Bullet Hell: The Globalized Growth of Danmaku Games and the Digital Culture of High Scores and World Records

Bullet Hell: The Globalized Growth of Danmaku Games and the Digital Culture of High Scores and World Records Johnson, M. R. (2016), ‘Bullet Hell: The Globalized Growth of danmaku games and the Digital Culture of High Scores and World Records’, in Pulos, A. & Lee, A. (Eds.), Transnational Contexts of Culture, Gender, Class, and Colonialism of Play: Video Games in East Asia , Palgrave Macmillan, 17-42. Introduction In danmaku ullet hell ga es the pla e is gi e o t ol of a s all a ata – often a spacecraft – a d tasked ith a oidi g a e upo a e of o ple patte s of ullets fi ed from enemy units. These games are most often considered a quintessentially Japanese genre (Crawford, 2013; Davison, 2013a; McMillan, 2013), and an extremely difficult game format which will only appeal to small crowds of players (Donovan, 2010; Clearwater, 2011; Bailey, 2013) who have the requisite skills to meaningfully compete at the highest level. This chapter seeks to challenge these assumptions and puts forward two interrelated arguments: first, that danmaku games have been significantly globalized beyond their native Japan, even if Japanese players remain strongly dominant in high-level competition; and secondly that the culture of danmaku competition has expanded outside the arcades from whence it originated, and now high-level players from many nations – as well as less skilled players who play for reasons other than pure competition – now enjoy and engage with the genre. In doing so the danmaku genre has become a more important aspect of global game culture than ever before. The chapter also argues that out present understandings of high-level gaming competition, primarily eSports and speedrunning, are inadequate to fully explain the model of danmaku competition created by these above two developments. This form of competition, involving as it now does a greater range of players (in terms of both geography and ability) than ever before, merits an analysis which draws upon, but does not rely upon, previous work on these other forms of competitive, professional and high-level gaming. The later parts of the chapter therefore examine danmaku competition in detail. The chapter has the following structure. It is firstly a historical analysis of the danmaku genre, its position within early video game arcade culture, and the shift in danmaku away from competition in the Japanese physical arcade and towards competition mediated in globalized digital forums and video sharing websites. The chapter is subsequently an analysis of the contemporary state of the danmaku genre in regards to world record competition in this genre, the similarities (and differences) between competition in singleplayer games and ultipla e e“po ts, the episte ologi al a iguit i the o ept of the o ld e o d when contrasted with the eSports o ept of the o ld ha pio , a d uestio s of collaboration and sportsmanship posed by this particular form of competition. The chapter lastly returns briefly to the question of geography and considers the continuing dominance of Japanese players in danmaku games, and the role of nationality as a determining factor behind the rise of the most skilled players in a range of game genres. This analysis will demonstrate that the appeal of danmaku games is considerable in both geographical scope and player ability, and that high-score competition is a model of competitive gaming which is noteworthy and highly distinctive from others which have been previously analyzed in game studies literature. To examine these three factors the chapter draws upon scholarly examinations of Japanese gaming culture, original research into the online community of danmaku pla e s, the autho ’s o e pe ie es as a danmaku world record holder and high- level competitor, and work on eSports and other forms of competitive gaming. This chapter therefore provides an overview which aims to both serve as a centralized descriptive source of information for future game studies scholars investigating the shmup/danmaku genres and the nature and history of arcade gaming more generally, and also as an original analysis of a particular form of extremely high-level video game play and the distinctive cultural, competitive and cooperative norms and expectations which have emerged around it. Amusement Arcades, Shmups, and Danmaku Games From the late 1970s until the middle of the 1980s – a pe iod o efe ed to as the golde e a of a ade ga es – amusement arcades were one of the most well-established and popular ways to experience and play new video games (Davison, 2013a). Arcade games were coin-ope ated oi -op , which meant that a player gaming in an arcade had to pay each time they died or lost a continue. As players must insert coins to keep playing, these games were designed to offer immediate rewards within the first few minutes of play. This was often in the form of early, easy bosses, encouraging players to commit to the game throug h an early feeling of achievement and mastery. Many arcade games were also structured to take advantage of the coin-op system in other ways; a five-stage game, for example, might make the first four stages comparatively simple or short whilst the