The Cartoonist, our winning entry in the 2010 Knight News Challenge, emerged from two research programs. For the past two years, my research group at the Georgia Institute of Technology has been cataloging and analyzing the burgeoning genre of “newsgames” — videogames about current and past real-world events. That research produced a book, Newsgames, which will be published next month by MIT Press.
Meanwhile, on the other side of the country, professor Michael Mateas and his Expressive Intelligence Studio at UC Santa Cruz have been working on the problem of game generation by creating artificial intelligence tools to create a virtually infinite number of games.
The goal of our two-year Knight grant is to create a tool for generating newsgames on the fly, making it viable to create a videogame about a breaking event. This is done by identifying the issue and an angle for editorial or reporting, boiling the story down into its constituent agents and their relationships, and selecting from a range of rhetorical archetypes. Anyone who understands how to use the tool will be able to create a newsgame, remixed from the structures and mechanics of popular arcade games, within five minutes or less. The game will output to Flash and HTML 5 for instant uploading to the web, where it can be paired with reportage, columns, video, infographics, and cartoons covering the same current event.
Read the rest of the post here at MediaShift Idea Lab.
This is a section from my critique of Left 4 Dead in my Master’s thesis, “The Judgment of Procedural Rhetoric.” I’m posting snippets here on my blog to drum up interest in the rest of the work. If you’d like a copy of the full document, please email me: chungkingDOTespressoATgmailDOTcom (because of copyright issues, please make sure to give me your full name and your website when you send the email).
Anybody familiar with the work of Anna Anthropy might recognize this as an introductory attempt to do for 3D maps what she does for 2D maps.
Structure as Literacy
Left 4 Dead alternates narrow interior spaces with open exteriors. While moving through the interiors, players often have multiple distinct avenues to choose from. These multilinear spaces encourage exploration, but they also have the potential to feel like mazes that disorient and separate players. It is also difficult to see upcoming dangers when indoors, as special Infected have numerous ways to hide themselves around corners or behind objects until they are ready to strike. The exteriors, on the other hand, provide better visibility and a single general axis of motion. These spaces afford strafing—the ability to physically pan sideways around an obstruction or threat—but they typically funnel the player to a single ultimate destination. In exterior spaces, disorientation comes primarily from partial decreases in visibility due to foliage or detritus.
In order to understand various types of modular level design in Left 4 Dead, we will make use of a series of maps below. The blue line represents the most efficient pathway through the level. Red lines represent distractions from this optimal path. White highlights delineate accessible space, and white lines signify obstructions (some of which can be entered or climbed upon). Yellow lines in the fourth map signify the “scatter” pattern needed to survive Tanks during the scene’s finale. Yellow dots represent places to remain still during attacks from Hordes. One must understand a few things about the way the AI Director works in order to understand why experienced players would ever stray from the blue, optimal path.
First, items such as ammunition, explosives, pain pills, and med-packs can be scattered anywhere throughout the level. The Director decides which of these items to provide, then randomizes their location throughout the level. This selection and location process changes on each attempt at the level, meaning it cannot be memorized; therefore, the primary temptation to follow red lines is to look for these items. The major secondary causes of diversion are Witches. Passing next to a Witch usually ends in disaster, but they typically rest in places that can be wholly avoided by choosing a less optimal path. Our only purpose for even recognizing paths as non-optimal is that enemies never stop spawning in Left 4 Dead (except right before finales). The best way to minimize casualties—the implied goal of the game’s design—is to move at a constant pace, as a group, along the shortest path possible.
Figure 5.1 “Blood Harvest” Intro
Figure 5.1 shows the first level of the “Blood Harvest” campaign, which takes place primarily outdoors.[i] Players begin at the bottom of the map. The white dots at the beginning of the stage represent dense forest. Movement through the first half of the map often proceeds slowly, as the group clusters together to eliminate straying common Infected that come running out of the woods and onto the path. Boomers and Smokers hide among these trees, pulling players into the darkness or leading them astray through blindness. Midway through the level is a trailer, which sometimes contains medical supplies. Lingering here often triggers a Horde, exacerbated by Boomers that hide behind the trailer or off in the woods to the right. The final L-shaped sprint to the saferoom opens visibility but also threatens to pull the team apart as injured teammates lag behind. Play in this level is much more complex in Versus than it is in Campaign mode, because the foliage and surrounding ravines provide tactical opportunities for the Infected team.
This level is basically a “track”-type space in Nitsche’s dichotomy.[ii] This is one of the best maps for new players to run in order to learn basic mechanics. It’s early in the campaign, so it’s a straight, narrow line in an exterior setting designed primarily to set the mood and help a newly formed team build trust. It affords only optimal, unilinear movement. The one major distraction point (the trailer) is one of the clearest learning opportunities for players who don’t understand the importance of constant motion. Even if the team becomes mired in a Horde onslaught, they will almost always have enough medical supplies to make it to safety. It is uncommon for Witches and Tanks to spawn in this level, but if they do the straight bath forward or backward provides ample opportunity for escape and defense. We can conclude from all of this that track-type spaces are the best for developing basic literacy and team dynamics.
Figure 5.2 “No Mercy” Intro
Figure 5.2 is from the “No Mercy” campaign, which is typically the first campaign played by new players and the most popular Versus mode map; therefore, it mixes interior and exterior spaces to form another kind of tutorial. Players begin on the roof of the southernmost building, and they work their way quickly to the ground floor. Following an alley, the team exits onto a street. A witch typically sits right around the corner from where the alley exits out; there are also cars that will summon a Horde if shot. Players choose here whether to proceed along the street itself or through the building in the middle marked with red lines. Moving through the building will increase overall travel time; it holds a higher density of common Infected, but it also might contain health packs. Wrecked trucks litter the streets, creating little pockets of space to entice players away from the optimal path. At the end of the level, a staircase tempts players to linger just before the protection of the saferoom.
