The Turn of the Tide

International E-Sports and the Undercurrency in Dota 2

And tell me Mr. Waugh, what do both baseball and business need? Someone to keep the books.

—Robert Coover, The Universal Baseball Association, Inc., J. Henry Waugh, Prop.

I realize I am not only watching the game differently but am watching a different game. . . . I’m watching the whole game, and responding the way an ordinary fan responds. . . . They’re watching fragments—not the game itself but derivatives of the game—and responding, so far as I can tell, not at all.

—Michael Lewis, Moneyball: The Art of Winning an Unfair Game

Well before the TAT-8 fiber optic cable was dragged across the bottom of the Atlantic Ocean in 1988, the deep sea had already been bisected by American and British business. Even a hundred years before the TAT-1 telephone line was engineered in 1956, the first transatlantic telegraph cables were already descending to the depths of the ocean.[1] The infrastructure and ideology supporting these kinds of bit torrents is not new. Since the Age of Sail, the Earth’s oceans have operated as a technology for trade. According to the logic of combat and capital in the sixteenth century, current and currency have been wed to each other for hundreds of years. Earth is not a desert planet, but a water world whose surface is traced by trade winds and shipping lanes—ecological phenomena labeled according to their geopolitical role, rather than the relation of atmosphere to lithosphere. Whereas the topographies of trade and travel mapped the movements of Empire on a continental territory, the smooth surface of the sea was also dissected and differentiated by capital.

If exchange is contingent on comparison and comparison requires the individuation of otherwise undifferentiated matter—the cut that differentiates quantitative measure from qualitative movement—then cutting characterizes capital. Capitalism cuts. The cut of competition, which sociologist and game theorist Roger Caillois (2001, 12) calls “agon,” also makes games possible. Videogames are not simply equipment for sequencing, serializing, and scoring the performance of various actors (both human and nonhuman alike), but in the process of measuring analog play in terms of digital rules they also enact a necessary comparison or competition. Numbers level the playing field and, in the process, reduce play to what can be measured, what can be cut. The precise character of this cut is what is at stake in the difference between videogame and metagame. Beginning with the history of statistical play—from the first German wargames in the eighteenth century to the tabletop encounters of Warhammer (1983–) to the real-time strategy of Warcraft (1994–) to the international competitions surrounding Dota 2 (2013–)—this chapter investigates games that operate according to the agonistic logic of what Steven Connor and McKenzie Wark call “cutting.” These games also mirror the agony of grammatization, a term originally used to describe the transformation of undifferentiated sounds into discrete words and a term with which Bernard Stiegler characterizes the relation between digital technology and labor in For a New Critique of Political Economy (2010). Like Claude Shannon’s definition of information as the probability of a given signal within a field of noise, for these theorists both games and money produce statistical information and behave probabilistically. Whereas Connor and Wark focus on their own brand of economic game theory, the management strategies of Gabe Newell, the co-founder and general manager of Valve, puts the homology between digital games and finance capital into play.

Since founding the game development studio, digital distribution retailer, and increasingly multifaceted tech company in the late 1990s, Newell’s moneygames have harnessed vectors of information to derive value not from gameplay as such, but from the metagames that operate outside the computer screen. From simply purchasing and playing videogames to modding software, selling virtual commodities, spectating live events, and predicting future markets, for the past decade metagaming has become synonymous with an untapped ocean of informatic and affective labor. The precarious labor of players is not a form of currency, but an undercurrency driving a deluge of vectoralist management strategies. The undercurrency carries with it the sedimentary traces of all forms of play. As play accretes within this digital undertow, different forms of metagaming are made exchangeable and flattened into one monolithic unit of measure: productivity. This chapter charts the undercurrency through an in-depth analysis of one of the most famous play in the history of Dota 2, a sea change in which the statistical play of two tide hunters transformed the metagame.

The Turn of the Tide

In the last weekend of August, in 2012, the audience assembled at Benaroya Hall were not enjoying the Seattle Philharmonic perform the Pines of Rome (1924). They had not gathered to watch an Indigo Girls concert or listen to Ira Glass reinvent the radio. Drawn forward as if automated, spines pneumatically straightening and hands moving to meet faces and mouths, the spectators in Seattle were caught up in The Turn of the Tide. Assembled on the stage, like some kind of ballet mécanique, stood two enormous glass boxes. Onlookers gazed beyond the wood-trimmed proscenium and oversized vitrines to an industrial spectacle of another sort: professional players packed behind computers, fingers clicking mechanical keyboards and multibutton mouses, faces lit by rows of LCD screens, and heads entombed in noise-cancelling headphones. Hanging above the booths, massive movie screens projected aggregate electronic performances alongside live commentary and statistical analysis in three languages. Rather than rioting, like the audiences attending Futurist concerts nearly a hundred years ago,[2] twenty-first-century spectators cheered from stuffed chairs, waving iPads in lieu of pennants as silk banners streamed from box seats, billowing in the air conditioning at The International: Valve Corporation’s 1.6-million-dollar Dota 2 tournament (see Figure 5.1).[3]

Figure 5.1. A panoramic photograph of The International at Benaroya Hall documents the crowded hall, box seats, and two teams competing at the heart of the event.

As its name implies, The International featured sixteen five-player teams from around the world digitally duking it out within Valve’s then-forthcoming remake of Defense of the Ancients or DotA (2003–), a classic Warcraft III (2002) mod independently developed by a community of player-programmers since 2003. Hailing from Sweden, Germany, Ukraine, Russia, China, Malaysia, Singapore, Australia, and the United States, at The International pro gamers not only competed for what was the largest cash prize in the history of electronic sports, or “e-sports”—an astounding million dollars to the winning team[4]—but also hoped to earn a living by playing a game that had not even been officially released to the public. Almost a century ago, Bertolt Brecht (2002, 183) derisively joked, “There seems to be nothing to stop the theatre having its own form of ‘sport.’ If only someone could take those buildings designed for theatrical purposes . . . and treat them as more or less empty spaces for the successful pursuit of ‘sport,’ then they would be used in a way that might mean something to a contemporary public that earns real contemporary money and eats real contemporary beef.” With Dota 2 and The International, Valve is producing an electronic sport for a new kind of spectator and in the process has revealed a corporate metagame in which “real contemporary money” has found new ways to imbricate itself within the industry, software, culture, spectatorship, and play of videogames.

Out of the hundreds of hours of tournament proceedings, seventeen seconds stand out. During the winner’s bracket semifinals, the Ukrainian team Natus Vincere (Na`Vi) and their star player Danil “Dendi” Ishutin competed against Invictus Gaming (iG), a Chinese team with an intimidating track record led by Wong Hock Chuan (see Figure 5.2). Na`Vi were the clear fan favorites. The beloved Ukrainians were defending champions, Internet celebrities, and rags-to-riches “Dota millionaires” after winning the first international tournament held at GamesCom in Cologne the previous year.[5] Whereas fans consider Estonian captain Clement “Puppey” Ivanov a tactical genius and refer to power-player Alexander “XBOCT” Dashkevich as a suicidal killing machine, Dendi is hailed as “the face of Dota” and a digital trickster likened to the “Lvivian batiar . . . a class of urban mischief maker” native to his hometown in Ukraine (riptide and McEntegart, 2012). Given the influx of utterly dominant teams from China at the second International, the Ukrainian champions had been recast as underdogs and the “last foreign hope” after all European and North American teams had been defeated.[6] In the prelims the five Chinese teams seemed unbeatable, and in the tournament brackets not much had changed.[7] After the first match with iG, a sixteen-minute-and-fifteen-second stomp and the shortest game at The International, Na`Vi’s teetering trajectory through the ranks of Chinese teams seemed all but over.[8]

Figure 5.2. From left to right, in 2012 Na`Vi included Sergey “ARS-ART” Revin, Danil “Dendi” Ishutin, captain Clement “Puppey” Ivanov, Alexander “XBOCT” Dashkevich, and Dmitry “LighTofHeaveN” Kupriyanov (top), while iG includes Luo “Ferrari 430” Feichi, Zeng “Faith” Hongda, Chen “Zhou” Yao, Jiang “YYF” Cen, and captain Wong Hock “ChuaN” Chuan (bottom).

Interpreted in terms of Na`Vi’s failure rather than iG’s success, shoutcasters and statisticians alike felt robbed by the anticlimactic first game. Throughout The International, and in Dota 2 more generally, Western teams are often described in terms of risky, aggressive play styles that result in thrilling, high octane spectacles loaded with the exceptional achievements of individual athletes. On the contrary, the rhetoric at The International characterized Chinese and East Asian teams as passive, patient, and predictable; their players were portrayed as anonymous agents of some Red machine grinding away behind the “Great Firewall.” Fans felt as though they “weren’t watching Na`Vi play against iG; [but] were watching Dendi, Puppey, XBOCT, ARS-ART, and LighTofHeaveN versus China” (emphasis original), five superheroes fighting an undifferentiated enemy on the battlefield of Dota 2 (riptide 2013).[9]

Racialized rhetoric is not uncommon in e-sports and, with the original Defense of the Ancients’ intense popularity and relative insularity in China and other parts of East and Southeast Asia, terms like “farming” and “ricing” have become common ways to describe long-term, economy-driven strategies in contrast to the micromanagement of team fighting (see Figure 5.3).[10] During iG’s victory over Na`Vi in game one, commentator David “Luminous” Zhang joked, “The Chinese [have] about two thousand years of farming experience so they have to be leading here” (Valve 2012b). As commentators attempted to generate compelling frame narratives on the fly, the often inscrutable and intensely complex teamwork taking place on the screen was reduced to common cultural stereotypes for the sake of spectacle. Lisa Nakamura (2009, 151) has argued that “early journalistic and academic accounts of the Internet stressed its utopian, democratic potential for erasing (or at least hiding) race and creating a ‘level playing field,’” but The International made it patently evident that there was nothing post-racial about Dota 2.

Figure 5.3. Around the time of The International, more than one racist meme emerged online depicting Chinese players (top) or Anti-Mage (bottom) (a powerful hero with a long experience curve) as rice farmers.

Popular narratives in which Western teams perform audience-friendly feats of daring and heroic individualism while Chinese teams opt for calculated collective wins at the expense of visual spectacle reveal how nationalism and race are still at play in digital games and how stereotypes function to reduce both the skill and, at the time, the winning strategies of East Asian players to racist tropes. In another essay on World of Warcraft (2004), Nakamura (2013, 189) discusses the ways in which certain play styles are portrayed online, arguing that “specific forms of gamic labor, such as goldfarming and selling, as well as specific styles of play, have become racialized as Chinese, producing new forms of networked racism that are particularly easy for players to disavow.”[11] Disavowal is contingent on the cognitive dissonance between, on one hand, the leisure activity of playing (and watching) games and, on the other hand, the labor associated with making money from those same activities. In the particular case of e-sports, it is with equal fanaticism that the right hand evangelizes the encroachment of global capital in the form of professionalization and monetization while the left casts out the value-seeking behaviors of so-called “farmers” that are coded (in both massively multiplayer online games and Dota 2 alike) as specifically Chinese. Alexander Galloway (2011, 120) flattens these differences by declaring that “we are the gold farmers” in a post-Fordist economy in which all forms of play and work, productive and nonproductive labor, are “impossible to differentiate.” The perception of two distinct play styles emerging at the second International not only reveals how racial difference is produced and negotiated within online digital communities (and the Orientalist image economies in which the Western commentators traffic), but also demonstrates the difference between the production of e-sports as a spectator sport and as a numbers-driven pursuit of wins in order to turn a profit (to the tune of 2.6 million dollars, no less). So despite the perceived tediousness of “Chinese Dota,” iG’s economic strategies, both inside and outside the game, seemed to be working.

When Na`Vi and iG logged in to the second game, so did over half a million people from around the world (Valve 2012g). More than two hundred thousand sleep-deprived Chinese fans pulled all-nighters on Sunday to watch the end of the tournament. The International was broadcasted with live commentary in English, Russian, and Chinese on at least two streaming video providers and, importantly, within the Dota 2 client itself.[12] Spectators were encouraged to download Dota 2 for free through Steam, Valve’s online store and digital distribution service, in order to watch the game in the game.[13] Within the software client, viewing Dota and playing Dota occur through the same actions. Audiences were invited to watch The International in real-time through the same technical apparatus and graphic interface as those players competing live. Rather than relying on the skilled hand and eye of an “observer” (a category of digital cameraman and real-time editor made popular in televised e-sports such as StarCraft [1998][14]), telematic spectators at The International became their own digital director, choosing the perspective from which to view the game and even spying on the precise mouse movements and interface manipulation of individual players half a world away. Beyond manipulating the perspective oneself or following the gestures of a live observer or professional player, Valve also implemented an artificially intelligent camera that exploited the delay between action and broadcast to algorithmically select and prematurely pan to those locations with the highest density of incoming information. Within this holistic software client, viewers selected their unique vantage point among a vast, in-game audience while, at the click of a button, they could jump from spectating to playing into their own 5 vs. 5 pickup games. In Dota 2, watching sports and playing sports are no longer easily separated. Even analytics gathered within each individual viewer’s Steam profile (e.g., total time played) do not distinguish between playing Dota and watching Dota.[15]

To this end, it is crucial that the company does not simply distribute free software, but requires users to first install Steam. Once the hurdles of installing the storefront (and hooking up a credit card) are overcome, the convenience of the digital marketplace, viral consumption within social networks, and impulse purchases (usually at extreme discount) far exceed the potential profit from the sale of a single title. For example, Valve’s multiplayer shooter, Team Fortress 2 (2007), made twelve times as much money as a free-to-play game with its booming cosmetics market[16] than it did while packaged in The Orange Box (2007) and sold on Steam (Miller 2012). Contrary to their label, free-to-play games are not the altruistic offerings of the game companies, but rather the giftlike distribution of software in order to combat piracy and create a temporary credit economy that, after a momentary delay or deferral, will generate greater profits through in-game advertising, cosmetic markets, and player-to-player evangelization. For Dota 2 and e-sports in general, videogame software ceases to operate simply as a game and transforms into service platforms, social networks, and online marketplaces for another type of play: a massive, million-dollar moneygame, in which free-to-play is actually a code word for free-to-pay.