fifth sta ge is extremely lo g a d e t e el halle gi g, to i du e i the pla e a false se se of ha i g ea l o the game and thereby encourage further depositing of coins. A striking example of this mentality is Metal Slug 3 (2000), a game where the fifth stage is longer than all the previous four stages combined, and is also arguably the most challenging stage of the game, and also o tai s a u e of false e di gs a d osses hi h appea to e the fi al oss , u til they are defeated, and the player learns there is yet more to do. Arcade games also featured another important component: a high score list (Taylor, 2012, 3), where players could generally enter a three- ha a te i k a e su h as the autho ’s p efe ed MRJ to represent themselves (this system yields 17576 possible nicknames, a volume of potential nicknames likely sufficient for any arcade). The crucial role played by the arcade high-score list ill e etu ed to late , a d fo s a ke o po e t of this hapte ’s a al sis. One of the most popula ge e of a ade ga es as the shoot-em-up , o sh up . Shmups are games where the player is in control of a single unit, normally a spacecraft but sometimes a tank, aircraft or submarine, and is faced with waves of enemies moving towards them and shooting at the player (Bailey, 2013). The player shoots their own bullets in return; in some games the player can change the direction of their bullets, whilst in other games the bullets are only fired forward. The genre is therefore defined by three major actions from the player – ai i g, shooti g, a d dodgi g Ga eO e, . The fi st of these games was arguably Space Invaders (Betts, 2005; Davison, 2013a), which saw waves of enemies descend from the top of the screen and occasionally fire bullets. Space Invaders and its derivatives soon gave way to horizontal-scrolling shmups such as Defender (1980), and vertical-scrolling shmups such as Xevious (1982), in both cases games where a fixed a kg ou d s olls u de eath the pla e ’s ship as the pla e o es up o a oss the terrain. The overwhelming majority of shmups follow this model, and those with horizontal scrolling are generally considered slightly more challenging since it is far harder for a human eye to track the path of a bullet travelling sideways (GameOne, 2009) than from the top of the screen to the bottom. In the early 1990s, the sub-genre of danmaku (which roughly t a slates as ullet u tai e e ged a d ui kl e a e a staple in Japanese arcades. Although there is slight ambiguity in all sub-genre definitions (Clearwater, 2011), danmaku games are generally agreed to be shmups defined by the overwhelming number of bullets fired at the player (Ng, 2005; GameOne, 2009; Crawford, 2013), the often highly complex geometric patterns these bullets etch out (McMillan, 2010, 273; Bailey, 2013), and a requirement for extremely fast reaction times from players (Clearwater, 2011). These appear to e i po ta t e ui e e ts fo defi i g the diffe e e et ee a sh up a d a da aku . A crucial component of danmaku games is that the pla e ’s hit o – the part of the pla e ’s spa e aft hi h a a tuall take da age – is extremely small (Ng, 2005; GameOne, 2009; Davison, 2013a), meaning that a far greater volume of bullets can be visible on screen at once whilst still making it possible for a player to avoid them. In some games the location of the hitbox is explicit and obvious – many DoDonPachi games contain a flashing circle at whose core lies the few pixels where the player can be hit, as do many others – whilst in some games it is less clear, and the player needs to learn which part of the ship sprite is the part which will actually take damage, and which parts bullets will go th ough ithout issue. The basic mechanics of all shmups and their danmaku subgenre are therefore extremely simple and easily picked up in an arcade environment – move, and shoot (Hock-koon, 2012, 5) – and offer an immediate ease of comprehension for any player (Crawford, 2013; cf. Taylor, 2012, 29). However, once the genre had become established – arguably (Davison, 2013a) via the game Batsugun (1993), whose first boss introduced a hail of bullets on a scale never before witnessed by players – danmaku games quickly became more and more challenging as game desig e s o peted to see ho a p oje tiles the ould get o the s ee at o e ti e C a fo d, . Fe e a d fe e pla e s e e su se ue tl a le to complete these games, although the most skilled players relished these new challenges (GameOne, 2009; Donovan, 2010, 97). Danmaku a e the efo e stagge i gl diffi ult Baile , ga es hi h test [a pla e ’s] ga i g a ilities to the a solute li it Da iso ,  a d appeal to ha d o e ga e s – those who enjoy deep and complex gameplay (Novak, 2005) and are highly skilled in the act of playing computer games (Dena, 2008) – whilst still always maintaining the chance of eventual player victory: as one interviewee in the F e h TV ha el Ga eO e’s do u e ta o sh ups put it – he a danmaku is well done, there is always a way to survive (GameOne, 2009, emphasis mine). Danmaku games are therefore extremely difficult to master – the e is a ua tu leap between completing o e a d a hie i g a de e t high s o e Ho k-koon, 2012, 5) – especially when coupled with the potentially intricate scoring systems which all danmaku games possess (McMillan, 2013), tracked on the high score list for each instance of the game. These scoring systems can be divided into three major categories (Betts, 2005), although some games use more than one of these in conjunction. The first is chaining – this is a score system where the player builds up a hai dest o i g e tai e e ies i a certain order, and the more enemies the player kills according to whatever rules the game possesses for chaining, the higher their score. Examples might be destroying enemies of the same colour – e.g. Ikaruga (2001) – or destroying enemies within a few seconds of each other, making sure to be quick enough with each subsequent kill that the chain is not d opped – e.g. Ketsui (2003). This system promotes a high level of memorization and strategizing to identify the best possible way to chain the finite number of enemies which will spawn. The second is collection – this is a score system where enemies drop items or toke s hi h the pla e a olle t, a d the pla e ’s s o e is i so e a ased o the collection of these tokens (how many are collected, or which tokens are collected, or what sequence tokens are collected in, or different tokens affect different score multipliers) – examples would be Mushihimesama Futari (2006) and Triggerheart Exelica (2006). The third is proximity – this scoring method can take two possible forms. The player either gains more poi ts the lose to a e e the pla e ’s ship is he the e e is dest o ed, o the pla e gai s poi ts fo g azi g e e ullets, hi h is to sa aki g su e that ullets u extremely close to the pla e ithout killi g the , athe tha hoosi g safe paths hi h take the pla e ’s ship fa a a f o e e fi e. Proximity is therefore a risk/reward score mechanic in both its forms, where the greater risk of proximity to enemies or their bullets yields greater rewards. Examples are Ketsui (2003) and Blue Wish Resurrection (2006). As danmaku games became more popular in Japanese arcades, it quickly became apparent that players were willing to practice to master both these complex scoring systems and the challenging game mechanics in pursuit of a coveted high score (McMillan, 2010, 183), and the recognition and mastery that came with it. From Japan to the West Although these impressive high scores were once limited to Japanese arcades, players from all over the world now hold records in the danmaku genre. To understand how these games found their way into the West and into the current global competition we now see, we must first examine arcade gaming in the era before the emergence of the danmaku subgenre. Arcade use within Japan has always been a primary medium for playing games (Aoyama & Izushi, 2003; Colwell & Kato, 2005) and Japan remains the only country which has maintained an active, vibrant and profitable arcade culture into the present day (McMillan, 2010, 368). Arcades were ubiquitous in Japan in the era of Space Invaders, and Space Invaders could in turn be found in a range of other non-arcade settings where brief amusements or distractions might be appropriate (GameOne, 2009) – there were even spe ialized esta lish e ts, I ade Houses , he e o l Space Invaders could be played (Crawford, 2013). Following the success of Space Invaders, shmups and their danmaku descendents quickly became extremely popular in Japanese arcades (Crawford, 2013; Davison, 2013a). In past decades there were even shmup and danmaku tournaments in Japan – these might contain elimination heats, and then a final competition between atte dees, ofte i the fo of a Ti e Atta k ode: a contest to see who can accrue the most points in a two or five minute period (GameOne, 2009). Although other games with extremely high skill ceilings were introduced to the Japanese arcades as time went by – such as fighting and rhythm games (Crawford, 2013) – shmups still retained an important place in the arcade repertoire. When shmups and danmaku games eventually began to expand outside this Japanese context, this change was characterized by two major trends: firstly, the shift of these games away from arcade cabinets and towards home consoles and personal computers when shmup developers began to release their work on these platforms, and secondly, a concurrent shift of high-score competition away from arcade high-score lists and towards a globally mediated community of skilled players. It is difficult to place an exact date on the shift away from arcade machine – Japanese danmaku arcade games trickled into the West beforehand, and there were always dedicated collectors who would be willing to purchase Japanese-region machines in order to play these games outside Japan, although Western arcades were generally devoid of danmaku games (McMillan, 2010, 368) – but the game which exemplified this move was arguably Ikaruga (2001), released worldwide on the GameCube. Ikaruga introduced many Westerners to the genre for the first time (Davison, 2013a;2013b) and has been re-released many times on multiple console platforms, and demonstrated the viability of the shmup/danmaku market outside Japan. Further exploiting this new and perhaps unexpected market, a range of danmaku games produced by Japanese games company Cave – Ketsui (2003), Espgaluda II (2005), Mushihimesama Futari (2006), Deathsmiles (2007) and DoDonPachi