... the [memorabilium?] was in the ability to engage in such an exciting and new way with a beloved story. It had huge flaws, but the concept was enchanting. Starcraft. The well-crafted storyline was compelling enough to get me to replay the game a second time after a few years. Storytelling per se was the draw – a great, ripping yarn, as it were. Uru [a.k.a. Myst Online]. The (flawed) ambition of creating a nicer, persistent world, with developments that evolved over the span of years, in some cases, and the intention of involving the community both in creating and evolving the story but also in determining the story developments to other members of the community – both through the virtual world but also through external sites – forums, fan sites, etc. The general premise of player as participant, in an identity more or less like their “real” one was also breathtaking. The Sims. Huge, deep, evolving personally-created stories involving a personal community. Very rich and creative and satisfying. Respondent: SquirrelWe began what would become one of our longest Dungeons & Dragons campaigns by the Dungeon Master describing us, the players, sitting around a table, much like we were just then in real life. Five minutes of introductory narrative later, we – the players – had been whisked away into a parallel world – the game world. We were our characters; our characters were us. We were playing ourselves, as characters, in a strange and unknown fantasy world. It was fascinating watching each of us roleplay ourselves as we would act in extraordinary circumstances – being held prisoner by a mysterious “benefactor” in a strange world in which magic was real, in which we could not understand a single spoken word. The better part of a year later, the campaign had progressed considerably. We had learned to survive, and even thrive, in this new world. We had learned to fight, to speak and read a new language, and even to use magic. Everything we did was as ourselves, as we would (we believed) truly act in such a situation. “A skeleton lives in that cave? Hell no, I’m not going in there!” At one point we recovered a book of “questionable” magic from a necromancer we’d defeated. While we sailed back from the necromancer’s island, we debated the morality of using the evil magic (even for good!) for nearly three hours. The argument ended when [my fellow player,] Benson, disgusted with our indecision, grabbed the book and threw it overboard. Discussion over. The most amazing moment of the campaign was when, during the climactic final battle scene, Chuck was crushed to death by a huge dragon that fell upon him. Jessica, his wife, had to roleplay herself watching her husband be killed in battle in front of her very eyes. Respondent: TonyAdventure (the original):
Myst:
World of Warcraft:
Portal Respondent: VictorIt was emergent when I was asked to run a mission in the Canadian Air Force simulation acceptance and I was briefed in to run it like a real mission. It was a pulse-pounding, exciting mission that felt totally real. Which is why it was so great – totally real and immersive. Respondent: ZackIco. Previously (and still) stories are very directly told through dialogue and narration, only secondarily focusing on animation, music, and setting. Ico eschewed the standard presentation by completely removing dialogue (for the most part) and narration and telling the entire story by the character interaction and gameplay elements. Even though the story was authored, it felt less like it was told to me and more experienced by me. The starkness of explicit detail allows us to fill in the gaps with our own powerful conclusions, while still being robust enough to preserve the integrity of the story from user to user. Respondent: Unknown AuthorIt’s hard to pick one, but I’d go with Portal. It made me laugh, and horrified me at the same time. And the narrative elements tied in perfectly with the game mechanics and level design... it seemed organic, as if narrative and gameplay had been created together instead of the more typical [feel of a] narrative grafted awkwardly on later.
section 5 |
Copyright 2000-2014, Fat Labs, Inc., ALL RIGHTS RESERVED |