The brevity of this level makes it a relatively safe place to learn the tradeoffs between searching through cramped hallways for items and simply charging forward to the safehouse. Because there is only one mini maze-type space with clear entrances and exits, the consequences of slowing down are minor. The only significant danger of this level is the event of a Tank or Horde spawn in the street crowded with cars. Cars that can set off alarms are placed nearby the entrance to the safehouse, meaning that in the event of an accident it is fairly easy to beat a hasty retreat.
Figure 5.3 “No Mercy” Sewers
Figure 5.3 is the beginning of the third level in “No Mercy.” It begins in a series of warehouses connected by darkened alleyways. The way forward is obvious once one knows in which direction to move, but non-optimal paths through ancillary warehouses may contain health packs. Proceeding along the blue line, players enter a courtyard. At one end of the courtyard is a gas station that explodes when shot. Once again, trucks create pockets of space to distract players from the blue path. Right next to the gas station, at the yellow dot, is a forklift that slowly ascends to allow access to the rooftops. The forklift triggers a Horde, and players must run along the rooftops to get back inside at the top left of the map. Smokers, hiding in between the trucks, can easily pull players off the rooftops before their teammates know what’s happening. Soon after this scene, not pictured, is a figure eight-shaped sewer system.
This is a moderately difficult area due to the need to stop to raise the forklift and the added vertical element of running along the rooftop, where there is a hazard of being pulled downward by Special Infected. This forces the rest of the team to track backward to protect the fallen player, and it’s one of the most common causes of a wipe. Before coming to this level, players have already encountered a hard defense point where they must wait out a Horde in order to proceed, but the forklift is much more open and lacking in supplies than previous defense points. The rooftop shows how much more complex a level gets when verticality comes into play. Players must simultaneously keep an eye on enemies descending from up and over a higher rooftop while keeping guard on the Special Infected lurking below. This area primes the team for No Mercy’s finale, which occurs in a two-story building with an open rooftop.
Figure 5.4 “Blood Harvest” Finale
Figure 5.4 is the final level of “Blood Harvest.” Players proceed down a narrow railroad track and climb on top of some train cars at the end. The mid-point of this section often contains a Tank, which requires players to backtrack or ascend the car to the right marked by a red line. Rounding the corner, players drop down into a cornfield to trigger a Horde. Players only have to travel in a straight line to exit the field, but the corn obscures vision almost completely. Considering this is the end of the campaign, multiple teammates may be injured and limping. Enemies can attack from every direction, further disorienting the player. Exiting the field, the team comes upon a house and adjoining barn that serve as a base for the finale. Players can hole up either in the house or in the barn, but they’ll probably have to run circles around the house during two Tank phases.
The house and its surrounding open field are the closest Left 4 Dead gets to the “arena”-type space in Nitsche’s dichotomy.[iii] A tacit assumption is that, by this point in the campaign, the team has learned to work together. The conceit of the finale, wherein the team holes up against overwhelming waves of enemies, takes much of the burden of providing challenge off of the level design—explaining the use of a somewhat nonlinear space. There is also much less clutter in the final arena, emphasizing tactical fluidity.
With the exception of the winding, track-type map of 5.1, it is simple to identify the discrete rectangular shapes used to construct all of these levels. One can observe in most of them a sort of pulsing between interior and exterior, wide and narrow. Interior spaces tend to have multiple avenues of possible movement, but they also feature dead-ends. Exteriors generally only afford unidirectional motion, but all of these open spaces feature objects such as trees or cars used to distract the player from that single direction. When placed in sequence, these basic variations create a rhythm of attack and defense, motion and pause, and centripetal and centrifugal force upon the team’s unity.
[i] Image source for all Left 4 Dead maps: http://l4dmapdb.com/, modified.
[ii] Michael Nitsche, Video Game Spaces (Cambridge, Mass.: MIT Press, 2009), 173.
[iii] Nitsche 183.
The Judgement of Procedural Rhetoric by Simon Ferrari
Committee: Ian Bogost (chair), Fox Harrell, Michael Nitsche, and Celia Pearce (in absentiae)
Thursday, 18 March 2010, 1:30-3:30p
This thesis establishes a theoretical framework for understanding virtual spaces and roleplaying in relation to Ian Bogost’s theory of “procedural rhetoric,” the art of persuading through rule systems alone. Bogost characterizes the persuasive power of games as setting up an Aristotelian enthymeme—an incomplete argument—that one completes through play; however, I argue that the dominant rhetoric intended by a team of game designers is subject to manipulation through player choice. Discrete structures within the play experience cause the meaning-making possibilities of a game object to pullulate in a number of directions. Procedural rhetoric is not comprehended or created when reflected back upon after play: we interrogate it, piece it together, and change it through play.
If rules are how the designers express themselves through videogames, then the player expresses herself by forming a personal ruleset—a modus operandi or ethical system—in response to the dominant rhetoric. Furthermore, game space is not merely the place where this dialectic occurs; it also embodies a ruleset in the way it organizes objects and directs the flow of play. The thesis proposes a model by which games, which are “half-real” according to theorist Jesper Juul, can be judged intersubjectively—that is, in a way that accounts for the objectivity of their rulesets and the subjectivity of player experience. By fully understanding the dynamic between the three procedural influences of rules, space, and identity, we can learn more about designing persuasive game systems and enhance the possibilities of subversive play.