Beneath the surface of spectatorship software, amidst the currents of international currency, and caught up in the undertow of tacit racism, the tide was about to turn in game two of the winner’s bracket semifinals at The International. Back at Benaroya Hall, the crowd’s response moved from agitation to elation. After their flawless victory in the first game, the hunters became the hunted as Na`Vi sprang an elaborate trap preying on iG’s predictable, economic strategies and that summer’s metagame, the in-game decisions and team configurations that were fashionable in 2012. The Ukrainians were able to invert what seemed like a perfectly executed ambush into a “reverse all-kill” that left the Chinese team gasping for air. This seventeen-second sea change became the breaking point of both The International 2012 and the strategic metagame that had governed much of the gameplay that summer. The English transcript for what has come to be known simply as “The Play” has since become the stuff of legends:

DAVID LUMINOUS” ZHANG: [Invictus Gaming is] coming to the left, smoking up, and here we go!

DAVID LD GORMAN: Oh man, big wrap-around gank is going to be the name of the game for iG. Who leads the way? I believe it should be Zhou. They’re going to cut for the shorter path. They storm up the river. Patience from Zhou, waiting in the wake. Na`Vi is about to be caught. Oh, there’s the sleep! The surge! He catches everyone! Oh, this could be a total disaster! Vacuum in! Ravage on everyone! The black hole as well! LighTofHeaveN turns it around! Ravage as well, stolen by Dendi! Are you kidding me? They turn it around! Four heroes dead for iG!

DAVID LUMINOUS” ZHANG: Five heroes dead! Chuan trying to survive. Chuan’s going to go down. Puppey talked about the Naga counter. It’s LighTofHeaveN, with his BKB! They turn it around. I don’t even know how they can do it! Standing ovation from the crowd! The last tower will be going down and I think they sense blood in the water! (Valve 2012c)

As Chuan’s onscreen avatar burned in effigy, Dendi leapt from his seat, spitting expletives in Russian as the packed hall began to rhythmically chant “Na`Vi! Na`Vi! Na`Vi! Na`Vi!” Awash in emotion, Russian shoutcaster Vitalii “V1lat” Volochai lost his voice completely, breaking into choked sobs as he watched his team turn it around. Whereas the English and Russian commentators worked themselves into a frenzy, the three Chinese commentators observed the reverse all-kill in a stunned radio silence, dead air hanging over their drowned team.

Despite the differing emotional responses based on obvious allegiances, the question still remains: What exactly happened here? These remediated, real-time narratives of the commentators do not adequately account for what is at play beneath the phenomenal surface of The International. There is an immense and incommensurable gulf between the human experience of The Turn of the Tide and the ultrafast electrical undercurrents governing the computational processes and network protocols that make up the mechanics of Dota 2. The spectacle of The Turn of the Tide dramatizes the phenomenological problem that Mark Hansen frames in Feed-Forward: On the “Future” of 21st Century Media (2015). First discussed in the introduction of this volume, in Feed-Forward Hansen (2015, 27) acknowledges the shifting status of human experience in an age of vast, ubiquitous, ultrafast, atmospheric media, and determines that the central question of twenty-first-century media is “How can consciousness continue to matter in a world where events no longer need it to occur, and, indeed, where they occur long before they manifest as contents of consciousness?” Videogames are one such media ecology in which players sink, swim, and surf the statistical undercurrents that exist beyond not only the horizon of consciousness but also the horizon of embodied sensation and perception. Rather than manipulating these currents directly, pro gamers and publishers alike rely on inductive techniques not to control, but to predict the atmospheric operations that exist along the vanishing point of human experience.

Beyond the mechanics of the game itself, Valve’s desire to redevelop and redistribute a decade-old modification of another company’s game is due to a different kind of undercurrent: that of an untapped attention economy in China. The original Defense of the Ancients has been popular in East Asia for a decade and, in 2010, boasted around 10 million suspected Chinese players.[17] By releasing a regularly updated, standalone version of the game for free (not to mention offering unparalleled access into a carefully cultivated competitive scene) on the condition that customers first download Steam, Valve hopes to exploit a previously inaccessible consumer (and prosumer) base—a leviathan of potential profit in China.[18] Selling the Steam service rather than a single piece of software is one endgame of Dota 2. Through Steam, Valve converts metagames into moneygames by investing in the productivity of their player base. From community-designed cosmetics to the spectacle produced by live spectators to player-programmed mods, Valve seeks out historically undervalued metagames and capitalizes on the undercurrent of player production, the undercurrency at work in The Turn of the Tide.

From Wargames to Warhammer to Warcraft

Over two hundred years ago, in the late eighteenth century, ripples of The Turn of the Tide were beginning to gather in the most unlikely places: in the classrooms and parlors of Johann Christian Ludwig Hellwig, the German professor of mathematics and natural science credited for inventing the wargame, or kriegsspiel. In his marvelous seven-hundred-page tome, Playing at the World: A History of Simulating Wars, People and Fantastic Adventures from Chess to Role-Playing Games (2012), Jon Peterson documents the first wargamers and their attempts to build miniature models of combat for recreation and research. Starting in 1780, Hellwig expanded and instrumentalized the rules of chess in an attempt to more faithfully represent the terrain and tactics of large-scale military combat for pedagogical purposes. Divorced from the consequences of battle and based on a familiar game, the first kriegsspiel was designed with the dual function of entertaining and educating Hellwig’s students, “the pages of Braunschweig . . . bound for military service” (Peterson 2012, 213). Although in principle his skeuomorphic design was accessible to anyone familiar with the rules of chess, Hellwig’s wargame expanded the chessboard from a manageable yet complex 8 by 8 grid to an enormous 49 by 33 and added thousands of movable, multicolor cells to designate different types of terrain. Under careful instruction, players began each game with over a hundred pieces adapted from pawns, knights, bishops, rooks, and queens to represent contemporary infantry, cavalry, and artillery with which they captured fortresses instead of slowly moving kings (see Figure 5.4).[19] Encouraged by the game’s modest success among military enthusiasts, Hellwig would continue to expand and modify his kriegspiel for the next twenty years, adding more pieces but retaining the basic mechanics of chess. Whether or not the wars of the eighteenth century could be reduced to a series of discrete commands corresponding to clearly demarcated rules of engagement (or were simply romanticized as such within the historical imaginary of Enlightenment Europe), Hellwig’s war was born on the chessboard and embraced the turn-based strategy and abstract topography of that gridlocked game.

Figure 5.4. As seen in Rolf F. Nohr’s careful reconstruction (right), Hellwig’s original kriegspiel attempted to convert both the board and pieces of chess into a game of military strategy while retaining their original logic (left). Photographs by Ralf Wegemann, 2010, reproduced with permission from Rolf F. Nohr.

Despite Hellwig’s attempts to transform chess into a historically accurate model of combat, the earliest wargames delighted in the visualization of the battlefield as an idealized, Cartesian space and reduced the complex movements of troops to the orders of their commanding officers. “The invention of wargames,” Peterson (2012, 218) explains, “depended on recent improvements to maps, which were, a century before Hellwig, only loosely anchored to the grid of longitude and latitude.” With their predictive powers, eighteenth-century mapping technologies became an “intense interest to the various military powers of Europe” (Peterson 2012, 218). In The Right to Look Nicholas Mirzoeff (2011, 475) also acknowledges that “visualizing was . . . the hallmark of the modern general from the late eighteenth century onward, as the battlefield became too extensive and complex for any one person to physically see.” As such, visualization became a technology for “classifying, separating, and aesthetizing” warfare, business, and, one might add, games (Mirzoeff 2011, 3). The term visuality itself, Mirzoeff (2011, 474) notes, was first coined in 1840 by Scottish historian Thomas Carlyle and was used not to refer to visual perception, but to an imaginative strategy of “making the processes of history perceptible to visual authority.” From the origin of the term, visualization has been an act of harnessing “information, images, and ideas” to assert the power by making legible acts that exist beyond the visible that were previously “too substantial for any one person to see” (Mirzoeff 2011, 474).

Along with the growth of mapping technologies as well as the mathematization and abstraction of vision at the turn of the nineteenth century, new disciplines like statistics and probability became of interest to governmental, economic, and military powers. Although ancient gamblers rolled astragalus bones for play, profit, and prophecy, Ian Hacking (2006, 11) designates the period around 1660 as the moment when contemporary Western concepts of probability emerged throughout Europe. As a result of these scientific developments, “society became statistical” in the nineteenth century and “a new type of law came into being, analogous to the laws of nature, but pertaining to people. These new laws were expressed in terms of probability” (Hacking 1990, 1). Whereas the desire for a more mimetic wargaming simulation established many of the genre’s conventions, it was the design innovations of the Reiswitz family, specifically Georg Leopold von Reiswitz and his son Georg Heinrich Rudolf von Reiswitz, in the 1820s that, according to Jon Peterson (2012, 205), “freed [wargaming] from the abstractions of boards and figurines, as well as primitive conflict resolution mechanisms. . . . These gave way to mathematical principles of probability that decides game events.” The development of wargaming was predicated on the reconfigured relationship to both image and number that was emerging out of contemporary scientific, economic, and military culture.

Whereas Hellwig’s kriegspiel was designed such that “nothing depend[ed] on chance,” the Reiswitzes’ emendations introduced dice and luck-based mechanics as a way of reflecting the indeterminacy of the battlefield (Peterson 2012, 231). And although the Reiswitzes shared Hellwig’s goal of maximizing the verisimilitude of military scenarios, Peterson (2012, 220) speculates that “with a consistent scale, a game ceases to be an abstraction like the game of chess, and begins to evolve towards something entirely novel: a simulation.” The ability to simulate historical events occurred through the application of probabilistic models based on empirical data gathered from studies such as Gerhard von Scharnhorst’s “On the Effects of Firearms” (1813). Von Scharnhorst’s work collated the data produced by firing weapons multiple times packed with differing amounts of gunpowder and ammunition in order to determine the odds of hitting a target. Although the results derived from statistical studies benefited military leaders, it was “less the conclusions than the data itself that would inspire the Reiswitz family” (Peterson 2012, 234). The Reiswitzes imported statistical models and interfaces into their wargames. Peterson (2012, 237) emphasizes that these innovations in simulation and wargaming are not to be underestimated, as the act of “combining empirical probability with implements of chance has no obvious precedent in intellectual history, and represents a paradigm shift that underlies a great deal of the science of simulation that followed in the twentieth century.”

What ties these historical developments in statistics, cartography, and gaming together is the reliance on numerical abstraction—the conversion and compartmentalization of otherwise undifferentiated material phenomenon into discrete quantities to be compared, exchanged, and procedurally operated upon. Following Steven Connor’s (2011, 151) A Philosophy of Sport (last mentioned in chapter 3), all games involve numbers in the form of scoring, and “scoring involves a conflict and convergence between two entirely incommensurable orders, the qualitative syntax of bodily motions and actions (kinesis), and the qualitative calculus of number (ratio).” It is no coincidence, therefore, that statistical forms of organizing space, time, and bodies pioneered in the eighteenth century translate effectively into gamelike simulations. From gridded topographies and mathematical probabilities, the history of wargaming negotiates the ontological tension between measurement and measured. In the twentieth century, tabletop games would further dissociate themselves from military simulation, displacing their historical diegesis with a newly popularized form of literary abstraction: the high fantasy of Dungeons & Dragons (1974).

Beyond their function as a leisure activity for aristocrats and training exercises for military students, wargames found a new audience and narrative storyworld in the twentieth century. Peterson (2012, 109) considers that “the study of wargaming in the late 1960s is the study of a conservative youth movement, a bastion of early post-war values preserved in middle-class suburban America.” As disillusionment with Vietnam steadily increased (particularly within a draftable player base) along with the growth of countercultural movements in America, wargaming received a makeover with the adoption of J. R. R. Tolkien–inspired fantasy settings.[20] Rather than reenacting the theater of World War I or II, orcs and elves replaced American GIs in the games played on tabletops around the world. An early pioneer of fantasy roleplaying, Gary Gygax co-developed Dungeons & Dragons (D&D) with Dave Arneson in 1974. A tabletop game that retained many of the mechanics of the wargame, D&D replaced the historical military context with a fantasy world more palatable to players seeking not to embrace but to escape the realities of contemporary warfare.

D&D is often cited as the first roleplaying game and added elements of storytelling and improvisational theater to the tabletop formula dating back to the eighteenth century. Matthew Kirschenbaum (2011) notes that in the nineteenth century wargames split into “rigid” and “free” kriegsspiel with the “elaborate rules and calculations of the von Reiswitz game” replaced by new mechanics: a human “umpire” (anticipating D&D’s Dungeon Master) who “made decisions about combat, intelligence, and other aspects of the battlefield.” In the twentieth century, not all wargames would become loregames, and, in addition to popular roleplaying games, there were plenty of fantasy-themed rollplaying games developed in the 1970s and 80s. One such example that would be of particular historical significance to the development of Dota 2 was Warhammer: The Game of Fantasy Battles. First published in 1983 by Bryan Ansell, Richard Halliwell, Graham Eckel, and Rick Priestley, Warhammer adapted the miniature tabletop combat of wargaming to the fantasy genre. Small, painstakingly detailed lead statuettes of wizards and warriors replaced traditional infantry and artillery. Although the roleplaying aspect was limited, Warhammer capitalized on a unique visual style that inspired blockbuster videogames like Blizzard Entertainment’s Warcraft: Orcs & Humans (1994), one of the most popular real-time strategy (RTS) computer games in the 1990s.

The RTS genre of videogames adapts elements of the classic wargame to a computational environment. Miniature figures are still positioned around a modular battlefield with the intention of simulating large-scale combat, but many of the gameplay elements are automated and, as such, are far more granular than most tabletop games in terms of space and speed. Instead of placing hand-painted figurines on massive, foam-core topologies, RTS games automate dice rolls, unit positions, and even the passage of time. Henry Lowood (2009, 410) stresses that “in contrast to the turn-based game . . . the states of the [RTS] game occur at specific times in a continuous stream—time flows in an uninterrupted (generally) and synchronous way for both players.” Borrowing many now-common features from Dune II (1992)—one of the earliest examples of modern RTS, often cited as the progenitor of the genre[21]—Blizzard’s first Warcraft title further calcified gameplay around the production and manipulation of military units without the down time or alternating turns in which players could patiently consider each command (see Figure 5.5).

