The last-minute visual overhaul of the original Left 4 Dead cast, revealed during EA’s 2008 E3 conference1, came mere months before the November 18 launch date for the game. The changes, though minuscule as well as practically-minded, felt like a crushing blow to some, yet wholly inconsequential to others.
Some players might have missed the debacle altogether. Me? In response to the news, I envisioned a future scenario wherein Beyoncé models for a revised Alyx in Half-Life 2 EP3. I was desperate to commentate on the topic right there and then, but ultimately decided against it. Phew. Now, despite the fact that Left 4 Dead 2’s characters have barely been cast out to light, however, I am bold enough to suggest that an intriguing pattern has emerged, that a trajectory of design can be seen in the way Valve and Turtle Rock Studios have designed the various characters of the Left 4 Dead series.
In this text – which is just as much a history of the games’ development cycle as it is an analysis of the concept of “character” in the gaming medium – I will first navigate us through a series of dates, occurrences and currents that ultimately affected the make-up of the casts of both games. I will also attempt to explain and pinpoint decisions related to the series’ art direction. Since we’re broadly three months removed from Left 4 Dead 2 release date of November 17th, you’ll have just about enough time to read through my admittedly thorough assessment. Thanks for reading!
GameSetWatch’s Thomas Cross in fact previously touched upon the characterization of the Left 4 Dead cast last January, albeit from a slightly different vantage point: In his article, “Making Storytelling Look Natural”, he concludes that “Each [Left 4 Dead] character has a recognizable personality, if you take the time to learn them”, adding, “And yet these are still horror movie archetypes, it’s just that they’ve been given a slightly different stage this time round.” While I do wholly agree with his assessments, the focal point of Cross’ article is decidedly on the “hows” than the “whys” and the “whats”, which I rather choose to focus on in this article instead.
Since there exists a bidirectional relationship between characters and the narrative (or “stage”, as Cross calls it) encompassing them, if these characters are not instantaneously classifiable, positionable and relatable (as per our pre-existing societal knowledge), players can have difficulty in properly perceiving the story as a whole. This is especially true of the gaming medium, where avatar readability has been a chief priority for designers and artists – and justly so. For obvious reasons, it is preferred and encouraged that players reflect and transport their own personalities onto avatars as successfully as possible. Rather than fleshing out our avatars, though, there is a common, shared tendency to either aim for instant audiovisual accessibility and/or customizability. If developers chose to, of course, they could also attempt to round out characters in games’ narratives.
This, however, was not originally meant to happen in Left 4 Dead.
Valve employees have let out that Turtle Rock Studios, the original developer of the series, were not “…thinking in the same story-based way that Valve thinks”.2 Though Turtle Rock Studios ended up being fully assimilated into Valve in January 20083, it’s nevertheless more than probable that the creative intention to include a narrative as well as more storytelling sprang up relatively early in development, definitely before the actual acquisition took place. After all, as early as May of 2007, Michael Booth described Turtle Rock Studios’ relationship with Valve as going “far beyond being an engine licensee or distribution partner. … Valve would say the same thing.”4 On August 4, 2008, Doug Lombardi characterized the acquisition as “a piece of paper that makes things official.”5
Nevertheless, even later in the production, it also became apparent that dialogue-oriented cutbacks would have to be made. There has been mention of considerable omissions – after all, it has been confirmed that a full storyline for the game exists (and can be found hypothesized in the Steam Forums), including a “…dictionary with different terms for areas and infected that describe how the Director will handle things in [a particular] area.”6 In an interview with Eurogamer, Doug Lombardi further clarified the reasons for these cuts, stating that “[Valve] saw a number of reasons to pull back with the first game, primarily because the replayability suffers with a scripted sequence” and that they had “much more dialogue amongst the characters originally, and that was really getting in the way of people understanding the co-op nature”.7
The forthcoming Left 4 Dead 2, on the other hand, is to delve deeper into establishing visible narrative progression, unlike the first game, where most storytelling is relayed by the spray-filled walls of the game’s safe houses. Though an ingenious way of sending information transparently to players, the downside to these sprays is their inherent disconnection from the members of the cast as well as their personalities, rendering the sprays largely unsuited for character growth. In relation to this, Lombardi has promised that Valve will “bring more story to [Left 4 Dead 2] in terms of the players’ dialogue”7 instead.