Required reading: Persuasive Games, Half-Real, What Video Games Have to Teach Us About Learning and Literacy, Video Game Spaces, and The Ethics of Computer Games
Core Ludography: Far Cry 2, The Elder Scrolls III: Morrowind, and Left 4 Dead
For my LARP field study I played a night full of Mafia with Paul, Pauline, and Jenifer from class (along with a number of their friends). Doug Wilson of IT Copenhagen calls Mafia “the most political game ever conceived.” The game is an ideal LARP for non-traditional roleplayers, because there are no combat rules to remember or stats to track. Typically the game is played with between 10 and 20 people, seated in a circle. We had ten for our session, a number which lends itself to a more intimate and competitive experience.
One player takes the role of the narrator (game master) who randomly doles out roles at the beginning of each play experience, tracks the state of the game, and provides a narrative context for every game action. There are two cycles in the game: night and day. The game begins at night, with all heads bowed. Six players were assigned the role of basic townsperson; they have no special abilities or duties. Two players constitute the Mafia, and each night they raise their heads to select one person to kill. One player is the detective, and each night they can point to one person, asking the narrator if that player is in the Mafia. Finally, one player is the doctor, able to select one person per night for protection. Nobody knows what role the other players bear.
During the day stage, the results of the Mafia’s activities are reported. If the marked player was not protected by the doctor, they die. If the detective accurately discerned a Mafia member, she may want to declare the fact. But if she reveals her identity, she becomes an easy target for the Mafia if the doctor is unable or unwilling to protect her. Then the townspeople begin accusing each other of being in the Mafia, stating their (usually tenuous) reasons for believing so. Players can choose not to condemn anyone, but usually the Mafia players will attempt to sway the townspeople toward killing each other (which leads to counter-accusations, etc.). An accused player gives a defense speech, then the players vote on which person to lynch.
When the Mafia murder somebody, the narrator does not reveal what role the dead player bore; however, when the townspeople lynch a player they are told what role the dying player held. The game ends when either all townspeople or all Mafia members are killed.
It took awhile for us to get the game started. During the first round, I forgot which role I had been given and ruined everything. Everybody forgave me when the narrator forgot what was going on during the second round and spoiled that one. The third attempt was a success, especially for me. Because I knew what roles everybody had been assigned during the first two unsuccessful attempts, I used fuzzy math to try to discern which players were the most likely to be Mafia. Basically I went on the false mathematical assumption that the chance of three successive “heads” in a game of coin-flip is 1/8 instead of 1/2 (I still want a look at the theorem that establishes that bit of nonsense).
As it turned out, my fuzzy math worked! I successfully picked the two Mafia even though I was only playing a lowly citizen. The first time I nominated one of the suspect players, nobody believed me and didn’t vote for him to die. So during the next round, I falsely stated that I was the detective and that I knew the second suspect was mafioso. The healer was dead at this point, so I knew I would be killed after the round was over. I gave an impassioned speech about self-sacrifice, everybody bought it, and we lynched the suspect player. I was right about the pick, and I was also right that the remaining Mafia player would off me that night. But the real detective was still alive, and he found out who the second murderer was in time to win the round for the townspeople.
The next round, I was killed straightaway. I assume it was because I had such good hunches during the first game. This is similar to the experiment of iterated prisoner’s dilemma in game theory, where bias from previous plays affects how the players within the dilemma choose in subsequent rounds. I watched the players to figure out if any of them had tells, and I discovered that one of the players giggled whenever he was in the Mafia. During the third game, I heard the distinctive giggle on the first night and outed him to everyone during the day. After I explained my reasoning, a few players believed me and we successfully lynched him. Then I got killed the next round. Playing Mafia too well usually means you’re going to get axed.
By the fifth and final match, I’d consumed a bit too much alcohol for my own good. This resulted in me persuading the townspeople to murder two innocents in a row. I’m glad we stopped after that round. So I’ve played Mafia twice now, and I’ve never actually gotten to be in the Mafia. As a result of this, I can’t speak for how to strategize a defense while playing one of them. The rounds that I was the healer and the detective were the rounds where I died the first day, so I also don’t know how to play as those roles. Mostly I’m good at playing a standard townsperson, and I’ve got a knack for picking at least one of the Mafia off before getting slaughtered the following night (healers tend to be very stupid; they never protect me, their star player).
Is there a difference in embodiment while playing something like Mafia over a videogame? I don’t believe so. Identification with avatars in first- and third-person camera views has been well-documented. There’s a palpable, giddy energy to live action play, but for calculating players such as myself the difference seems negligible. This is probably because of the principle Gee calls the “psychosocial moratorium,” or what Huizinga calls “the magic circle”; this is a protection from real-life consequences and harm that some believe is intrinsic to play (perhaps the only exception would be in what Caillois identifies as Ilinx, or “vertigo,” play… there is a real danger present with things like roller coasters and skydiving).
I have no problem sacrificing myself for the team in Mafia, because I know I’m not dying in real life. The act of taking on a role is always a necessary step away from absolute embodiment and identification. I shun anonymity in online play, so I’m always just playing an accentuated fraction of my real self when I play any game. This appears to hold true in live play: I was sarcastic, calm, and reasonable (except when I became inebriated… which can affect performance in online games as well).