In RTS games, combat takes the form of mouse clicks and relies on manual dexterity to direct troops, expand territory, defeat enemies, and complete missions from an isometric, bird’s- or god’s-eye perspective. In a retrospective article, Patrick Wyatt (2012a), the producer and lead programmer for Warcraft: Orcs & Humans, recalls that the inspiration for the graphic user interface and game controls were based on his experiences with “low-end ‘Computer Assisted Design’ (CAD) programs like MacDraw and MacDraft . . . so it seemed natural to use the ‘click & drag’ rectangle-selection metaphor to round up a group of units to command.” Starting slowly with a focus on basic subsistence and resource management, in Warcraft, players manipulate common metaphors of desktop computing as the main interface for scrolling over a large, multiscreen map while directing peasants to mine mountains and mill trees in order to build a digital fief. The game’s simplified macroeconomics are tailored to support the steady production of military units which, in turn, must be strategically micromanaged against enemy elves or orcs controlled by either human or AI opponents.

Figure 5.5. Blizzard’s original Warcraft (bottom) borrowed many elements from Dune II (top), including its graphic user interface, point-and-click controls, and real-time combat.

The built structures in the Warcraft series range from the familiar to the fantastic with farmhouses, barracks, and blacksmiths sharing the same geographic space with goblin apothecaries, gryphon aviaries, and gnomish ateliers. In his postmortem account, Wyatt (2012a) admitted that “Warhammer was a huge inspiration for the art-style of Warcraft, but a combination of factors, including a lack of traction on business terms and a fervent desire on the part of virtually everyone else on the development team (myself included) to control our own universe nixed any potential for a [licensing] deal.” Even Blizzard’s 1998 follow-up to the Warcraft series, StarCraft, closely mirrored the narrative themes and visual styles of Warhammer 40,000 (1987), the dystopian science fiction game in which miniature models of space marines, noble aliens, and big bugs engage in tabletop combat.[22] From wargaming to Warhammer to Warcraft, the homology between the abstract rules of chess and historical combat in the first kriegspiel evolved through the incorporation of statistics and the application of fantasy in a way that would eventually lead to the procedural operations and RTS of a new genre of digital games like Defense of the Ancients and, eventually, to The Turn of the Tide.

Whereas military strategy and stochastic combat are the ostensible thematics of wargaming (anticipating the emphasis on competitive violence that has come to define both board games and computer games in the following centuries) the abstractions on which these games are based inaugurate what both Steven Connor and McKenzie Wark see as a more primary form of violence or “cut.” As Wark (2007, 023) explains in Gamer Theory, “the real violence of gamespace is its dicing of everything analog into the digital, cutting continuums into bits. That games present the digital in its most pure form are reason enough to embrace them, for here violence is at its most extreme—and its most harmless” (emphasis added).[23] Even the production of oppositional categories like “digital” and “analog” is a binary logic that enacts a discursive, digital violence. Like Wark, in A Philosophy of Sport Connor (2011, 154, 151) argues that “the most general and abstract kind of agon [competition], which characterizes all sports, is that between number and motion, score and play” noting “[‘score’] has as its primary signification, from Old Teutonic sker-, the act of cutting . . . [and] the etymology of the word ‘decision’, from Latin caedere, to cut.” It is not the diegetic or representational violence of mortal combat, grand theft auto, or modern warfare that produces the agony of videogames, but the transformation of continuous play—of mind, motion, materiality, and matter—into “cut” abstractions.

Confronted with the formal and mathematical mechanisms driving these virtual battlefields, human play tarries with the number, the statistic, and the score that Connor calls ratio. Ratio renders otherwise irreducible material components, spatial gestures, or temporal events abstract and exchangeable. The urge to compare scores, to reduce unique moments of play and make them fungible with one another, is also the motor driving what Roger Caillois (2001, 12) calls agon, or competition. Agon lies at the heart of the most ancient games and, not coincidentally, the heart of capital. Following Marx’s distinction between use value and exchange value, Bernard Stiegler’s concept of grammatization applies the philosophical cut between quality and quantity to characterize the operations of both technical media and capitalist economies. As discussed later, the fantasy of a level playing field represented by both fair games and the free market are conflated within Valve Corporation’s “flatland” in which production and play are linked through a quantifying impulse.

With the drama and agony of competition, it is easy to forget that actions and events of both games and commodities must, at some point, be abstracted, or grammatized, into number. In the process, play becomes precisely the act of negotiating the discontinuity between phenomenal experience (or, on the other hand, material substance) and the abstract rules of the game (or market). If all games are constituted in terms of their formal rules, then, as McKenzie Wark (2007, 79) boldly determines, “all games are digital.” “Without exception,” Wark continues, “[games] all come down to a strict decision: out or in, foul or fair, goal or no goal. Anything else is just ‘play.’” Connor (2011, 149) reiterates the distinction between game and play when he writes that “sporting rules and scores always in a sense limit or finitize the infinite possibilities of play.” Whereas play is theorized as a voluntary and unnecessary activity that can never be wholly predicted, measured, or even known, games offer the chance for uncertainty, the units to measure play, and the conditions for unknowing in the form of rules.

If, as Bernard Suits (2005, 55) writes in The Grasshopper, play is the “voluntary pursuit of unnecessary obstacles,” then the player must freely submit to an absolute set of rules and play for the sake of play alone. When play for play’s sake occurs within rules occurring in and of themselves, the game transforms into an aesthetic object and play must be, in a Kantian sense, disinterested in results obtaining outside its magic circle. Both play and game, in this cut up form, entail utopian horizons—two divergent lines of flight propelling human desire. This dream of an immersive, escapist, autonomous, and fantastic gamespace structures consumption and production within the videogame industry. More than merely selling entertainment products, the games industry sells an ideology. As the term game becomes codified and calcified in the form of fetishized commodity, videogames become this ideology’s avatar. Here the practical possibilities of play are expropriated into yet another form of statistical labor, an affective economy privatized within an industry designed from the ground up to capture and mobilize desire.

However, despite these utopian fantasies, play and game do not stand absolutely apart, nor do they precisely correlate. There has never been a game that is absolutely unnecessary, immaterial, and ahistorical. There has never been a player able to resist involuntary action like the process of metabolism or the Earth’s gravity. Since the Reiswitz family, the material and historical friction that has always existed between rules and play has been expressed and allegorized through the inclusion of probabilistic or chance-based mechanics. Connor (2011, 171) writes, “Just as probability can neither be distinguished from nor wholly identified with number, so number can neither be extricated from nor entirely exhaust play. Probability is the play of number that number itself makes possible.” Probabilistic games are, in part, structured around this form of playful unknowing. From wargames to Warhammer to Warcraft, it is probability that most closely resembles the deeply commingled, historical relation between game and play—and the primary operation driving The Turn of the Tide.

After the emergence of wargaming in the eighteenth century, Connor (2011, 150) states, “the nineteenth century was the most decisive era in the formalization of rules, which is to say, the development of explicit codes, standing abstractly apart from the play, and obtaining in all circumstances” and notes that “as the rules became more abstracted from the playing of the game, the regulative role of the official, as the mediator between rule and play, became more important.” Following this trajectory, in the mechanical, electrical, and computational games of the twentieth century the rules of the game cease to behave like social contracts that can be officiated over by an umpire and agreed upon by players and instead are replaced by mechanics. In pinball cabinets, electric toys, and videogames physical forces like gravity, momentum, friction, acceleration, and electricity—along with the infelicities of MacDraw- and MacDraft-style interfaces—replace the rules of the game (sometimes skeuomorphically simulating the rules of analog games, sometimes disavowing rules altogether). In digital games rules become mechanized; in analog games mechanics are often ignored. Whereas the official rulebook of Major League Baseball details the composition of each team, methods for scoring, and the precise geometry of the field, it does not include the gravity of the Earth (see Figure 5.6).[24]

Whereas videogames attempt to reduce all player actions and social rules to their mechanics, exposing the ideological drive to conflate the game with play, metagaming entails the material, historical, and experiential complexities that videogames that the ideological avatar of play rule out. How disabled players negotiate standardized control schemes, how siblings take turns according to house rules, how partygoers drink for each digital death, and even how probabilities predict play are all metagames. Metagaming accounts for those external rules or social customs built in, on, around, and through videogames. The strategies of speedrunners, preferences of pro gamers, and even the tendencies of conspicuous consumers are not what is packaged and sold in stores, distributed on digital networks, or produced by hardworking programmers. Whereas the videogame can be reduced to an abstract set of binary numerals or algorithmic instructions that nevertheless structure the way we play, metagaming constitutes the material histories exceeding the algorithm. And one of the most intricate and influential metagames evolved (and continues to evolve) within Warcraft III is a mod called The Defense of the Ancients.

Figure 5.6. Like Dota 2’s map (right), Diagram No. 1 from the Official Baseball Rules (left) defines a field of play in terms of absolute, geometric coordinates. Unlike Dota 2’s polygonal geometry, however, a physical baseball stadium can never precisely correlate to the abstract diagram (Major League Baseball 2013).

The Evolution of Dota

When Zheng “Faith” Hongda logged in to the second game of the winner’s bracket semifinals, he could not predict exactly what would happen, only what would probably happen. Although Faith could not know what hero he would be assigned, what lane he would be playing, or which Na`Vi players he would be matched against, Invictus Gaming’s de facto “support” player understood what professional players and e-sports enthusiasts call the metagame. The personal preferences and quirks of individual players coupled with the coevolution of common strategies and patch updates is the metagame of Dota 2. Just as with StarCraft, Counter-Strike, and other multiplayer e-sports, Dota 2 competitions are based not simply on the game’s mechanics but on the psychological and probabilistic metagame with which teams wage war against one another: strategy, counterstrategy, counter-counterstrategy, and maybe back to square one to surprise or confuse an opponent.[25] Dota 2’s metagame is perhaps most evident in the draft phase of the game. In the draft, each team’s captain begins the game by banning and picking not players, but characters from the game’s massive roster.[26] At the time of The International in 2012, there were ninety-two heroes, each with diverse attributes (from physical size and movement speed to stats like agility, strength, and intelligence) and, importantly, at least four special abilities. Ninety-two heroes with about four abilities each who can use one hundred and thirty-eight items offer an almost inconceivable range of statistical variation (not to mention the specific strategies executed in real time within a match) (see Figure 5.7). With the draft, Dota 2 offloads game design decisions like the balance between various hero combinations to the players, as they collectively design—via the ever evolving metagame—what kind of Dota 2 they will play.

Before The Turn of the Tide, iG planned to follow that summer’s dominant metagame in which Naga Siren and Dark Seer were regarded as an overpowered pair able to unleash a game-winning “wombo combo” by combining their particular abilities (Valve 2012c). The Siren-Seer metagame is all too evident in the statistics for The International. Out of ninety-two possible heroes, the top three banned and picked throughout the entire tournament were Dark Seer at 155 times and Naga Siren (tied with Rubick) at 152.[27] It is not a coincidence that these heroes saw the most action during the draft phase of The International in 2012. Captains used their limited number of picks and bans to prevent the competition from lining up overpowered combinations of heroes. Knowing all this, it must have looked odd when Na`Vi, seemingly unfazed by the loss of Game 1, did not pick or ban either Naga Siren or Dark Seer. Clearly disappointed at the moment iG drafts both heroes, English commentator “LD” Gorman explains:

For anybody who doesn’t know Naga’s Song of the Siren obviously puts everyone to sleep. You can then Vacuum them in [with Dark Seer] while they’re asleep, drop [Dark Seer’s Wall of Replica], unleash the pain train. . . . Oh god, if they give up the Tidehunter. How do you take a fight? You don’t take a fight. (Valve 2012c)

Considering their swift defeat in game one, Na`Vi’s picks looked suicidal. Yet, the footage of the team gathered together in their glass box shows the Ukrainians grinning ear-to-ear and giggling mischievously. LighTofHeaveN and XBOCT crowd around their captain’s computer as Puppey picks the Juggernaut, an unconventional hero who was banned and picked a mere ten times during the entire tournament. With Na`Vi’s third pick out of the way, there was only one option for Faith, iG’s support player: the Tidehunter.

Figure 5.7. As Dota’s heroes (top) gain experience and gold throughout a game, they can learn a host of spells (middle) and purchase a plethora of items (bottom). Images by sweetlikepete, 2013 (top); BnJx, 2013 (middle); and Patrick LeMieux, 2016 (bottom).

Evolved from the polygonal Sea Giant Hunter in Blizzard’s Warcraft III, Valve’s version of the Tidehunter is a bipedal fish that carries a fish (see Figure 5.8). While not the strongest or fastest hero in Dota 2, in 2012, Tidehunter was a very defensive pick who could be used to support fragile teammates and buy them time to level up and get some gold. Beyond his early game utility, Tidehunter also has a powerful ultimate ability called Ravage that can stun a whole team if they are clustered together and is often used to initiate large 5 vs. 5 battles called team fights. Having seized the opportunity to select an unbanned Naga Siren and Dark Seer, Faith’s fate seemed inevitable. In this tournament, in this series, in this match against this team, and with this specific metagame, how could picking Tidehunter go wrong?

Figure 5.8. When compared to the original Tidehunter (aka Leviathan) from Warcraft III (left), Valve’s version (right) sports a graphic overhaul while retaining the exact same abilities: Gush, Kraken Shell, Anchor Smash, and Tidehunter’s ultimate ability, Ravage.

Whereas iG picked Tidehunter in accordance with the strategic metagame in the summer of 2012, Dota 2 itself is also a game about another game—a mod originally designed within and disseminated through Blizzard’s Warcraft III. Beyond appropriating content and components from Warcraft III (a fan-made way to play with Blizzard’s IP), Dota[28] also represents a hybrid videogame genre that recontextualized and reinvented RTS game mechanics within the ecology of multiplayer online games in the early 2000s. Like Counter-Strike (1999), Team Fortress (1996), and Day of Defeat (2003) before it, Valve has built a business on converting player-produced metagames into service platforms that flatten leisure activities alongside other forms of labor. Whether playing matches, competing in tournaments, talking on forums, analyzing statistics, authoring tutorials, designing cosmetics, debugging beta software, modding new mechanics, or even working as an employee at Valve, Dota’s users operate according to a decade-long metagame whose origins are notoriously difficult to identify. As Nicholas Werner’s documentary series All Your History aptly puts it, Dota

came from nowhere with no corporate backing, no advertising campaign, and no business model. It’s worth millions of dollars yet only a handful of studios have done any work with it. It’s a genre, except it’s a game, except there are dozens of them, except it’s just a mod. It’s a brand new word, except it’s an acronym that is trying to be trademarked by its founder, except there is no founder. It’s been called MOBA, it’s been called ARTS but to the millions of players around the world who play it, it’s just Dota. (Werner 2012)[29]

In All Your History, and in games journalism more broadly, Dota is presented as a historical anomaly, a mutant mod with evolutionary origins outside the umbrella of the games industry. The industry—a complex ecosystem of corporate stockholders, development studios, trade shows, global distribution companies, retail vendors, and both independent and completely dependent journalists—attempts to subsume much of the history of videogames (and games in general) within its sphere of influence.