“Not only are we telling the story of these characters, but we’re telling the story of this world, we’re seeing how things fall apart, and a new way of interacting with the infection” – Chet Faliszek6
This admittedly long-winded history lesson serves to clarify to us three major points present in the Left 4 Dead development timeline: First, the original Turtle Rock design of Left 4 Dead did not contain as much story as it does now. Second, relatively early on, Valve’s influence on the game nevertheless resulted in a proper storyline being implemented. Third, late in the development, playtests indicated that players reacted disappointingly to a more fully fleshed narrative, and some upgrades to these elements were thus once again scaled back. A fourth point, if we were to discuss the future now and not later on in the article, would be the return of the deeper storyline in the forthcoming sequel.
All these stages ended up influencing the characters’ make-up, and in this light, it’s perhaps easier for us to see the motivations that drove Valve to brusquely revise important player models so late. Above all, the changes can be seen to be a direct response to the needs and requirements of the arising situation: When storytelling was to be omitted, the team sought to find a way to keep the cast accessible and understandable to players of the game. With less story to round and flesh out the characters, something else had to be done to maintain this relationship. As a result, the revised survivors could be seen to be intentionally designed to be more easily identifiable and recognizable out of the box.
In this sense, the revision is closely connected to the characters’ primary function as players’ avatars: The now erased, first cast of characters was, for the lack of better description, damaged goods (Take a look at how adorable Francis looks on the left); Without the cut storyline elements functioning as padding, this type of character conceptualization ultimately made the less iconic cast harder for players to quickly define, classify and relate to. After all, storytelling in games still primarily relies upon a foundation laid down by other media: Literature, theatre, the movies.
In the classic arts, “actors” are used to play “roles” that in turn represent “characters”, by displaying personality through movement and action, and by emphasizing traits: Outwardly looks, responses, use of language (both bodily and verbal). The concept of the (Jungian) archetype, especially in the extended meaning found in literary criticism as referring to recurring generic motifs, is very useful here. In the realm of literature and theatre, the role of the archetype principally as a narrative method is relatively clear-cut, but in the gaming medium, the archetype-as-avatar relationship causes additional factors to consider: In designing avatars, developers have to take into account functions beyond the aforementioned types of characterization.
This is because characters in video games have functions in the programmatic rules of the game, especially so if we perceive games to be “limited formal systems”.8 In this case, the characters’ role as agents for narrative exposition of the overarching storyline is secondary to their primary function as player avatar, as vessels for player interaction inside the “game world”. Valve, in turn, have displayed a very deep commitment to primarily function-based avatar art direction that takes into account shape, form and lighting – right down to codified principles. They are effecting visual design that is philosophically oriented toward maximum functionalism: Portal, for instance, had a distinct, stylized aesthetic informed by playtesting and feedback; To quote Lombardi again, “the first iteration [of Portal] was much more lush in terms of graphics…and people were having trouble identifying the pathways… We had to make it this sparse environment to get to the gameplay.”7
Valve designer Jason Mitchell’s paper, “Illustrative Rendering in Team Fortress 2”, was an astonishing explanation of the multitudinous rendering processes that went into the graphical design of Team Fortress 2. (For a concise explanation of “illustrative rendering”, a video was released). Similarly, it is no accident that Valve are planning on publishing a comic book adaptation of Team Fortress 2: The primary source of inspiration for the graphical look of TF2 was early 20th century commercial illustrators – Cornwell, Leyendecker, Rockwell. Advertisements are, at least in theory (so diluted, the postmodern TV spot!), designed to convey as much information as possible in the least amount of space and time. This idea, of maximum potential identification, heavily factored into the art design of Team Fortress 2. Elaborate techniques were applied “so that players are able to easily identify other players in the game, and assess the possible threat”; Characters were designed to possess “…distinct silhouettes that can be easily identified”.9
During the aforementioned 2008 E3 reveal of the new survivors, Gabe Newell’s presentation briefly read: “Taking the lessons learned from Team Fortress 2 about read hierarchy and providing useful and immediate information via character design.”10 This is where the team’s specific focus on readability, identification and distinction bled into the final stages of Left 4 Dead development, and had as much effect on the characters as did the process of cutting down the storyline. This is not at all surprising considering Valve’s “cabal” development model that puts great weight on spreading, sharing and mobilizing innovations as well as individuals. When this particular methodology of design, of visual distinction and identification, was applied to the Left 4 Dead cast in the form of the revision, a distinct collision of the role and function of “character” arose, causing the upheaval I referenced at the very beginning of this article.