As for the strategic difference between NPCs and real human players, I hold, along with Jason Rohrer, that there isn’t much of one. I didn’t know any of my fellow Mafia players exceedingly well, so I tested and prodded them much as I would an alien computer intelligence. As a material and physical determinist, I think people behave with predictable regularity (except in panic situations). I read the one player’s giggle-tell much as I would a sound cue in a videogame. If I’d been playing with family or close friends, this might have been different–but only because I would know them and their personal rulesets all the better. They could act to upset my predictions, but I would probably be able to counter-predict that if I were playing carefully enough.
One notable exception to this rule was that we had a player named Akido who spoke little English. His defense was always, “Why do you think I’m in the Mafia? I am innocent!” It was impossible to read him, because he wasn’t fluent enough to craft different responses based on his current role and situation. I correctly identified him by luck during the first round, but every time after that (if he were mafioso) nobody was able to nail him. We avoided accusing him, perhaps out of fear that we would be discriminating against him. I wonder how this could be simulated in an NPC?
Jenifer made two videos of the experience, but I can’t speak to their quality because I don’t want to download them:
End of Life is an interactive fiction about family life and decision-making. It started as an idea in Ian Bogost’s newsgame project studio. One of the branches of newsgames we have identified for our book is the documentary game. Typically these have a medium-length (20 minutes to two hours) playthrough time and are built as a mod for a 3D engine. There are three major types: spatial, procedural, and personal. Personal documentary games mix spatial and system-based models in order to tell share a story from a unique, subjective point-of-view. End of Life is a text-based adaptation of the documentary game form, addressing the real-world issue of “end of life counseling” or the decision whether to pull life support from a dying loved one.
The high concept pitch for EoL would sound something like, “It’s Ruben & Lullaby meets The Sound and The Fury.” Point-of-view switching is a powerful literary device, but in static texts this typically implies a forced perspective. In EoL, the player can switch back and forth between five family members at any moment and in any order. If they don’t like a character, they can ignore her for the course of the playthrough. The invalid family patriarch is our Benjy Compson (the mentally handicapped member of Faulkner’s fictional family), providing commentary that the active family members do not have access to. Some characters always do the same things in every playthrough; most have branching choices based on their moods at certain points in the day. When there is no choice in action, mood will instead dictate how the character mentally reacts to her situation.
Ruben & Lullaby provides the inspiration for the interaction model: the player controls a wisp that can nudge the emotions of one family member per hour. I see this as a direct contradiction of the interaction model of The Sims, where players are cued to a desire or feeling in the Sim that they can rectify or not by dictating action. Players of R&L and Facade are often frustrated when their commands don’t lead to tangible results in game, and I wanted to capture a similar frustration in EoL. Each family member begins the playthrough in a randomized mood. Each is variably susceptible to particular mood swings, leading to healthy dose of guesswork and replay value. The player can also choose to abstain from influencing the characters, letting the drama play out based on the beginning values.
At the end of the game, the family convenes to decide the fate of the patriarch; some will vote to keep him alive if they are in a good mood, some if they are in a bad mood. This decision takes place offscreen, much as in the violent sections of Greek tragedy (mostly because I wasn’t good enough to code it dramatically). The player has gleaned parts of their personalities in the playthrough, but he doesn’t know everything about each family member. Most importantly, their ethics aren’t considered. The game argues that people make decisions based on who they are and the mood they are in. Ethics certainly make up who we are, but they tend to be remarkably malleable under duress. Decisions are also relational; some people, under some circumstances, will take radical action to counteract what they see as the controlling influence of others.
In discussing digital media, we often fall back on an essentialist logic that says that an artifact is aesthetically legitimate if it maximizes the affordances of the medium; however, there is a slightly older aesthetic criterion, coming from Theodor Adorno and the Frankfurt School, which states that aesthetic legitimacy arises not from essentialist qualities but from the reflection of the work’s means of production–it has to reify the cultural milieu of a time and place, adopting a suitable form for conveying it. End of Life draws from the latter school of thought, directly confronting a relevant public issue and encapsulating how one specific family deals with it.
The suggestion that a digital artifact should provide always immersion, embodiment, and agency is perverse. It only makes sense if one views digital media as escapism, created to fully engage the user in the place of the real world around them. A brute fact of human life is that we don’t have control over much of our lives or the lives of others. Aarseth argues that games become more “gamelike” if they are configurative, that the player should be able to see the meaningful influence her actions have on a virtual world. I would argue that agency and embodiment mean more in configurative work when they are directly contradicted in non-configurative work. By taking these essential qualities away sometimes, we make them more cherished. Such qualities should be selected from to suit the work, not the other way around. Defaulting to what is important to us robs it of importance. This is an educational opportunity, an antidote to the intoxicating sense of power that most digital artifacts provide. Some things simply aren’t configurative in the real world; families are a good example.
A week before finishing this project, I finally found published theoretical grounding for my position. In their early work on augmented reality games, Jay Bolter and Blair MacIntyre argue that point-of-view switching provides adequate embodiment in lieu of actual agency in a digital environment.2 I actually don’t find their particular example of this principle compelling; basically they simplified Twelve Angry Jurors to Three Jurors, strapped a backpack computer and a virtual reality visor to a player, and then allowed the player to switch between inhabiting the mindset of one of the three characters as a static drama played out. I think EoL takes point-of-view switching one step further and provides a better proof-of-concept for their argument.
I consider End of Life no small success. My writing is admittedly the weakest element; mentally I finished the piece the moment I finished coding the framework girding the story. This project combines everything I’ve learned how to do in Flash thus far (excepting animation), and it constitutes the first true state machine I’ve ever made completely by myself in the platform. Even though the writing is somewhat trite, pulling from every cliche of everyday family life I’m familiar with, it becomes true in that I pulled it from one specific, real-world family (my own).