But if “all your history belong”[30] to a linear model of production based on the rigid schedules, intellectual property rights, sales figures, and numerical ratings surrounding discrete pieces of software, Dota 2 arrives on the scene as an outsider, its historical origins obfuscated by its genesis from the seabed of collective play rather than corporate pay. It is no surprise that most major videogame journals, magazines, and blogs did not cover Dota’s longer history as a popular mod, a competitive sport, or even its public beta, preferring instead to write articles only after its “official” release by Valve on July 9, 2013 (Valve 2013b). Although Dota is a mutant, it did not spontaneously emerge. From wargaming the statistical and probabilistic mathematics of the eighteenth century to restaging tabletop combat according to the logic of fantasy literature to automating the complex calculations of role- (and roll-) playing games with digital computers, Dota’s evolution can be traced back through time though it has only recently been absorbed within the videogame industry proper. As such, Dota represents a remarkably disruptive and deconstructive entry in the history of games, upsetting the stable, smooth, and simply chronological timeline so often deployed by videogame companies to further the ideological avatar of play—what Erkki Huhtamo (2005, 4) calls the “chronicle era” of videogame history.

Based on the records of software revisions and the firsthand accounts of fans, Dota’s ancient ancestors first crawled up onto land from the primordial ooze of the RTS genre in the early 2000s.[31] Life didn’t spring from the software itself, but from player-designed games made within Warcraft and StarCraft. These early metagames were distributed in the form of unit mods and map packs uploaded to online forums or shared across peer-to-peer networks like Blizzard Entertainment’s[32] One such custom map, Aeon of Strife (AoS) by Aeon64, was produced with StarEdit, a standalone interface for building levels and designing new campaigns shipped with StarCraft and its expansion, StarCraft: Broodwar (1998).[33] Instead of echoing StarCraft’s standard gameplay by focusing on the real-time production and deployment of a large military force, AoS featured four individually controlled “heroes” protecting defensive structures or “towers” against endless waves of artificially intelligent enemies or “creeps” that traveled down four “lanes” connecting the two opposing bases (CtChocula 2011) (see Figure 5.9). Featuring a seemingly counterintuitive mixture of gameplay borrowed from RTS, action adventure, roleplaying, and tower defense games, AoS instituted many of the defining (albeit nascent) features of Dota: the map, the mechanics, and the mission.[34] A game of AoS would end when the players were able to offset the balance between the otherwise equally matched AI units and destroy the opposing team’s base. Yet the mod was underplayed. The bizarre mixture of mechanics would not prove viable until remade in another game engine: the Warcraft III World Editor (2002).

Figure 5.9. Compared to an early four-lane tower defense map like Aeon of Strife (top), the onscreen mini-maps from Dota 2 (left), Heroes of Newerth (middle), and League of Legends (right) reveal a very similar, symmetric design for team versus team gameplay. Image by Softmints at, 2013 (top).

Warcraft III: Reign of Chaos (2002) continued Blizzard’s trend of releasing limited versions of their in-house development tools for fan-based game design. Like StarEdit before it, Blizzard’s Warcraft III World Editor included tools for modifying terrain, objects, and triggers as well as an Import Manager which allowed players to import their own art assets like sounds, textures, and 3D models (Feak and Mescon 2009). Furthermore, the World Editor’s triggers made use of JASS, “just another scripting syntax,” which afforded users the flexibility to both choose their development environment and dabble with more complex aspects of game design than was possible with the utilities released with Warcraft II and StarCraft.[35] Fans recall that “in 2003, the game changed. . . . Blizzard released Warcraft III, along with its editing tools, and modders went crazy. . . . [Use Map Settings] had become prolific and fun, with games like Lurker Defense and Zone Control” (Dimirti 2013). In 2003, around same time that other AoS-inspired games (like The Valley of Descent by Karukef) were being produced, a modder named Eul remade Aeon of Strife in Warcraft III’s more powerful engine.

Titled Defense of the Ancients, or DotA, Eul’s mod expanded AoS’s basic formula, adding more players and more mechanics, some inspired by Warcraft III itself. Blizzard’s game featured RPG-inspired “hero units” which gain experience, level up, and collect items alongside mass produced, disposable troops spawned from production facilities and controlled within a typical RTS setting. Importantly, the original DotA focused more on player-versus-player (PvP) competition than AoS by allowing two teams of five players aided by AI rather than a single set of players working together against a computerized opponent. The goal of the game remained the same: to upset the balance of the incessantly marching creeps, exploit tactical advantages to overtake towers, and destroy the opposing team’s “Ancient,” the largest structure at the heart of the enemy base. In Eul’s mod, teams of five players sparred against one another for experience, earned gold by “last hitting” enemies, purchased items to boost their abilities, and engaged in team fights. With these basic elements added to AoS, DotA began to gain steam. The mod became so popular that Eul’s follow-up project, DotA: Thirst for Gamma, could not even compete with remakes of his previous game within Blizzard’s Warcraft III: The Frozen Throne (2003) expansion. Soon after The Frozen Throne was released on July 1, 2003, Eul retired from the modding community but not before publishing the JASS code for the original WCIII DotA, code that would inspire an explosion of DotA games like the DotA DX Series, DotA Unforgiven, DotA Mercenaries, and DotA Outland (CtChocula 2011).

With The Frozen Throne expansion and Eul’s open source code, DotA began to multiply. In 2003, there were dozens of DotAs and AoSes as well as DotA-clones, DotA-likes, DotA-based, and DotA-esques. With the growing popularity of the World Editor and the number of players selecting “use map setting,” served as the Cambrian sea, incubating more and more versions of DotA with new heroes, spells, items, and maps. With so many tweaks and changes being tested out within this ad hoc, player-driven development community, compilations emerged over time featuring the most popular mechanics and hero units borrowed from dozens of different games. Some of the earliest attempts to build the ur-DotA were Meian and Ragn0r’s various Allstars titles in 2003 that compiled popular heroes from across the DotA-verse into one downloadable map launched on February 3, 2004 (CtChocula 2011). Following these mods, Stephen “Guinsoo” Feak began to work on DotA Allstars 2.0 in February, and released 3.0 in March that year. More than simply building a clearinghouse for favorite heroes, items, and abilities, Feak took on the mercurial task of consolidating, tweaking, balancing, and debugging the game with a loose team of modders who designed many of the features that have come to define the genre.[36]

Between 2004 and 2005, Feak oversaw the development of more than fifty heroes and over fifty items, a system for item creation and combination based on recipes, the randomly appearing “runes” that temporarily alter hero abilities, “super creeps” that spawned after the destruction of all barracks, a large and centrally located neutral creep named “Roshan,”[37] and the addition of booming “kill streak” voiceovers borrowed from Epic’s Unreal Tournament 99 and 2004 (e.g., “FIRST BLOOD,” “OWNAGE,” “M-M-M-M-MONSTER KILL,” etc.), signature sound effects that still exist in Valve’s version to this day (Feak and Mescon 2009). Beyond these now familiar features, Feak’s most important contribution was guiding the team of amateur developers to balance heroes for competitive player-versus-player competition rather than focusing strictly on tower defense and lane pushing against AI opponents. As Feak and Steve “Pendragon” Mescon (2009) explain in their postmortem account of DotA Allstars, “balance was tuned each version, and the only way for [Feak] to know if he got it right was to see if the players stopped complaining.” To further organize the player base, Mescon built a set of forums that “became the main medium of communication between Feak and the community, and contained everything from player guides to player-created hero and item suggestions, and the organizations of the first real DotA tournaments” (Dimirti 2013). Feak worked on DotA Allstars until version 6.0 when he left the Warcraft III modding scene, it is rumored, to focus on playing Blizzard’s new blockbuster, World of Warcraft.

In many ways, DotA could be figured not only as an evolutionary mutant but also a missing link between two distinct species of game represented in Blizzard’s Warcraft franchise: RTS games like Warcraft III and massively multiplayer online (MMO) games like World of Warcraft.[38] Between these two major bookends of the Warcraft series, DotA wrested Warcraft III’s mechanics and point-and-click interface away from the macroeconomics and micromanagement of the RTS genre to create a different kind of collaborative and competitive online play. But rather than building a game around leveling a single unit for cooperative player versus environment (PvE) challenges like World of Warcraft, DotA focused on manipulating individual units in a team-based, PvP setting.[39] For a number of reasons, the historical origins and evolutionary influences of DotA have been obfuscated—within corporate-, industry-, and even the player-created histories surrounding Blizzard’s franchises. Because the game emerged from a community of modders rather than the games industry proper, for many years DotA was undervalued in terms of both economic and cultural capital. In an era of MMOs dominated by games like World of Warcraft, DotA continued to flourish, particularly in non–North American contexts. After the newfound euphoria for the MMO and a rocky shift of power, Neichus picked up where Feak left off before handing Allstars over to IceFrog, who continued to labor in relative obscurity (and without pay) to maintain DotA.

Not much is known about IceFrog, the elusive and longest-serving developer of DotA who began modding the game with Feak and Neichus in October 2004. The first new hero shipped with IceFrog’s input in Allstars 5.74 that year was none other than the Tidehunter. According to Michael Walbridge (2008), he “may be named Jeremy . . . may be from Boston and . . . may study at UCLA” and may have been born in 1983. IceFrog may be many things, but to the DotA community he is a mysterious and “almost spiritual figurehead,” responsible not only for coding and distributing the game but for organizing a team of beta testers in order to playtest and adjust the most arcane and nuanced aspects of the game (Walbridge 2008). Like Tarn “Toady One” Adams, Marcus “Notch” Persson, and Dean “Rocket” Hall, IceFrog belongs to a class of game developers who operate as virtual landlords responsible for maintaining games such as Dwarf Fortress (2006–), Minecraft (2009–), DayZ (2013–), and, of course, DotA for digital renters who pay with their attention, time, and money as the game is developed. Whereas Eul propagated the explosion of experiments with his JASS source code and, following the first Allstars mods, Feak carefully selected features which would come to define DotA, IceFrog introduced a half-decade-long, glacial age in which the mod stabilized and hardened around a centralized director, a unified design philosophy, and an international scope. The long evolution of DotA as a species of software slowed as player productivity migrated from the creation of ever-multiplying mods to the culture of competitive play. While Neichus co-authored a few versions of the map starting with 6.0, version 6.1 ushered in the era of IceFrog as the game began “to take large leaps toward competitiveness” (CtChocula 2009).[40] This moment in DotA’s history also marks the introduction of China to the competitive scene as another modder, Heintje, translated the game into Chinese for the first time.

With a stable, international version of DotA Allstars available in 2005, version 5.83B appeared in the Malaysian and Singaporean World Cyber Games while 6.12B was featured as a side event at the World Cyber Games championships in the Suntec City convention center in Singapore alongside corporate titles like Warcraft III: The Frozen Throne and StarCraft: Broodwar. In 2005, international teams were beginning to treat DotA like a sport and with its aging technical requirements the mod could be played on almost any personal computer, from the newest gaming laptops to the beige towers and CRT screens of the late nineties. It was relatively easy to find (or pirate) a cheap copy of Warcraft III in the mid-2000s, and at this point DotA’s gameplay was so diverse that the recombinatory potential of the game far outpaced any one player or team. In 2009, IceFrog estimated between five and eight million new downloads, especially in Asia and South America, and by 2010 “it [was] roughly estimated . . . that the Chinese DotA audience is about 40–50 percent of the worldwide audience. Not counting China, the player base is estimated to be somewhere between 7–11 million” (IceFrog 2009a, 2010). With a thriving community and informal global economy, it was only a matter of time before DotA moved from the realm of play to that of profit.

In the late 2000s, while DotA was emerging as “arguably one of the most popular game mods of all time,” the corporate game industry began co-opting the map, mechanics, interface, and even specific statistics of each hero in order to simulate the original mod (Feak and Mescon 2009).[41] Whereas Feak and Mescon were hired by Riot Games to design League of Legends (2009), IceFrog assisted S2 Games during the development of Heroes of Newerth (2010) before he and Eul, the original creator of Defense of the Ancients, were recruited by Valve. Rather than design a new game with original characters, edited map, and different items (like League of Legends) or attempt a faithful remake of the Warcraft III mod with similar frame rates and polygon counts (like Heroes of Newerth), Valve tasked IceFrog and Eul with translating DotA to their Source Engine. With Dota 2, vestigial features that arose due to the eccentricities of the Warcraft III World Editor become deliberate game design choices. As Chris Thursten (2013, 44) notes in the first formal announcement of Dota 2, “several unintentional side-effects of Dota’s design have become fullblown features” and “bizarre behaviour” like creep stacking, pulling, and juking that developed based on the limitations of the original mod have become defining mechanics of a whole genre.[42] Even in 2015, IceFrog continued to develop Dota 2 within the Warcraft III World Editor before translating game design decisions into Source. And although IceFrog’s identity remains a mystery, it is now believed that he may be living in Seattle, may have turned 30 in 2013, is still stewarding DotA Allstars (version 6.83d was released on September 6, 2015), and is employed at a games and digital distribution company known for its experimental and unorthodox approach to game design, economics, and labor organization.

At the end of Defense of the Ancient’s long evolution, Valve filed a trademark application for the name Dota in 2009. What resulted was a series of countersuits initiated by Riot Games and Blizzard Entertainment—corporations made up of ex-modders with vested interests in the future of the game. Under the aegis of their old company, Dota Allstars, Riot employees Feak and Mescon filed for the Defense of the Ancients trademark in order to “protect the work that dozens of authors have invested to create the game and on behalf of the millions of DotA players all over the world” (Augustine, 2010). Blizzard, on the other hand, argued that “by attempting to register the mark DotA, Valve seeks to appropriate the more than seven years of goodwill that Blizzard has developed in the mark DotA and in its Warcraft III computer game and take for itself a name that has come to signify the product of years of time and energy expended by Blizzard and by fans of Warcraft III” (Plunkett, 2012). Whereas both Riot and Blizzard appealed to the rights of the community of modders and players, each planned to protect the name of the game for its own interests.[43] In the end, though, Valve settled out of court and agreed to forgo both Defense of the Ancients and DotA, instead trademarking Dota 2—a sequel to a Dota 1 that never existed. The land grab that took place between 2010 and 2012 in the form of these trademark lawsuits demonstrates one way that corporations enclose the metagame. It was not until Valve’s attempt to trademark the name Dota that those companies directly responsible for the evolution of the mod began to value the longer history of community play and production in, on, around, and through their games.