From an avatar-, character-oriented point of view, after all, there are obvious differences in-between the two games: Firstly, each character in Team Fortress 2 is a distinct base-level representation of abstract class (or “role”, rather, should we get into a semantic discussion); It’s acceptable and even desired for these avatars to be visually generic, exaggerated and archetypal because the very meaning of the word “class” is the same as the word “genus”. Left 4 Dead’s characters, however, are derived from a less programmatic register, from everyday society: Instead of being absolute base-level manifestations, the four survivors are slightly higher-level: We have Bill, a Vietnam veteran; Louis, an assistant manager at a retail electronics store; Zoey, the teenage daughter of a wealthy family, and Francis, a tattooed biker. Intriguingly, as stereotypical as the characters are, it is nevertheless possible to further compress, if you will, the characters into lower-level archetypes.
I would personally affix the characters the following tags: The Wise Man (in Jungian11 terms, again…), the white-collar everyman, the newbie and the lonely ranger. (Feel free to correct me in the comments!)
Now, as a distinct look and feel had already been established for the characters, some players had already identified (with) the characters. Semantically speaking, the revised cast of characters no longer represented the exact same “person”. Yes, one could potentially argue that all Francis did was lose a few pounds and shave, and that Louis got a hot TV makeover, but what about Zoey? Herein lies a perfect collision of the two aforementioned types of character function, of character-as-narrative-vehicle and character-as-avatar-as-game-mechanic.
All in all, we can conclude that the revised cast of Left 4 Dead 1 fulfilled two separate design-related needs: On the one hand, the developers had to find a way to narrow the narrative gap now present between players and the characters due to cutbacks being exacted on the game. On the other hand, Valve’s general, company-wide art direction had been simultaneously moving towards a more functional implementation. These two sides melded, blending appeal with function, ultimately creating a slightly more iconic set.
Finally, to steamroll the entire article into Left 4 Dead 2. Beyond promises of more dialogue and narrative exposition, we can already see for ourselves how the cast, this time, is decidedly less archetypal and consists of higher-level stereotypes; The mere existence of Coach’s religious leanings and Rochelle’s taste in music tell us this much! There also exist early indications of interpersonal schism (not in small part due to the inclusion of Nick), which obviously means that the characters’ position and function as avatars is no longer as strictly dictated by player identification alone. To briefly reiterate, the reasons for this are the following: Firstly, Left 4 Dead is now an established brand. Secondly, players are aware of the gameplay mechanisms and no longer need to be introduced to it. Thirdly, Valve have artistic ambitions to produce more storytelling.
As briefly illustrated in my earlier post on the topic, the posters of the sequel’s campaigns already exhibit a larger degree of interpersonal relationships and personality. It is also evident that the roles affixed to Coach, Rochelle, Ellis and Nick are higher-level than those assigned to Zoey et al: A high-school football coach, a news reporter, a mechanic and a conman. As we did with the first game, here too we can assign a compressed, lower-level set of stereo-/archetypes to match the four. They might be, for instance, the teacher, the heroine, the kid and the cynic. The very fact that I am struggling to distil the characters into lower-level stereotypes tells us that Left 4 Dead 2’s characters are going to be more fully rounded than those of Left 4 Dead.
What you’ve just read is an interpretation of the overarching processes and phenomena – the whats and the whys – that had an effect on how each these three casts ultimately turned out visually, narratively and functionally. At this juncture, I do want to emphasise how great a share of things outlined here are nothing new in itself, as it is only natural for us to expect many of the aforementioned upgrades and improvements from a video game sequel – these are the very things that we expect sequels to be and to do. Despite these facts, it remains nevertheless both intriguing and therapeutic to attempt to map out the various processes and mechanisms that go into game development in order to achieve a more comprehensive understanding of the functions and features of the characters we play as.
In closing, I offer a thought: I always play as Francis, for deep down he loves life, the hating just a façade; how else could he smile so, in the face of a zombie apocalypse?
- http://www.destructoid.com/e3-08-left-4-dead-main-characters-get-redesigned-95661.phtml [↩]
- http://www.bit-tech.net/gaming/pc/2009/07/06/left-4-dead-2-interview-a-chat-with-chet/2 [↩]
- http://www.valvesoftware.com/news.php?id=1401 [↩]
- http://www.gamesindustry.biz/articles/turtle-power [↩]
- http://www.gamasutra.com/php-bin/news_index.php?story=19643 [↩]
- http://arstechnica.com/gaming/news/2009/06/before-its-time-valve-explains-left-4-dead-sequel-to-ars.ars [↩] [↩]
- http://www.eurogamer.net/articles/left-4-dead-2-interview [↩] [↩] [↩]
- Koster, Raph. A Theory of Fun for Game Design. 1999. p38. [↩]
- http://www.gametrailers.com/video/illustrative-rendering-team-fortress/23520 [↩]
- http://www.shacknews.com/onearticle.x/53637 [↩]
- Jacobi, Jolande. The Psychology of C.G. Jung. 1948. p64 [↩]