There is some room for future development here, both graphically and procedurally. Right now there are two variations for every character in every round based on there mood. Given the way the structure is set up, I could add mood variations to the branching story sections or add a third mood variation (neutral) given enough time and literary inspiration. I would also love to try to remake this project as a true documentary game, in a 3D engine, with unique art assets and dialogue. The current iteration of this project represents the utmost level of my design and programming abilities given the time constraints and the specifications of the assignment.
I should note that this situation didn’t actually happen to my family, and the personalities have been a bit blown own to be more compelling. My grandfather died five years ago from Alzheimer’s disease, asleep in his bed, in the room that I grew up in. This isn’t meant to be a universal story, though it can be generalized to the extent that families are, after all, families; it is a directed experience featuring characters with largely determinate personalities. This is the way I wanted it, and I hope the player enjoys what I crafted for them. A big thank you goes out to Graham Jans for teaching me how to randomize variables in Flash. I’m also indebted to my family for providing me with the strong personalities embedded in the family members of this fiction. Thank you to my father, who used to work as an intensive care nurse, for describing the hour-by-hour care of a comatose patient.
Adorno, Theodor. Aesthetic Theory (Minneapolis: University of Minnesota Press, 1997), 16-45.
MacIntrye, Blair and Jay Bolter. “Single-narrative, multiple point-of-view dramatic experiences in augmented reality” in Virtual Reality 7 (London: Springer-Verlag, 2003), 10-16.
“Speaking in Djinni” spoke directly to my childhood self: I remember quite clearly, after seeing Disney’s Aladdin, pondering for hours how I would perfectly phrase my three wishes so as to maximize their potential and avoid fatal misunderstanding (I was terrified of the tale of Midas). Harrell relates the difference between human language, which vaguely describes phenomena in highly subjective ways, and the imperative languages of djinni and computers, which literally have the power to create but only produces satisfactory results when worded carefully and in the proper grammar. The argument that follows shows how the peculiarities and affordances of programming languages inform the software and development kits that are built upon them, which in turn constrain and guide the actions allowable within the artifacts constructed with those digital tools. This work can be seen as a direct antecedent of the work of Bogost and Montfort on creating the Platform Studies series for MIT Press.
“Benchmark Fictions” seems to be a relatively early work in comparative media studies; I say this because of its matching of a strong conceptual frame with a disappointing proof-of-concept executed in tedious early digital standbys such as wikis and chatbots. Benchmarking fiction takes its name from benchmarking software, which tests the performance of the computer hardware that runs it through a series of procedural pings. The prime directive is to separate the “content” of a literary work from its original printed or digital “form” in order to learn more about both through a series of procedural translations or adaptations. A major problem here seems to be that the authors quickly write off the suggestion that form and content really can’t be separated, just before they mangle a fairly revelatory short story by attempting a number of crude digital “adaptations” that have little or nothing to do with what or how the story means. They also note how the interaction models of games require novel forms of adaptation, but the best they can muster is a chatterbot that responds to pre-scripted questions with coy hints as to who is The Lady and who is The Tiger.
Mateas and Stern finally answer that final question for the authors of “Benchmark Fictions,” by recognizing the possible antagonism between interactivity and narrative before finding an out in Laurel’s work in interactive drama. Drama is an ideal model because it already involves actively constructing a “story” arc through acting. It also operates on a model of causally connected actions with a tight rising in tension. Mateas takes the Aristotelian hierarchy of drama and substitutes “character” for “user/player,” showing that in an interactive drama that player is the inferred formal cause of all meaning except the action/plot which undergirds the experience. We can see this at work in their Facade, a game that will always progress from introduction, to initial signs of unrest, to drinks, to open conflict, and finally to either a happy resolution or the player’s expulsion from the apartment. Yet within each of these major stages, procedural variation and player choice lead to a number of possible conversations and revelations.
Years before “Benchmark Fictions” had been written, two satisfactory procedural translations of complex source texts had already been undertaken. Chris Crawford left his job at Atari to create Balance of Power for the Apple II. The game allows the player to take on the role of lead negotiator either for the United States or the U.S.S.R.; the goal is to avert elevating tensions leading toward mutually-assured destruction. In fact, this game was Crawford’s interpretation of the memoirs of Henry Kissinger; while working under Nixon, Kissinger speculated that the U.S.S.R. would run out of fiscal security sometime during the 1980’s–the way to “win” the cold war was to survive until that occurred. Nixon and Kissinger embarked on a number of compromises with Soviet authorities that both conservatives and liberals in the U.S. disparaged. Balance of Power communicates this tangible sense of danger and walking-the-tightrope.
Will Wright and his team at Maxis, on the other hand, created one of the first popular citybuilding simulations called SimCity. This game allows players to build basic public utilities such as transportation and power, specify three city zoning types (Residential, Commercial, and Industrial), and maintain growth and the public interest through taxation rates and law enforcement. Wright came up with the idea for the game after studying Forrester’s Urban Dynamics, a book about urban growth and decay cycles containing both reflective analysis and prescriptive suggestions for managing public welfare, sprawl, and re-gentrification. Most infamously, Forrester argued that social spending on underprivileged minorities in the inner city would decrease a city’s worth instead of increasing it; on the other hand, he also had the foresight to predict that the construction of the Interstate system would lead to the neglect of areas in between major highway hubs. Wright attempted to model as much of these principles and arguments in SimCity as he could, including the famous example of encoding a correlation between rising tax rates and social unrest.