Following Valve’s settlement with Blizzard Entertainment, the corporate struggle shifted from the name of the game to the name of the genre. Despite its generality, Riot promoted the acronym MOBA (multiplayer online battle arena) as a way to distance their product, League of Legends, from Dota after Valve acquired the proprietary eponym. In opposition to Riot’s acronym Valve offers ARTS, a slightly more descriptive if less common term because, as Gabe Newell (Nutt 2011) states, “action RTS seems to make a lot of sense to customers. If you say that, they have a pretty good idea what you’re talking about. I don’t even know what MOBA stands for.” The debate over nomenclature and lack of consensus on how to define this genre of 5 vs. 5, lane pushing games stands in as a metonym not only for the growing pains of DotA and its ilk, but the difficulty of institutionalizing and branding a mod. Even the transition from the acronym DotA to the word Dota reified the game as a commodity and relocated the metagame from early communities of play built on the exchange of unique designs to a service platform for cosmetic markets, e-sports, and emerging forms of spectatorship.

The technical evolution of Dota began in the antediluvian wake of RTS games and, after surfacing as one of the most successful mods of all time, generated an entire genre of competitive play. From a mutant mod to the genesis of a genre to a spectator sport and international marketing tool and cottage industry of user-generated content, Dota 2’s evolution outside the industry extends the horizon of videogaming as a medium, encapsulating previously undervalued ways to play and converting them into organs of productivity and profitability. Dota’s evolution grinds to a halt precisely at the point at which the game’s design is at its most general and most open to divergent possibilities defined not by what is materially present in the software, but by the diverse play made possible by its noisy, entropic systems. Under the management of IceFrog and newly funded by Valve, Dota 2 is less a sequel and more of a service platform, expropriating the metagaming practices that have long supported the game as a global sport. In biology, the term sport refers to organisms that have undergone dramatic mutations, and in the early years of Dota, many “sports” erupted out of the recombinatory potential of early network culture. Yet, the calcification of the play into a professional sport—specifically what has come to be known as e-sports—marks a moment when the evolution of Dota stabilizes and moves away from the production of software to culture.

From Sports to E-Sports

Well beyond the evolution of Dota and its recent acquisition by Valve, the origins of e-sports stretch back to the development of the earliest electronic games. In Raising the Stakes: E-Sports and the Professionalization of Computer Gaming (2012) T. L. Taylor offers a rich history of competitive videogaming and documents a half-century of attempts to position computer games within the arena of spectator sport. In one respect, games have always engaged agon, the cut between ratio and kinesis. The technical operations of digital media automate this kind of cutting as switches, transistors, and semiconductors cleave binary bits of data from otherwise analog pulses of electricity. Even at the beginning, videogames conflated these two related forms of agony, allegorizing their digital processes of grammatization, individuation, and quantification via the cut of competition. Tennis for Two (1958) was, of course, for two players, and Spacewar! (1962) inspired the “Spacewar Olympics,” a four-hour tournament at the Stanford Artificial Intelligence Laboratory (SAIL) sponsored by Rolling Stone magazine in 1972—perhaps the first e-sport event.[44] Later, single-player games continued to incorporate player-versus-player competition like Sea Wolf (1976), an early arcade game with the first persistent scoreboard. Although individual players played early videogames games serially, scoreboards added a social component, a historical record that could be competed and compared with over an extended time period. As Taylor (2012, 4–5) recounts, in the early eighties there were televised game shows in which players competed against each other, such as Starcade (1982–84), that stand as further attempts to convert videogames into a spectator sport. These two traditions manifest in two distinct types of sport: the player-versus-player competitions of the strategy, fighting, and first-person shooter communities and the player-versus-records of the speedrunning, high-scoring, and some MMO communities.[45]

According to Taylor (2012, 36), “long-standing debates about what constitutes a ‘real’ sport intersect deeply held notions about masculinity (and femininity), class, and culture . . . Only the most naive and ahistorical would suggest the coveted legitimacy of sport is bestowed objectively, outside of any deep cultural values about what constitutes meaningful human and social action.” Rather than engage in a semantic discussion of sport versus game, Taylor observes the way in which computer gaming communities have created a competitive architecture that follows many of the same patterns as professional sports: the emphasis on athleticism and the embodied actions of the player, the importance of standardized equipment, the negotiation of both written and unwritten game rules, the development of professional tournament rules (what Steven Connor [2011, 150] calls “metarules”), gambling and betting on the outcomes of matches, and even the proliferation of scandals in which e-sports players collude to throw important games for profit.[46] From the growth of corporate sponsorship to the broadcasts on television stations like TBS and ESPN to the U.S. government issuing “professional athlete” visas when players compete abroad, the perception of videogames as a sport is less of a philosophical question than a social, legal, and political issue with the most important factor contributing to the changing perceptions of competitive videogaming being the realization of its profitability.

From another angle, all sports are e-sports. Taylor (2012, 40) emphasizes that “the sporting body has always been tied up with technology,” but, more importantly, that the precondition for sports is the digitization of analog action—the conversion of continuous motion into discrete metrics for scoring and judging.[47] Based on the primacy of this relationship between kinesis and ratio, the homology between sports and e-sports can be inverted: rather than seeking to legitimize videogames as sport, it is important to recognize the computational kernel driving all sporting competitions.[48] “From the start,” Wark (2007, 079) suggests, all “games were . . . proto-computer—machines assembled out of human motion, inanimate materials, and the occasional dubious call by the referee.” As much as videogames may be sport, all sports are also videogames. Undifferentiated and irreducible physical fields, material equipment, human bodies, and other forms of play are reduced to a calculation, a score, a cut. Moreover, this form of abstraction that marks one player a winner and the other a loser could be said to go by another name: capital.

As evidenced by The International in 2012, the burgeoning movement of electronic sports models itself closely on the culture and business of professional sports. Major League Gaming (MLG) adopts the branding idioms and iconography of other national leagues such as Major League Baseball (MLB), the National Football League (NFL), the National Basketball Association (NBA), and the National Hockey League (NHL) (see Figure 5.10). Games like Dota 2 inspire international tournaments, lavish prize pools, diverse media coverage, lucrative sponsorship deals, live commentary in dozens of languages, and communities of fans who tune in weekly not to play, but to watch their favorite videogames. The discourse undergirding the administration, advertising, and dissemination of e-sports co-opts myths of human exceptionalism, the pleasure of visual spectacle, and the monetization strategies of professional sports in order to orient audiences toward alternative forms of play based on watching computer games.[49] On the other hand, contemporary sports like football, basketball, and baseball have adopted digital technologies for broadcasting, officiating, and otherwise abstracting the movements of players into exchangeable numbers for transmission, consumption, analysis, and prediction. The mathematical probabilities governing play within MLB, for example, extend well beyond the diamond and the dugout to include marketing, management, and, most of all, money. The more money saturates the culture of videogaming, the more sportslike it becomes. After all, money is the ultimate referee and one of the first digital medium.

Figure 5.10. Mirroring Jerry Dior’s red and blue Major League Baseball logo (left), Major League Gaming (right) has added a curved, skeuomorphic highlight popularized by Apple’s Aqua interface in the late 2000s. The subtitle uses a square sans serif font, connoting the pixelated graphics of videogames. These small changes adapt the iconography of national sports to a digital medium.

When IceFrog (2009b) announced his collaboration with Valve to port Defense of the Ancients from Warcraft III World Editor to the Source Engine in 2009, he reassured the game’s global fan base that the “goal and top priority in the future is to solve the surrounding issues that affect the DotA experience in order to allow it to reach new heights” (emphasis added). While “gameplay, mechanics, and in-game feel” were still important to the developer, Valve offered IceFrog a once-in-a-lifetime opportunity to change not the way Dota was played but how it was distributed, spectated, networked, and monetized in relation to the larger Steam marketplace. By treating Dota 2 as a platform, Valve has not only developed experimental forms of biometric and statistical analysis that expropriate attention and create new ways of “seeing,” but has also built spectatorship into the game as a form of play. With Dota 2, spectatorship is no longer a superfluous byproduct of gaming, but a productive part of a much broader ecology in which all player activities are measured and funneled back into the development of the game.

In a report titled “Gamers on YouTube,” Google researchers James Getomer, Michael Okimoto, and Brad Johnsmeyer undertook an analysis of the company’s video sharing service to demonstrate how watching footage of videogames has become a market in and of itself. Their study shows how YouTubers watch games as much as play them and that the “growth rate of time spent viewing gaming videos was greater than YouTube’s overall growth in the U.S.” (Getomer et al., 2013, 2). What is striking about “Gamers on YouTube” is its resolute focus on one specific model of value derived from the relationship of view count to game sales. Rather than valuing individual views as such, view count serves as an index of the indirect monetary benefit derived from the viewer’s attention. Turning toward mainstream videogames, Google reports that “82% of game console sales occur in first 4 months” and the top ten best-selling console games represented “49% of 2012 total console game revenue” (Getomer et al. 2013; 6, 7). Focusing on these games, when Getomer, Okimoto, and Johnsmeyer (2013, 7) “compared all pre-launch video views . . . for the top 2012 games to their sales in the first four months and a compelling correlation of 0.99 emerged.” This marketing study suggests that YouTube serves as an accurate forecasting device, predicting the “correlation between views and sales” of videogames (Getomer et al. 2013, 7). However, Google’s analysts fail to see the more complex relationship between consumption, production, and attention at stake in the widespread transformation of spectatorship online.

Watching is not merely an index of revenue (or of play) but a form of capital accumulation based on attention. As Christian Fuchs (2012, 704) writes, “On Facebook, YouTube, Twitter, etc., all consumption time is commodity production time.” Although selling consumption itself is the primary condition of social media and user-generated content in general, nowhere is the attempt to capture previously undefined sources of value and productivity more evident than in Valve. The reconfiguration of modes of spectatorship in games like Dota 2 is just a small part of a much larger project of redefining the systems of value that circulate around gaming and recasting play as a form of productivity within a flattened corporate gamespace. Whereas the Google study focuses only on how viewing videogames correlates to future sales, Valve’s model of valorization and evaluation assumes that spectatorship and, more generally, all forms of play are, in themselves, valuable commodities. The problem is how to capture these forms of undervalued production.

With Dota 2, Valve has invented increasingly complex and compelling ways to augment, gamify, track, and play with spectatorship. E-sports flourishes in a complex media ecology designed around converged modes of playing, laboring, and watching. In “The Cinema of Attraction,” Tom Gunning (1986, 70) observes that “every change in film history implies a change in its address to the spectator, and each change constructs its spectator in a new way.” Even at the first International tournament in 2011, a publicity stunt in which Valve unveiled Dota 2 by inviting teams to first learn then compete for a million dollars in a game they had never played, commentators (and viewers at home) were treated to a suite of in-game analytic tools that were far more advanced and centralized within the game’s client than those offered in other e-sport platforms or streaming services such as From simple gold and experience graphs to visualizations of galvanic skin response units hooked up to competing players, reviewer Chris Thursten (2013, 41) remembers this shift from individual virtuosity to statistical inevitability as “commentators [were] just as likely to get excited about a character pulling ahead on a graph as they are a particularly skillful play.”

In 2012, Valve introduced further means to quantify and gamify watching by offering 99-cent “digital pennants” that fans flew in hopes of both increasing the viewer count for their favorite team—a metric clearly visible to competing players in-game—and increasing the odds of winning cosmetic items and other digital commodities distributed to random viewers with each “FIRST BLOOD” or “ULTRA KILL” achieved by their respective teams (shouted by the familiar voices of Unreal Tournament). These light gambling mechanics were extended outward to social media platforms and video streaming providers that, if synced with a player’s Dota 2 account, will also reward that player with in-game giveaways outside the game. Beyond the virtual pennants, 2013’s Interactive Compendium was initially conceived as an up-to-the-minute digital almanac that aggregated statistical information about the players, teams, and matches of the third International (Valve 2013a). Described as a “living document” and “virtual passport,” the Compendium allowed viewers to assemble automated fantasy leagues, collect virtual trading cards of their favorite Dota 2 players, enter into lotteries for exclusive items, and even vote on two dream teams that would compete face-to-face in a local exhibition match. Whereas part of each 99-cent digital pennant sale went to each respective team, 25 percent of each ten-dollar Interactive Compendium was added to the tournament’s prize pool—a sum crowdsourced from $1.6 million to over $2.8 million after fans bought 509,752 copies of the e-book. This number exponentially increased in the following years, expanding Valve’s $1,600,000 prizepool to $10,931,105 in 2014, $18,429,613 in 2015, and $20,770,460 in 2016. Beyond the hype of these astronomical figures and despite myriad analytical tools, players and spectators alike had trouble decoding The Turn of the Tide.

Seventeen Seconds of Dota 2

Returning to Benaroya Hall and the second match of the winner’s bracket semifinals at The International in 2012, a question still lingers: What exactly happened here? After the glow of particle effects, the standing ovation from the Seattle crowd, the stunned silence of the Chinese commentators, and Na`Vi’s improbable solution to that summer’s metagame, spectators new to Dota 2, e-sports, or LAN tournaments must have wondered how iG’s winning strategy failed and what mechanics Na`Vi had exploited to turn it around. Did the virtuosic play of pro gamers reverse the outcome of the match, or did the deterministic, microtemporal processes of technical media overshadow their agency? And how do these seventeen seconds play into Valve’s corporate interests in China? Of course, the answer is in the metagame. Along with narrative explanations and MS Paint schematics scattered across various online forums, two alternative forms of spectatorship surfaced in the months following The Turn of the Tide: a one-thousand-frame-per-second (FPS) slow motion replay and a ‘pataphysical performance of strange statistics.