Crawford worked alone, rigorously working and reworking his procedures and datasets until he could fit all the information and complexity he desired into the constrained memory that he had available. The programming language and platform had a massive influence on the finished product, so much so that Crawford spent nearly twenty years crafting his own “Storyworld” development kit and scripting language in order to present Balance of Power again. SimCity, on the other hand, has gone through a number of translations and versioning for every new operating system that emerges. The countless number of spin-offs to the series, some more popular than the original thread itself, provides us with an ideal model for how benchmarking fiction should work. These games also show us how the expression of the same idea changes based on the imperative languages and development kits used to create each new iteration.
1) Do you think it’s possible to separate form from content? Or do you think a better experiment toward finding the essence of digital media art would be to construct novel works such as Facade?
2) How convinced are you by the solution of substituting a dramatic arc for narrative? Are Mateas and Laurel overly relying on old Greek paradigms of meaning creation in their insistence of an arc-like structure with tension, climax, and denouement?
3) Benchmarking experiment, following my work in Ian’s project studio: in groups, pitch two different kinds of games (from editorial, documentary, infographic, or puzzle) dealing with the same public issue (traffic, healthcare, the war in Iraq, etc.).
Short Essay: Analyze and compare the narrative grammar of Propp, Greimas, and Aarseth.
From earlier studies in film history and comparative literature, I’ve been familiar with Vladimir Propp’s narrative grammar for quite some time. Propp broke down a selection of Russian folk tales into 31 functions and 7 generic characters, elaborating possible combinations and causal sequences. That Propp was able to create his typology was no surprise to me, because I had already learned about oral mnemonic techniques used in commedia dell’arte and the codification of Platonic dialogue and Homeric epic. Once you understand that oral storytellers memorize a set amount of objects, characters, and events along with a structure for connecting them and then improvising, the revelation that folk stories carried down from an oral tradition follow such a grammar is almost trivial. My major problem with this grammar is that it ignores everything that’s wonderful about folklore—the flourishes and improvisations. In the face of Mark Turner’s ability to create compelling prose building upon otherwise stale research in cognition and early childhood development, Propp’s grammar strikes a dull chord. I am reminded of Janet Murray’s ability to find personal meaning in the rote act of manipulating the falling bricks of Tetris. Perhaps Propp too felt this lacking, reflected in his later decision to study literature instead of linguistics (Wikipedia).
Greimas attempts to bridge the gap between deep linguistic structure and surface narrative structure, explaining “the fact that a narrative enonce is represented at the linguistic level by a whole paragraph” (797). Much of his work goes into breaking down subjects, object, and verbs (which he renames “functions”) into the form of signs. Greimas expands narrative grammar into story grammar when he dichotomizes narrative and non-narrative enonces. A non-narrative enonce builds from a stative verb, or one that addresses being and qualification (800). Toward the end, Greimas conceives of how to represent the literary device of asyndeton (he merely labels it “ellipsis”) as symbolic logic in the form of a series of conditional statements—or a narrative syntagm (804). Greimas understands that a compelling narrative grammar must explain the structural affordances that allow for story grammars and literary devices.
Propp was a Russian formalist, meaning he identified with the goal of separating the artist from the text and then showing how formal elements such as syntax and structure were inherent in how a text means. Greimas, on the other hand, was closely associated with structuralists such as Levi-Strauss. Structuralism attempts to break down a text into signs and the structures by which they’re related. These structures are held as “real,” whereas the signs they order merely refer to the signified objects and events that exist outside the literary artifact. From the introduction to his Cybertext, Espen Aarseth appears to adhere to a variation on post-structuralism—a movement that critiqued such assumptions as the importance of the author’s intent and the inherent “deep meaning” of a text. Poststructuralism originated primarily in France, but I’d argue that Quine’s indeterminacy of translation principle made the first step toward establishing the cultural relativity of the connection between signifier and signified. Poststructualists (according to Wikipedia) break down the distinction between signifier and signified to hold the combination as “real,” but I’d add to this that they recognize that there exists a culturally-specific (and personally-specific) version of each of these constructs.
Aarseth’s model of an ergodic textual machine—placing the “text” inside the vertices “operator,” “verbal sign,” and “medium” (21)—seems to uphold the idea that there is no single meaning to a cybertext; rather, this meaning is generated through the conflict of the vertices. These textual machines constitute localized microcosms of the general poststructuralist mission of critiquing social structures through playful deconstruction. Although Aarseth specifically establishes the textual machine in order to explain cybertexts and not textuality-in-general, I think it also applies back all the way to oral storytelling in a way that Propp was unable to capture. In “Double-Scope Stories,” Mark Turner imagines a dialogue between mother and child to accompany the reading of a bedtime story—it is this dialectic (trialectic?) that Aarseth ends up capturing with the ergodic machine.
Assignment: Write a design sketch for a narrative engine to be coded next week.
For my narrative generation project, I’d like to create flash fiction that describe the varying experiences of moving through discrete zones in a level from a generic FPS game. The initial idea for this comes from an earlier essay I wrote for Michael Nitsche last semester, about reading Left 4 Dead as a team-based rhythm game. In that essay I elaborated on what I saw as a somewhat vague but valid set of ideas from Henry Jenkins and Celia Pearce about “evocative space” and “game design as narrative architecture.” I delineated a few basic binary options for any zone in a level, such as whether it is wide or narrow, linear or multi-linear, light or dark, defense or offense, enclosed or open. Although I didn’t fully flesh the idea out, I grappled with explaining the psychological effects that various combinations of these attributes—and the act of moving between different zones—would have on the player.