Shortly after videos of The Turn of the Tide made waves across the Internet in August 2012, Michael Krukar typed “host_timescale 0.001, host_framerate 1000” into the Dota 2 command line interface before replaying the .DEM or “demo file” of Na`Vi and IG’s historic match.[50] Instead of a real-time record of the second game of the winner’s bracket semifinals, what resulted was a microtemporal version of The Turn of the Tide. Like high-speed film appears in slow motion when played back at thirty or sixty frames per second, Krukar rendered the sub-second record of mouse clicks and key presses in Dota 2 at a thousand frames per second. Whereas the commentators at The International could not cope with the density of pixels, polygons, and particle effects erupting for the seventeen seconds the two teams collided, the glacial images of Krukar’s decelerated replay tell a different story. Krukar’s one thousand FPS animation obfuscates the ecstatic narratives produced in the heat of The International in favor of the aesthetic genre associated with slow motion cinematography. Accompanied by the French electronic music group M83’s ostentatiously titled “Lower Your Eyelids to Die with the Sun,” the sequence channels the majestic and melodramatic images of Ron Fricke’s cinematography and the repetitive sonic landscapes of Phillip Glass’s minimalist compositions in Koyaanisqatsi (1982), a film in which techniques like slow motion and time lapse cinematography are used to signal temporal events operating outside the range of human sensation. The statistical information stored in Dota’s .DEM files begin to tell a different story as the procedural rhythms of Valve’s Source Engine appear from beneath the visual spectacle in Benaroya Hall. Instead of illegible explosions of frenetic energy, the mouse movements and button mashing of professional players are organized according to the serial mechanisms and statistical record of The Turn of the Tide. As avatars float calmly from coordinate to coordinate, The Turn of the Tide drifts in and out according to the pull of the Dota’s probabilistic mechanics as much as the pull of human will.

Krukar’s evocative animation points to a deeper source of information: the record of inputs within Dota 2’s .DEM. Using a tool produced by Bruno “Statsman” Carlucci, these files may be mined for the statistical data and textual description of any given Dota 2 match. With 17,463 lines of logged combat events, the spectacular narrative of Ukrainian underdogs facing the impenetrable wall of Chinese Dota may be replaced by an exact set of serial events. Returning to the postplay confusion surrounding The Turn of the Tide, the textual inscriptions reveal a discrete sequence of recorded events from the moment iG’s Chen “Zhou” Yao initiates the team fight to the moment Na`Vi’s Dendi turns the tide:

  1. 29:06 (17:35). Juggernaut loses the modifier_naga_siren_song_of_the_siren.
  2. 29:06 (17:35). Shadow Shaman loses the modifier_naga_siren_song_of_the_siren.
  3. 29:06 (17:35). Rubick loses the modifier_naga_siren_song_of_the_siren.
  4. 29:06 (17:35). Enigma loses the modifier_naga_siren_song_of_the_siren.
  5. 29:06 (17:35). Enigma gets the Black King Bar Immunity.
  6. 29:06 (17:35). Enchantress loses the modifier_naga_siren_song_of_the_siren.
  7. 29:06 (17:35). Juggernaut gets the Blade Fury Buff.
  8. 29:06 (17:35). Rubick gets the Force Staff Push.
  9. 29:06 (17:35). Rubick loses the Force Staff Push.
  10. 29:06 (17:36). Shadow Shaman gets the Ravage Debuff.
  11. 29:06 (17:36). Enchantress gets the Ravage Debuff.
  12. 29:07 (17:36). Tidehunter deals 187 damage to Shadow Shaman using Ravage.
  13. 29:07 (17:36). Tidehunter deals 178 damage to Enchantress using Ravage.
  14. 29:07 (17:36). Juggernaut deals 21 damage to Naga Siren using Bladefury.
  15. 29:07 (17:36). Juggernaut deals 21 damage to Dark Seer using Bladefury.
  16. 29:07 (17:36). Juggernaut deals 21 damage to Tidehunter using Bladefury.
  17. 29:07 (17:37). Lina gets the modifier_rubick_telekinesis.
  18. 29:07 (17:37). Naga Siren gets the modifier_enigma_black_hole_pull.
  19. 29:07 (17:37). Dark Seer gets the modifier_enigma_black_hole_pull.
  20. 29:07 (17:37). Tidehunter gets the modifier_enigma_black_hole_pull.
  21. 29:08 (17:37). Puck gets the modifier_enigma_black_hole_pull.
  22. 29:08 (17:38). Shadow Shaman loses the Ravage Debuff.
  23. 29:08 (17:38). Enchantress loses the Ravage Debuff.
  24. 29:08 (17:38). Enigma deals 45 damage to Naga Siren using Black hole.
  25. 29:08 (17:38). Enigma deals 22 damage to Tidehunter using Black hole.
  26. 29:08 (17:38). Enigma deals 45 damage to Dark Seer using Black hole.
  27. 29:09 (17:38). Enigma deals 45 damage to Puck using Black hole.
  28. 29:09 (17:38). Rubick gets the modifier_rubick_spell_steal.
  29. 29:10 (17:39). Lina loses the modifier_rubick_telekinesis.
  30. 29:10 (17:39). Lina gets the Ravage Debuff.
  31. 29:10 (17:39). Naga Siren gets the Ravage Debuff.
  32. 29:10 (17:39). Dark Seer gets the Ravage Debuff.
  33. 29:10 (17:39). Puck gets the Ravage Debuff.
  34. 29:10 (17:39). Tidehunter gets the Ravage Debuff.

These timestamps dredge up the precise sequence of events after iG “cut for the shorter path . . . [and] storm up the river” and unleash the “wombo combo” that defined the metagame in the summer of 2012 (Valve 2012c). Following these numeric footsteps, the play begins when the Chinese team’s “big wrap-around gank” takes the form of an incognito, five-man blitz that hopes to take Na`Vi from the rear and pinch the Ukrainians between the incoming team and iG’s second tier of towers (see Figure 5.11). Zhou initiates the aquatic ambush with Naga Siren’s ultimate ability: a powerful “Song of the Siren” that locks the enemy team into an invulnerable slumber. While incapacitated, Na`Vi are immobilized but, importantly, physically immune from taking damage. At this point, YYF’s Dark Seer uses “Vacuum” to suck Na`Vi into one centralized position. Everything is ready, but Naga Siren must stop singing before IG can enact their coup de grâce and win the team fight (and, inevitably, the game). Due to the time scale of human reflexes and the delay of team coordination, a window emerges for another probability. Between Song of the Siren and iG’s next important move, the Tidehunter’s 2.77-second, area-of-effect stun and damage ability called “Ravage,” there is a small yet critical window. Between Zhou cancelling the Song and Faith initiating Ravage, Na`Vi have a chance to change the metagame. Crucially, these two moves cannot overlap or Na`Vi will still be invulnerable and the Tidehunter’s ultimate ability will go wasted. Operating with as much temporal efficiency as possible, it took Zhou and Faith 0.46 seconds to stop one spell and cast another. But a half second was all Na`Vi needed to mount their counterattack.

From the moment Puppey gave away Seer and Siren in the draft phase, Na`Vi had engaged in a gamble. Knowing iG would select the popular heroes, Na`Vi also bet they would attempt to use them in a predictable way: first singing, then stunning. While under the effects of Siren’s song, Na`Vi engage in another gamble. Although they have no way of knowing exactly when Zhou’s spell will end, they hope to exploit the brief window between the syncopated, serial execution of Song of the Siren and Ravage. Rather than relying on a gut reaction, Na`Vi address the probability of play by becoming machine. While their avatars are immobilized by the sleep spell, each member frantically mashes buttons. The Ukrainians wager that they can tap faster than Zhou and Faith can coordinate their spells. Even if a tap lands a single frame before Song of the Siren ends, the next rote tap should theoretically occur before Faith can coordinate his single button press. By contrast, Faith cannot engage in the same frenzied button mashing but must deliver a single button press at microtemporal scales of precision, what photographers call the “decisive moment.”

Figure 5.11. Under the cover of the Smoke of Deceit, iG ambushes Na`Vi with Naga Siren’s ultimate ability (top), putting them to sleep while Dark Seer Vacuums them together (bottom).

Na`Vi’s prediction is correct and exactly 0.46 seconds after the Naga Siren ceases her five-second song, as the Tidehunter’s tentacles begin to spread across the screen, they break the metagame.[51] The series of actions that follow have not only brought celebrity status to Na`Vi but serve as one of the most memorable moments to date in the history of e-sports (see Figure 5.12). First, the Grand Magus, Rubick, piloted by Dendi, deploys a force staff to fly outside of the reach of Tidehunter’s Ravage, avoiding the 2.77-second stun. Within the same fourteen-frame window, XBOTC’s Juggernaut activates his invulnerable “Blade Fury” and LightofHeaven’s Enigma, standing in the center of the screen, uses an item called a “Black King Bar” in order to evade the Tidehunter’s ultimate ability as it stretches across the field. The result of these carefully coordinated countermaneuvers is that only two of the five heroes, Puppey’s Enchantress and Ars-Art’s Shadow Shaman, are stunlocked by the Tidehunter’s Ravage. As if the result of Na`Vi’s clever counter picks, in-game preparation, and quick button clicks is not enough, at this moment Faith makes a critical mistake.

Figure 5.12. As the Tidehunter’s ultimate ability, Ravage, begins to cross the field, Na`Vi springs their countermaneuver (top) so that only two of their five heroes are affected by the area-of-effect stun (bottom).

Beyond the metagame of popular picks and bans at The International in 2012, Rubick, the Grand Magus, represents a kind of metagame in and of himself. Rubick’s ultimate ability, “Spell Steal,” allows the player to capture the last ability used by any hero in the game. Thus, like Masahiro Sakurai’s pink puffball, Kirby (discussed in chapter 1), Rubick is a kind of meta-hero who represents the recombinatorial possibility space generated by the almost 400 spells in Dota 2 that summer.[52] The only way to guard against Rubick’s Spell Steal is for an opponent to immediately follow any powerful or game-changing spell with something inconsequential. The tighter the window, the lower the chance a skilled Rubick can steal the coveted spell. Faith understood this. Yet instead of adopting Na`Vi’s frantic and mechanistic mashing to prevent Dendi from stealing the Tidehunter’s Ravage skill, Faith waits just over a full second—an eternity in this time scale—before attempting to block the Spell Steal. Whereas Faith hesitates for 1.33 seconds while repositioning, Dendi does not waste this narrow window of opportunity. After dodging Ravage with his force staff, Dendi turns around, disables Chuan with “Telekinesis,” and then steals Tidehunter’s ultimate ability before Faith can use “Anchor Smash.”[53] At this point the tide finally turns (see Figure 5.13).

Figure 5.13. With iG engulfed by the Enigma’s Black Hole, Dendi is able to steal the Tidehunter’s Ravage (top) and turn it against the Chinese team (bottom), resulting in a teamkill that wins the match.

LightofHeaven’s Enigma initiates his ultimate ability, “Black Hole,” to gather and stun the entire iG team. Next, Dendi turns the purloined Ravage against iG by letting loose his stolen spell on the tightly clustered team. With the Chinese team disabled and damaged, Na`Vi operate in unison to swiftly deliver the kill. The 0.46- and 1.33-second windows of opportunity are the decisive moments that reverse the course of the game. From that point they easily go on to win the match and the winner’s bracket semifinals. Here The Turn of the Tide ceases to correlate to the human experience of seeing the drama unfold at The International or even playing Dota 2, but signals a different turn altogether. Na`Vi does not deliver an act of athletic beauty as much as they execute a statistical exploit. By first picking heroes with abilities that specifically counter the metagame, then spamming their spells in order to rebuff the “wombo combo,” Na`Vi gives Dendi room to cast out and hook the machinic undercurrent to reverse the flow of the game.

As a way of establishing more robust play, anticipating probable outcomes, and tracking the metagame, a cottage industry of statistical aggregation and analysis has flourished around competitive Dota 2. Lacking the institutionalization of large databases available to more established professional sports, websites like have developed their own datasets produced by particularly inventive and industrious players. Representing the first archive of Dota 2 statistics, Dota Academy was designed by two South Americans, Bruno “Statsman” Carlucci from Argentina and Bruno “Shostakovich” Tomaz from Brazil. The duo began collecting data the hard way, Carlucci (2013) explains, first by simply “watching the [video-on-demand (VOD) or] replays and entering the info by hand” before he “decided to finish [his] early parser in order to start using it on [Dota Academy].” By developing software to analyze Valve’s .DEM files and scrape game statistics, Carlucci (2013) realized “that there was lots more info to extract than [he] initially thought” (emphasis original). As Steven Connor explains, “we are seeing the statistical ecstasis of sport” in the sense that statistics operate both within and around the game. In the case of the game statistics surrounding Dota 2, they have taken on a life of their own and become a metagame in their own right. As with Krukar’s one thousand FPS animation, images Dota differently and opens up alternative avenues for spectatorship. In addition to Dota Academy and the Dota 2 Replay Parser, Carlucci produced the first Dota 2 Fantasy League[54] and found himself, after a stroke of luck, extemporaneously commentating The International in 2012.

Affectionately nicknamed “Bruno the Statsman,” Carlucci’s presentation of statistics is as playful as it is productive. Making his first official appearance as a live commenter during The International in 2012, Carlucci donned a diverse array of flamboyant leisure suits (with unbuttoned collared shirts that reveal both chest hair and a silver Dota 2 medallion). Bruno’s performance of nerd-inflected machismo operates not only in contrast to his more homogenous North American and European colleagues on the commentary desk (almost universally dressed in uniform-like hoodies and plaid button ups that year), but also offers another example of how racial difference operates within the culture of Dota 2 apart from the narrative of China versus the world. At The International, Carlucci leveraged his outsider status and nationality to play the part of the court jester and eventually appeared, for example, in an Orientalized fortuneteller outfit and even as Star Wars’ C-3PO in later tournaments (see Figure 5.14). Beyond wearing fluorescent costumes and adopting colorful affectations, the statistics Bruno presents while broadcasting are not particularly useful, or at least not useful in a traditional, positivist sense. Whereas Dota Academy, the Replay Parser, and the Fantasy League represent rigorous investigations into the untapped potential of data analytics in Dota 2, during The International the Statsman worked behind the scenes to produce statistical information while simultaneously undermining the authority of statistics.