Around six months later, I came across this short article by Justin Keverne. In it, he breaks Resistance: Fall of Man down into seven distinct models of gameplay and attempts to show how various combinations of these can elicit aesthetic responses such as “pushing through to teammates” and “camaraderie followed by loneliness.” His explication of this design method lacks the binary structure of mine, but he takes the critical step toward integrating it into the MDA model of starting from an aesthetic goal and working backward to determine the dynamics and mechanics required to elicit it. In the MDA model, level design seems to be a bounded box surrounding and structuring dynamics—it is a conduit through which dynamics can be fed directly into an aesthetic grammar.
Thus, for this project I will attempt to create a narrative generator that asks for desired aesthetic responses (in sequence) as input. The engine will generate an introductory zone based on the binary attributes I delineate, then it will attempt to create a sequence of zones to match the emotional flow described by the user’s input. Another desired feature will be the user’s ability to constrain the choice of binary attributes (such as ordering the machine to only use enclosed and linear spaces). The output of the generator will be in the form of a short story. This is a decidedly structuralist approach to the assignment, but I don’t mean for the output to be the final product of the endeavor; rather, I see the project as being a tool for brainstorming and design-sketching for level designers. I have no idea if my coding abilities will be adequate for this project, but I hope to be able to at least mock up a convincing prototype using PHP and mySQL. Even if I can’t hack the back-end sufficiently, I at least hope to develop the vague notion of level-design-as-narrative into a comprehensive grammar.
Images from Wikipedia, Creative Commons, etc.
Simultaneously posted on Bogost’s News Games blog.
Pictures for Truth is a newsgame funded by Amnesty International, produced using Microsoft’s XNA software development kit. You play an American journalist in China just prior to the Beijing Olympics. You have a date to meet with a Chinese journalist covering poor living conditions at a toxic electronics dump. When you arrive at your hotel, you receive a call informing you that your friend has been detained by authorities at the dump.
A police officer at the dump confiscates your camera and hauls your friend off to jail. You must find a new camera, interview people at the dump and outside a jail, and take pictures to accompany the “stories” generated by the interviews. You write three stories: about the health issues surrounding the dump, the working conditions of those living near the dump, and about China’s municipal system in regards to the death penalty (this story is unlocked by completing the first two).
Aesthetically the game is rather beautiful. Unlike many of these investigative reporter games we’ve played (like Homeland Guantanamo), PFT is rendered in realtime 3D. So instead of clicking between discrete composed scenes, you get to move through sensory-immersive recreations of the dump and the Chinese jail. The game is in black and white, presumably to compare the game with a newspaper. Texturing is spare, mostly hatched greyscale or pencil scribbling. The characters look like white paper cutouts.
A lot of effort has gone into making PFT “gamey.” You receive fame points for every interview question you ask and for adding pictures relevant to the stories you are composing. Fame unlocks three “power-ups:” a zoom lens, an extended hard drive for your PDA, and a hidden camera.
Unfortunately, only the hidden camera is actually useful. You must have this in order to take pictures inside the jail cell where your friend is being held. The zoom may increase the fame points you receive for taking pictures, but if this is true it would violate the photographic rule of thirds (placing the focus of your pictures in a third of the screen, with axis of action from the subject aiming toward the unoccupied two thirds (this assertion is disputable, but the fact that the NPCs remain static means that there’s no real reason to zoom up close to their faces to catch, say, teardrops forming in their eyes). The PDA hard drive space is only needed to store pictures; you can delete unwanted or used pictures, and there aren’t really enough subjects to require massive amounts of space.
The biggest problem with the game is that there’s no real room for agency on the part of the player. All one has to do is click through every available conversation piece with each NPC. Anyone used to playing games made by Bethesda knows the drill: swiftly click through the conversation tree without paying attention to much of anything in order to unlock everything, and then find the one piece you need to progress the game state and read it carefully. Adding insult to injury, the game only takes two lines of conversation and forces you to use them for the article you’re writing. It would have made much more sense to add a score for each line and then allow you to combine then to maximize the “fame” points for the article. This would at least provide some feedback educating the player on the quality of different types of information. The picture mechanic does some work to remedy this: much like in Dead Rising, capturing points of interests (represented as nodes or “hot points” once the cameras snaps) awards more points; re-using the same subject twice in an article subtracts points.
In the end it’s obvious what Amnesty’s purpose was for this game: not to teach one how to be a journalist, but to teach one how difficult it is to be a journalist in China. There’s also the ancillary educational goal of teaching players about living, working, and municipal conditions in China.
The makers do connect these three educational points with their narrative thread. Another occupant of the jail holding your friend is a woman who has been jailed for trafficking heroin. She did this in order to buy medicine for her daughter, a girl you encountered earlier in the dump standing around by herself. You can choose whether or not to enlist a doctor’s help for the child, but it really doesn’t make sense not to do it and your ending condition doesn’t change if you do this or not.
As a side note, this game did highlight for me the difficulty of using XNA to create one of these games. Only a Windows machine will run games made in this way, and they have a somewhat high barrier of entry on account of the fact that Windows has to install the .NET Framework 3.5 in order to run them. This requires roughly 15 minutes of downloading, installing, and configuring (plus a mandatory system restart) – of course, this is a one-time only thing, and now your computer is set up to play anything else built in XNA. The payoff seems worth it in the end, however, because the product comes off as much more polished than something developed in Flash. Realtime 3D rendering is always a plus, and at the very least this game didn’t require as much of a download time as a game distributed through Kuma War’s download client.