Although Carlucci’s performances can be read simply as quirky jokes or a defense mechanism against the inevitability of failed predictions, they also challenge the naive correlation between probabilistic projection and how the tournament actually unfolded. For example, at times Carlucci argued for statistics founded on elementary arithmetic errors: “You pick an Anti-Mage who has a 50% chance of winning. You pick a Lycan who has a 65% chance of winning. That makes . . . a 115% chance of winning. There’s no way to lose!” (Valve 2012d). In another instance, statistics are culled from seemingly irrelevant correlations with racist punchlines: “Tobi, whenever he casts, he’s 5–3 on China team’s winning. Lumi is 5–5. But Ayesee whenever he casts he’s 7–2 for China teams winning. So it’s confirmed. Ayesee is secretly a Chinaman” (Valve 2012e). (As discussed earlier in the chapter, Chinese stereotypes and racism saturated The International.) Finally, Carlucci drew attention to the absurd specificity of certain statistical anomalies as a means of questioning the very possibility of generalizable or comparable actions and events. After a match between Evil Geniuses and TongFu the Statsman jokingly observed that “the problem with EG was a surplus of legs. Look at TongFu’s lineup: Naga, no legs; Venomancer, no legs; Morphling, no legs; Enigma, no legs; Lina has two legs but she’s flying so she doesn’t use it. That’s the strat right there” (Valve 2012f). Just as Rosencrantz and Guildenstern’s coin landing on heads ninety-two times in a row is improbable but not impossible, Carlucci’s deeply ironic and skeptical statistics suggest that anything is possible, everything is equally meaningful, and nothing is exchangeable despite the projected probabilities.

Figure 5.14. A colorful character, Carlucci has performed as a fortuneteller, as Star Wars’ C-3PO droid, in fan-made memes that edit his appearance to make him look like the Riddler from Batman, and also replaced Mark Zuckerberg’s portrait in parodies of the poster from The Social Network.

Carlucci’s statistics parody the supposed generality and predictive potential of quantitative data by suggesting that any given statistical record is absolutely specific to a given historical moment. Like Borges’ cartographic critique of “exactitude in science,” Bruno has conflated the statistical map with the historical territory. This kind of statistician is no longer interested in the probability of signal to noise, but in the significance of all noises. Following a “science of the exception” rather than that of the common denominator, the Statsman offers an example of ‘pataphysical play in which the probability of an event is absolutely contingent with the event itself. Alfred Jarry (1911, 21–22), best known for his shocking and much-reviled play Ubu Roi (1896), defines ‘pataphysics as

the science of the particular, despite the common opinion that the only science is that of the general. Pataphysics will examine the laws governing the exceptions . . . since the laws that are supposed to have been discovered in the traditional universe are also correlations of exceptions, albeit more frequent ones, but in any case accidental data which, reduced to the status of unexceptional exceptions, possess no longer even the virtue of originality.

In the tradition of Jarry’s turn-of-the-century satire, Carlucci’s jokes speak to the impossibility of generality in physics and critique the Enlightenment dream of finally resolving the relation between reason and reality. In this way they caution the viewer against the narrative potential of numbers by a surrealist statsman. Carlucci’s buffoonery constantly footnotes the tenuous relation between math and matter, ratio and kinesis. The agony (and augury) of the statsman is the constant, nagging knowledge of this philosophical tension. If, as Jarry (1911, 21) writes, ‘pataphysics “is the science of that which is superinduced upon metaphysics . . . extending as far beyond metaphysics as the latter extends beyond physics,” then where are the ‘patagames that extend beyond metagames?

Back in IG’s booth, although Zheng “Faith” Hongda knew what probably should have happened in the second game of the winner’s bracket semifinals, he could not have predicted how Dendi’s Rubick would turn the Tidehunter against him and his team. In the end, although Na`Vi’s twenty-second victory has been lionized on forums and archived on video sharing websites, their success in the second and third games of the winner’s bracket was ultimately a pyrrhic one. What probably should have happened in the match actually happened across the greater Dota tournament. iG would go on to dominate the loser’s bracket finals and return to rematch Na`Vi for the grand prize. With very little fanfare and lots of safe, highly calculated plays, iG won The International and a million dollars in 2012. And, in retrospect, Dendi’s Turn of the Tide figured more as statistical anomaly in the face of a long, macro strategy that ultimately outpaced the mechanical, micro movements of the Ukrainian team. Na`Vi won the seventeen-second battle but iG won the million-dollar war. The far less crowd-pleasing but overall more effective method of playing to the numbers rather than to the audience or even to individual games is what would ultimately take the million-dollar prize—and this result suited Valve just fine. Beyond The International tournament, there was another international game the company was winning.

Since the turn of the millennium, the Chinese Ministry of Culture has enacted a ban “forbidding any company or individual to produce and sell electronic game equipment and accessories to China” (Ashcroft 2010). Discussed in relation to Narcissa Wright’s speedrunning practices in chapter 1 and 3, part an act of censorship (to “protect the mental and physical development of the nation’s youth”) and part protectionist economic strategy, the “console ban” deeply affected Chinese gaming in the 2000s, producing a profitable gray market of international goods, a host of alternative and after-market consoles,[55] a culture of both software and hardware piracy, and a boom in PC gaming (Clark 2013). With over a hundred million PC gamers in China—worth a whopping 6.8 billion dollars of the twenty-billion-dollar global market—both domestic and international publishers of MMO games and free-to-play web games were quick to capitalize on the fastest-growing market for PC games over the last decade (McNew 2014, DeCarlo 2013). Furthermore, freely downloadable mods like the original DotA were able to run on almost any PC given the age, popularity, and piracy of Warcraft III. As stated earlier, at its height in 2010 IceFrog (2010) estimated around 10 million Chinese DotA players (IceFrog 2010).

Thus, hiring IceFrog, rebuilding DotA as Dota 2 in 2010, and then inviting five Chinese teams to compete in (and eventually win) The International in 2012 was not simply a way to generate enthusiasm among the Chinese players but was an important part of Valve’s global strategy to expand their digital distribution platform, Steam, into China as a means to tap the country’s PC gaming (and piracy) market. It is already estimated that 75 percent of all videogame purchases take place through Steam, but Valve has taken up the challenge of converting pirated games to purchased games in Russia and China through their “frictionless” marketplace (Edwards 2013).[56] Only a few months after iG won The International, Valve contracted East Asian publishers Perfect World and Nexon to publish and distribute Dota 2 in China and Korea respectively. Though Valve cannot simply release Steam in China due to the country’s stringent economic regulations, Perfect World’s Dota 2 client is driven, nonetheless, by the Steam backend. The undercurrency at the heart of The International is not the ebb and flow of Na`Vi’s mechanical Turn of the Tide, nor is it the explosion of sales on the other side of the world—no, The Turn of the Tide is nothing less than the redefinition of the videogame from a leisure activity (a waste of time) to productivity (labor time). And the undercurrency in which Valve traffics is not fame, not profit, not entertainment, not fun (though these can be shored up), but productivity.

Welcome to Flatland

Before their incorporation in 2003 and well before their release of Dota 2 a decade later, Valve was in the business of managing metagames. In 1996, Gabe Newell and Mike Harrington left their jobs at Microsoft to found a new company based on three undervalued business strategies. In their next venture, the two “Microsoft millionaires” planned to (1) invest in the exponentially decreasing costs of communicating with customers, (2) acquire and retain the most expensive employees instead of outsourcing, and, eventually, (3) capture player productivity through videogames—a genre of productivity software not entirely different from the Windows operating systems that Newell and Harrington worked on at Microsoft (Newell 2013). The duo’s counterintuitive business plan arose after they noticed something astounding: even with an install base of thirty million PCs in the mid-1990s, Windows 3.0 was only the second most used piece of software in America. Towering above the operating system was the shareware release of Doom (1993), the early first-person shooter developed by id software. Although most remember Doom’s brutal 3D graphics and blazingly fast gameplay, Newell (2013) still marvels at the illogical idea that “somehow the largest software company in the world was being out-distributed by a twelve-person company in Mesquite, Texas. . . . Doom [had] a completely different approach to connecting users with value.” If a company like id could stand on even footing with Microsoft, suddenly the world seemed pretty flat.

Following in the footsteps of Michael Abrash, who left Microsoft in 1995 to work on id’s next game, Quake (1996), in 1996 Newell and Harrington quit their jobs and licensed the Quake Engine to begin building what would become the Valve’s first game: Half-Life (1998). As anticipated in chapter 2, Half-Life is not just a first-person shooter operating according to the same gameplay idioms as “Quake and Doom”; it is also a metagame made within Quake and about Doom—or at least Doom’s distribution model. Through its dissemination via a not-quite-frictionless, early version of Steam released first in 2003, Half-Life functioned like a sequel to Doom not only in terms of its genre conventions, but because it attempted to emulate Doom’s ubiquitous dissemination and id’s flattened approach to “connecting users with value.” In the late nineties, Valve sought to become an economic sequel to id and Half-Life would be their Doom, a vector connecting the company directly to their customers and a major platform for generating productive play in the form of metagames.

By actively courting customers cum content producers and mining the productivity of their player base, Valve’s history of products coincides with the acquisition not only of popular mods, but the players responsible for making them. Team Fortress 2 is a class-based multiplayer deathmatch mod originally released in 2007 as part of Valve’s The Orange Box collection, but was only released after the company hired John Cook, Robin Walker, and Ian Caughley, the original team that made the Team Fortress mod for Quake in 1996. Counter-Strike, the globally popular competitive shooter, began as a Half-Life mod by Minh “Gooseman” Le and Jess “Cliffe” Cliffe in 1999 before Valve hired the pair of programmers to institutionalize their work at the company in 2000. Portal’s (2007) revolutionary mechanics started as a student project at DigiPen Institute in Seattle titled Narbacular Drop (2005) before Valve hired the whole team along with writers Erik Wolpaw and Chet Faliszek,[57] who worked to adapt the gameplay to both the company’s proprietary Source Engine and the Half-Life universe. And Dota 2, of course, had an extensive history within Blizzard’s Warcraft III before Valve hired IceFrog and Eul to remake it hero by hero in Source. Very few of Valve’s innovations center around the creation of original, in-house IP; instead, they have developed a business model based on colonizing, expropriating, and assimilating metagames into a framework of benevolent capitalism.[58] The company typifies the post-Fordist corporation designed not around the creation of new products ex nihilo, but the organization of a social factory that extends beyond the labor of employees to the leisure of customers. As Gabe Newell (2013) is fond of saying, “the only company we’ve ever met that kicks our ass is our customers. We’ll go up against Bungie, or Blizzard, or anybody but we won’t try to compete with our own user base, because we already know we’re going to lose.” Why compete with the work of metagamers when you can organize their communities, analyze their practices, absorb their modifications, and profit from it?

For McKenzie Wark, the locus of power no longer rests with those who control the mode of production, but lies instead in the management of the vectors, connections, and networks organizing information—from the financial instruments responsible for high frequency trading to the social media platforms that collect user-generated content. In an information economy, the bourgeois capitalist has been replaced by a new ruling order that Wark (2004, 21) calls the “vectoral class.” After first introducing the term in A Hacker Manifesto, in Telesthesia Wark (2012, 72) asks,

What if the ruling class of our time were not exactly capitalist any more, but more properly vectoralist? What if a fraction of that ruling class acquired its political-economic power through the ownership and control of vectors along which financial information flows, and with it the flows of that information, not to mention stocks of these weird para-things, these instruments of a purely digital private property, with somewhat attenuated relations back to referents in other natures, other worlds.

The vectoral class represents a new species of managers who do not simply control the mode of production, but crucially, mediate connection, dissemination, and access to information. In an era in which production and consumption frequently take place in the same act, in which Marx’s use value and exchange value are no longer mutually dependent antinomies, companies like Valve generate profit not only by making and selling games but through the capture, steering, and control of information flows. Whether appropriating metagames by hiring successful modders, building economic infrastructures to cultivate an international e-sport, or mining Steam stats to predict future sales, Valve’s business model is based on locating those practices that exist outside the market and monetizing relations that previously were regarded as having no value (even though this extra-market activity operates as the precondition for exchange in the first place).[59] As Wark (2013) states “the new stage of commodification is less about extracting surplus value from labor as extracting surplus information from play. It extracts value by offering information for free, but extracting more information in return—surplus information.” The vectoral class does not traffic in fixed capital, but instead profits from continuous and promiscuous exchange within a flattened economy. And nowhere is this flat world better represented than in Valve’s business model and corporate culture.

“Welcome to Flatland.” Valve’s (2012, 37) official Handbook for New Employees opens with an invitation to join a company without higher-ups, corporate ladders, or corner offices. Like Gilles Deleuze and Félix Guattarri’s (1987, 5) declaration in A Thousand Plateaus, Valve decided they were “tired of trees” and “stop[ped] believing in trees, roots, and . . . all of arborescent culture.” Instead, Valve attempts to operate rhizomatically, starting with their office space (see Figure 5.15). In terms of a flat, physical space, the company’s headquarters in Bellevue, Washington is famous for its open floor plan and reconfigurable desks on wheels that represent the effervescent flexibility of affective, informatic, creative knowledge work and the dream of a job that is not work.[60] In the Handbook, Valve (2012) suggests new employees “think of those wheels as a symbolic reminder that you should always be considering where you could move yourself to be more valuable.” The flat office space symbolizes Valve’s flattened corporate hierarchy. Valve has no management, no quality assurance, no marketing department, and no job titles. Newell (2013) argues that much like offices, “titles are actually [Valve’s] enemy rather than something that makes them more productive,” and the Handbook confirms, “we don’t have any management” (Valve 2012h). Following the company motto, “boss-free since 1996,” every employee determines for herself how to be productive and generate value (Valve 2012h).[61]

Valve’s spatial and hierarchical flatness, expressed by their wheel-friendly workplace and advertised in their Handbook, is a tool for attracting, acquiring, generating, and retaining valuable employees. When Newell and Harrington organized the company in the mid-1990s, they not only hoped to exploit the exponentially decreasing friction between producer and consumer (implicitly understanding that their player base was ultimately the more significant substrate of outsourced labor), but began investing, counterintuitively, in the most valuable employees instead of outsourcing cheap labor. In 1996, the two “became convinced . . . that everybody was going in the wrong direction [by outsourcing their employees] for the lowest cost English language speakers in the world” (Newell 2013). So Valve decided to buy “the most expensive talent that was out there . . . [because] those were the people that were least correctly valued” (Newell 2013). Newell (2013) clarifies, “By talent, which is a word I hate, I just mean the ability to be productive.” Valve’s model of value flattens all activity and energy into a single, dereferentialized and abstracted metric: not money, but productivity.

Figure 5.15. Where Figure 1-3 from Valve’s Handbook for New Employees does away with vertical hierarchy, Figures 2-2, 2-4, and 3-1 offer 1950s-inflected illustrations of the company’s flattened management philosophy.

For Valve, productivity is measured in two ways. First, the pure volume of output is quantified. For example, “in IBM in the 1980s, typical productivity would be 1000 debugged, shipped lines of code per year. . . . Whereas when [Valve] was shipping Half-Life 1, one employee, Yahn Bernier, was shipping 4k lines of code per day” (Newell 2013). Beyond lines of code, Newell notes, “Even though it’s easy to see there is huge variation [of productivity] in software programmers, there was probably that same variation in a lot of other roles.” So, second, on an internal level, productivity is peer reviewed to assess employee contributions and award bonuses based on four main categories:

  1. 1. Skill Level/Technical Ability: How difficult and valuable are the kinds of problems you solve?
  2. 2. Productivity/Output: How much shippable (not necessarily shipped to outside customers), valuable, finished work did you get done?
  3. 3. Group Contribution: How much do you contribute to studio process, hiring, integrating people into the team, improving workflow, amplifying your colleagues, or writing tools used by others?
  4. 4. Product Contribution: How much do you contribute at a larger scope than your core skill? (Valve 2012h, 30–32)

Although the Handbook introduces new employees to the company’s culture and philosophy, Valve’s assessment of value ideologically extends beyond the Flatland located in Bellevue, Washington. There are about 330 employees at Valve, but the productivity and value of the players is as significant as the contributions of those on salary. One could imagine a fifth category added to Valve’s “Stack Ranking” to account for play.

Valve’s nonhierarchic, flattened philosophy not only applies to the architecture of their office and corporate hierarchy at Bellevue, but also informs the company’s approach to game design.[62] The salutation “Welcome to Flatland” applies just as much to its community of players as to the employees. Few of Valve’s games originate in-house but they do not outsource. In the case of Dota, they chose to insource a preexistent development team, flattening play into a form of productivity that could then be measured in terms of the work logs of any Valve employee. In an article in PC Gamer, Erik Johnson (Thursten 2013, 40) outlines Valve’s relationship to the Dota players: “We look at every single person in the game as creating user-generated content. . . . A person who just plays the game is generating some value for the other nine people playing.” The line between Valve’s data-driven game design philosophy and data-driven economics is porous. Players are no longer thought of simply as customers for whom the company must deliver, but as fellow producers within a gamespace of global capital who voluntarily generate surplus. By offering three thousand unique applications to over 75 million active users, generating more Internet traffic than most countries as one of the largest bandwidth consumers in the world, and growing about 50 percent a year since 1996, for Valve play is not only a byproduct, but a valuable resource.

Whereas Dota’s producers see play as an undercurrency of flattened productivity, in turn many players see the ability to financialize play is one of the appealing aspects of the Dota 2:

Dota 2 is built around these transactions, beginning at the individual player level, passing through the Workshop—where user-created cosmetic items begin their journey to the in-game store, earning six-figure sums for some—all the way to the competitive scene, dispersed among dozens of tournament showrunners who sell tickets and merchandise through the game client. These surrounding systems are a vital part of Dota 2’s identity—they’re game mechanics too, in a sense. (Thursten 2013, 41)

With the cosmetic markets associated with Team Fortress 2, Counter-Strike: Global Offensive, and Dota 2, Valve has created an economic system that flourishes through the contributions of both local employees in Bellevue and remote players on Steam. In 2013, for example, Stephanie “Anuxi” Everett was able to quit her job and make a living modeling original Dota 2 cosmetic items while collector PAADA sold a rare “Ethereal Flame Pink War Dog Courier” for $38,000 within the same market (Cameron 2013; PAADA 2013). With Valve taking a cut of the tens of thousands of dollars circulating between individual players (not to mention twenty million dollar collective prize pools), it would appear the metagame is quite profitable.

Whether working at Valve or playing Dota 2, “money needs to flow as a signaling tool. . . . In order for people to really assess that what they’re doing is valuable or not you need currency” (Newell 2013). Rather than treat money as the ultimate product, it has become an integral part of a generative, productive process. Newell (2013) tempers any claims of success with a simple question: “But is [Valve] interesting beyond being another company that makes a bunch of money?” Money is not merely an outcome, but has become a core aspect of Valve’s game design philosophy. No longer simply a medium of exchange, money is the equipment with which both the company and its customers play.[63] The mechanics of money, the informatics of money, and the communicability of money is meaningful to Valve. Beyond the flattening of an office space, the flattening of a corporate hierarchy, the flattening of software design, and the flattening of play, Valve is also flattening one videogame to another as a means of building a metaverse in which money can flow to the most valuable parts of the network.

Most games exist independent of one another, as magic software circles or secret Internet enclaves whose forms of production (virtual currency or otherwise) are often imagined as incommensurable with one another. As anyone moving from one MMO to another or switching e-sports knows, exchanging content or migrating production between two games is the equivalent of trying to mail a postcard to Mars—content is returned to sender to waste away within an abandoned virtual world. In the absence of a flattened gamespace, videogames are designed with what Newell (2013) describes as a “whimsical notion of property rights,” because the labor time invested and value generated in a game easily obsolesces and cannot be translated between platforms. Steam is designed to “plumb [the] notion of both ownership and authorship throughout the entire system. So [a player will] be able to create something in one game and exchange it for value with somebody else” (Newell 2013). Rather than attempt to create a “game of games” in the form of a universal game engine—a totalizing metaverse[64] representing the sum of all games (although Steam certainly aspires to this standardization), money has become the general equivalent. Money is the code through which Valve is building its metagame of metagames.

The act of making money is also tied to the generation of alternative forms of currency. Considering the markets surrounding the Team Fortress 2 hats and Dota cosmetics dropped randomly in-game, bundled with other purchases, or available for limited times in their online store, Newell (2013) ponders some of the ethical and practical issues surrounding the production of money and markets in Valve’s games:

We started to see things like inflation. We started to see deflation. We started to see users creating their own versions of currencies, mediums of exchange. Countries started to create regulatory structures so in Korea you actually have to create the equivalent of a W-4 form for your players to account for the virtual income they get in playing your game. . . . Should we increase the drop rate for those customers to offset the implicit sales or income tax that they’re having? And what do we do about purchase price parity? Should we adjust drop rates to provide welfare to people who are playing in lower incomes or do we drop the value of their drops to reflect the fact that they can trade their hats for cheeseburgers?

Once players begin to create value and that value can be translated between different domains, there is nothing to stop the equivalency of in-game object to real-world goods and services. So, what at first seemed to be a game design problem becomes a moral, ethical, and political problem within Valve’s economic Flatland. As Steam continues to extend its functions to grammatize material platforms and capture attention through biometric devices like the Steam Controller, console-like hardware platforms like Steam Machines, Linux-based operating systems like Steam OS, crowdsourcing spaces for developing new games like Steam Greenlight, and even player-driven prediction markets like Steam Curators, Flatland becomes Flatworld—something akin to what McKenzie Wark (2007, 1) labels the “gamespace”—the smooth topology of global capital in which everything is different in the same way and abstract exchange proliferates without the friction of play.

Much like Anuxi, PAATA, Carlucci, or IceFrog himself produce value through the way they play Dota 2, one of the three featured players in Valve’s 2014 documentary Free to Play, Na`Vi’s Dendi is described by Gabe Newell as one of the most high-value Dota players not only for his virtuosity, but for the value he adds to e-sports as an entertaining personality. According to Newell (2013), “being a really good player is a super valuable thing for the community. And the challenge isn’t that you created value, the challenge is coming up with the monetization method.” So rather than selling YouTube ads, hawking Twitch subscriptions, or establishing a PayPal tip jar (though Dendi invests in these forms of monetization as well), Valve’s pennants and tournaments offer a possibility for financialized play within the Flatland. Dendi’s hobby has become a way of life, a metagame that turned into a moneygame once his value had been abstracted and broken down into discrete units and made profitable within the world of Steam.

Similarly, Valve’s employees have claimed that Dota 2 originated as “a post-work diversion” (Thursten 2013, 37). Gabe Newell himself, who confessed to playing Dota around twenty hours a week during its development, has experienced the same kind of flattened productivity throughout his life. Newell has been described as “brilliant and wildly productive” and former colleagues recall him “doing 30 products a year” (Barret, 2005). At the end of a 2013 lecture at the University of Texas, Newell briefly turned to a discussion of his own hobbies, sheepishly admitting that even his play tends to eventually become work. Newell (2013) explains,

I have a million dollar CNC [milling machine]. . . . [At first] I thought, oh good, I’m going to have a hobby, it’s going to have nothing to do with my day job. Well, the reality is all the problems related to machining are software problems. They’re problems of how do you have engagement angles that look a lot like rendering problems. The difference between carving something out of aluminum and drawing something in 3D space are astonishingly similar. And the way that you keep this thing busy is you standardize a lot of stuff and create a network interface so jobs can flow through this thing. And this hobby of mine ends up looking like Counter-Strike in terms of the sets of decisions you would make to improve it.

Recalling his work on Windows at Microsoft and Steam at Valve, Newell perversely finds himself confronting non-standardized interfaces and attempting to maximize the efficiency of industrial processes in his spare time. Newell’s attempt to escape the circuits of productivity ultimately falls flat.

In 1961, Roger Caillois (2001, 32) argued that “industrial civilization has given birth to a special form of ludus, the hobby, a secondary and gratuitous activity, undertaken and pursued for pleasure.” For Caillois (2001, 32), the hobby is “a compensation for the injury to personality caused by bondage to work” and “the hobby of the worker-turned-artisan readily takes the form of constructing complete scale models of the machines in the fabrication” in which the worker ultimately “avenges himself upon reality, but in a positive and creative way.” This homeopathic act of reproduction as revenge against labor within a distinct sphere of leisure (that simultaneously serves to reinforce and reconstitute the system) no longer applies. Under a vectoral regime there is no hobby. There is no fun, there is no work: only productivity for player, laborer, and even millionaire manager alike.

Two Tide Hunters

From wargaming to Warhammer to Warcraft, the tension between ratio (e.g., digital quantity, numeral, abstraction) and kinesis (e.g., analog quality, motion, materiality) operates in all games and is activated and allegorized through the informatic play of real-time strategy games. Steven Connor (2011, 171) reminds us, “Just as probability can neither be distinguished from nor wholly identified with number, so number can neither be extricated from nor entirely exhaust play. Probability is the play of number that number itself makes possible.” If, according to Claude Shannon, information can be defined by the probability of signal to noise, Dota 2’s game design is extremely noisy, information rich, and constantly tending toward states of stability. Instead of entropically evolving more explicit genre conventions or mechanical specializations (like, perhaps, Riot’s League of Legends or the multitude of contemporary Dota-likes and even Dota-like-likes), IceFrog’s approach to game design is glacial in the sense that Dota’s evolution has stalled at the cusp of a phase change—between genres—where the immense possibility space offers a wealth of signals that the player must pick out of the noise. The homeostasis of Dota’s adaptability, rather than any one specific adaptation, is what allows so many metagames to proliferate as different players explore the possibilities of a noisy game.

Despite the human experience of these freely chosen signals—the material, historical, and phenomenal instance of play—videogames like Dota and sports like baseball still cut. Like capital, digital games grammatize quality into digital quantities so particulate (in terms of time, space, and scale) that, on some level, the statistical relation between game and play escapes human experience entirely. Bernard Stiegler calls this form of proletarianization “systemic stupidity” as players no longer know what game they are playing. Similarly, Mark Hansen (2015, 55) argues that “in the face of contemporary data capitalism, time itself becomes an agent of surplus value extraction that operates within a system structurally dedicated to exploiting the imbalance between microtemporal, machinic sensibility and human consciousness.” Both Stiegler and Hansen clarify the philosophical stakes of Newell’s managerial strategies that cut human consciousness out of the loop to better serve the production of quantitative information and accelerate the rate at which labor time is turned into abstract productivity. The metagame, while continuing to function as a human history, no longer seems to matter as human activity itself is flattened by an ever more granular statistical average operating beneath the surface of experience.

In response, different aesthetic practices have been adopted to counteract the explosion of information and the precarity of human subjectivity within digital game spaces. Michael Krukar’s one thousand FPS video applies traditional narrative and hermeneutic forms of knowledge to numeric figures, naively correlating consciousness and cutting. Bruno Carlucci’s commentary, on the other hand, attempts to resist a narrative or conscious conception of numeral as a disruptive, aesthetic form of spectatorship that articulates the disconnect between experience and abstraction. Despite these approaches to spectatorship, Valve generates value from every interaction within their flatworld. Like Dota, Steam is designed around a radical generality and homeostatic adaptability. The scale of each grammatizing function within Valve’s network is so small that rather than conventional, corporate modes of productivity, Valve produces not only Dota-like games, but a Dota-like economy.

Rather than specialize, Valve’s “Dota economics” generalizes and becomes not merely a metonymy for but operates as a fractalized symbiote with the wider circulatory system of Steam. In the same way that steam is invisible and cannot be seen, Valve functions as the valve that controls the flow of productivity and profit, of both currency and the much larger undercurrents that take the form of player labor time, server space, and bandwidth. Within this digital flatworld, play has been cut up and the immaterial labor of thousands of players has been quantified. The sweat of players’ brows within this emergent ecology precipitates into Steam’s atmospheric media. This computational vapor becomes the primary energy source powering a distributed global machine, a world system designed around a ludic mode of production. Within this steam-powered ideology, Valve and value are synonymous.

Hearkening back to medieval moments in history prior to the emergence of industrial capitalism in European culture, when the “v” and the “u” had yet to be divided into distinct linguistic operations, Valve has created a system of neofeudalism in which the distinctions between work and play, leisure and labor, no longer hold water. Indifferent to such Protestant moralizing that compartmentalizes life into different categories, Valve oversees an agnostic economic system in which player-tenants are not farming for gold or money or even for fun, but to be productive despite themselves—a form of data capitalism that captures player attention and productivity on both macro- and microtemporal scales. In a world where the ability to manipulate an avatar can turn you into a millionaire and millionaires find themselves thinking about steel fabrication in terms of first-person shooters, Valve embodies the new value-form.

Within the undercurrency of untapped productivity, managers like Gabe Newell and players like Dendi are tide hunters. Each hunts, speculatively, for a certain kind of undercurrent. The undercurrent cannot be seen, felt, or experienced because each tide hunter is, in their own way, adrift in a sea of data. The undercurrent they seek, though fluid, is not aqueous. Not an undercurrent, but an undercurrency, a deep-seated form of finance capital at the heart of each of their respective games. The undercurrency moves the tide hunter. It rises and falls, yet its influence seems as distant as that of the moon. The waxing and waning of these abstract liquidities is beyond the horizon of human experience as we, humanity, are marooned at sea.

All rights reserved