This is more for my own record-keeping than for your reading pleasure, but some of my friends might find it fun to tear into me for how I present this information and/or be inspired to go check out the stuff we’ve got on the News Games blog.
The News Games blog has enjoyed fairly consistent amounts of visitors throughout its existence. November and December (2008) saw the highest number of unique and repeat visitors (10,479 and 605, respectively) because of some high-profile linking by mainstream gaming websites. Q1 of 2009 saw a slight dip in readership despite increased recognition by other websites (7,694 unique visitors), and numbers have been steadily rising through the beginning of Q2 (with 5,691 unique visitors thus far). The top referring links were from (in descending order) Kotaku, Slashdot, Game Set Watch, and Digg—although the most consistent flow of readers came from no referring link at all, which means that a number of our regular readers bookmarked the website themselves.
Recorded visit lengths show that roughly 20% of visitors stayed long enough to thoroughly read at least one article, three percent read multiple articles (spending up to an hour on the site), while 2% browsed the website for longer than an hour! Readers came from at least 21 countries, with all continents represented except Antarctica. Seventy-six percent of readers came from the United States, 6% came from Canada, 4% from the UK, 3.4% from Brazil, 1.6% from Australia, and 1.2% from Singapore (other countries represent less than 1% of the readership).
News Games has published 91 articles over the past 6 months, which averages to one post every other day. Once normalized, the posting schedule was one article on every Monday, Wednesday, and Friday. The articles cover topics such as (but not limited to) journalistic connections to mainstream games; analyses of news, editorial, and documentary games; connections between games and the journalistic values of verification and transparency; explorations of interactive info-vizualization and traditional newspaper puzzles; and general musings on the current state of print journalism and the nature of games as a serious artistic/informational medium. Readers and contributors have authored 116 unique, non-spam comments on articles over the course of the project.
The most popular search query used to reach the site is “journalism games,” followed closely by “news games” and “news games georgia.” This implies that either a decent number of Internet readers were interested in the subject beforehand or became interested in the subject after the project had been linked by popular gaming websites. The latter query also shows that the project has been recognized and explicitly associated with Georgia Tech. Following the most common queries are “newspaper games,” “tenants of journalism,” “ethics in games,” and “journalism kids.” The searches show that our project has become associated with the disciplines and business models of traditional journalism, the question of morality in games, and the potential educational uses of news games.
News Games enjoyed a healthy dose of mainstream attention throughout the past six months. Gawker Media’s Kotaku, one of the most popular and influential videogames-related websites in the world, linked an article called Do Really Games Qualify As Escapist? in late November of 2008, netting a decent amount of outside attention for the blog and challenging mainstream gamers to consider the ability of games to address social issues. This article was also picked up by Digg, a popular bookmarking and crowd-sourcing website. Game Set Watch, the respected casual arm of the games industry mega-site Gamasutra, linked four articles from News Games through Q4 2008 and Q1 2009:
Newsgames in the Pipe (a hypothetical business model for making newsgames in the newsroom)
Technical Aspects of Breaking Newsgames (a followup to the previous article)
Dead Rising and Interventionist Media Ethics (about an action reporting videogame and its possible relation to the media ethics of Conrad Fink)
Huys/Hope – Turkey’s First Political Game (an analysis of the connection between Turkey’s first editorial game and its relation to journalism and Gonzalo Frasca’s Madrid)
The Online Journalism Blog published an article in early Q2 2009 called Now That Journalism is In Trouble, Why Not Play With It? that aggregated numerous writings from the News Games blog, setting up a discourse on the question of videogames as a possible solution to the problem of news media companies failing to reach wider audiences in the era of the Internet. This article was picked up by the UK Guardian newspaper, a bookmarking website called Slashdot, and Kotaku. Readers treated the idea with a mix of excitement and fear, thinking that both games and the news might be possibly tarnished or at least changed irrevocably if the connection between the two were strengthened in the future.
The most recent recognition of the News Games project comes from Professor Roger Travis (of UConn) and his Video Games & Human Values Initiative, which linked the article Beyond Good & Evil and Photographic “Truth” (about another action reporting game and its reliance on a naïve view of documentary reality).
Our project received some attention from academics as well, which was exciting (and, frankly, amazing) considering the relative youth of the project and its niche focus. Two professors had the News Games blog as required reading for media studies classes in the spring semester of 2009. Professor Ed Halter of Bard College drew upon the site for case studies in his Film 106: Introduction to Documentary Media. Professor Casey O’Donnell of the University of Georgia blogrolled News Games for his Introduction to Game Design class; students in this class were required to write 250-word synopses of games-related articles from the blogroll each week, many of which ended up being drawn from News Games.
Because the University of Georgia was close by Georgia Tech geographically, the project’s spring 2009 graduate assistant (Ferrari) visited UGA to deliver three lectures in February. The subject of the lectures was an introduction to news, editorial, and documentary games (as well as the work of the News Games project); he spoke to a 150-person class on Introduction to Reporting taught by Professor Barry Hollander—as well as smaller classes taught by Casey O’Donnell, Game Design and Introduction to New Media. The latter lecture was attended by Dr. Hugh Martin, a senior faculty member of UGA’s Grady School of Journalism, who responded to the idea of teaching newsgame production in the context of J-school positively. Journalism students and professors responded well to the idea of integrating newsgame production into the newsroom/reporting process. Ferrari shared some resources on learning Adobe Flash-based game development to the students who were interested in pursuing the idea further in their own coursework.