DaiFukkatsu (2008), to name but a few – can now be found on the Xbox 360, whilst a number of Western-made danmaku games aiming to mirror the work of the Japanese masters – Jamestown (2011), Shogun: Rise of the Renegade (2012) and Danmaku Unlimited 2 (2014) – have been released on online game platform Steam, and sometimes also on mobile phones (Davison, 2013a), alongside a small number of Japanese independent danmaku games like Crimzon Clover (2011) which are sold online rather than distributed to arcades. As danmaku games have expanded into the rest of the world, the competition for high scores has undergone a coterminous shift in structure and player expectations – the competition has shifted from the physical into the digital. No longer is it the norm to strive for the highest score in your local arcade – and possibly, in some instances, to discuss with players from other popular arcades or visit other arcades to see which machines held the highest scores – but instead to strive for the highest score in the world (although highest arcade and highest global scores were often the same thing in the past, and in some cases are still so today). This shift means the methods by which records are viewed and disseminated have also changed. Danmaku records and particularly noteworthy playthroughs in Japan have always been recorded and distributed via commercial DVDs (Crawford, 2013) – and before that, VHS tapes (Clearwater, 2011) and photographs (Taylor, 2012, 6) – showing expert players handling particularly challenging levels of some of the most well-played shmups and discussing the strategies they use (Betts, 2005). Such videos ep ese t a staged pe fo a e of o petiti e i al , i ol i g skilled pla e s a d thei a ati es “eo & Ju g, , . The p opagation of these media speaks to what Reeves et al call the is e al pleasu e Ree es et al, 2008, 210) of the construction of a sequence of highly skilled danmaku pla , de o st ati g it to o e’s o-players, and in turn being one of those co-players consuming and appreciating the skill of another. In the last decade, however, an expectation of submitting scores to fan-operated websites has become an increasingly important part of the shmup/danmaku communities, especially for anyone competing in the genre at a high level, and these discussion forums and video services have begun to take over some of the roles of older arcade high score lists (now that danmaku is no longer unique to Japan). Despite this growth of danmaku games beyond Japan and the increasing importance of online recognition, there is no one centralized online repository for danmaku records and videos. The most well-known forum and score repository is the System 11 forum, which contains a number of subforums – Shmups Chat, Hi Scores, Strategy, Reviews, Development, Hardware, and Shmupmeets (for real-world meetings of shmup players) – and a highly active community1. The running of these websites, and the management of records and videos, is carried out entirely by grass-roots enthusiasts, as it is with many other competitive games (Rambusch et al, 2007, 161). These environments are digital archives where players upload their highest scores and discuss strategies for maximizing scores, both (as we shall explore in more depth later in this chapter) cooperating and competing with their fellow players to push the records ever higher. They also serve as a centralized repository of video links. Videos are just as important as scores; whereas in the past videos were distributed in physical copy as above, now many videos are uploaded to video sharing sites (although DVD circulation remains in Japan). Gamers of all genres and subcultures often post their exploits 1Other sites include the mostly-inactive Restart Syndrome, the French shmup.com, the site Shmuplations which contains a range of shmup developer interviews as well as scores, and the Chinese Niyaozao. on YouTube and other video sharing sites (Witkowski, 2013), a trend which has now been carried over into the danmaku community. The danmaku community has therefore geographically expanded, become partly virtualized, and overseen a greater distribution of world records, although there remains no universal repository for the storage of, or agreed- upon method for the submission and management of, records and scores. Having explored the origins and growth of this community and the circulation of impressive playthroughs and world records, the next part of the chapter will now delve into more detail about the nature of these world records, and how we might best analytically conceptualize these danmaku world records in relation to our contemporary understandings of professional and high-level gaming. High Scores, eSports and Speedrunning There are two currently two dominant forms of high-level video game competition which have been given attention by scholars. A very significant body of work has emerged around p ofessio al ga i g , o oe o o l k o as e“po ts e.g. Witko ski, ; Rambusch, 2011; Taylor, 2012; Ferrari, 2013), which has deepened our understanding of competitio i ultipla e ga es. This is o ple e ted ith o k e a i i g the o u it of speed u i g F anklin, 2009; Menotti, 2014; Scully-Blaker, 2014) has explored high-le el o petitio i si glepla e ga es. This se tio ill e a i e i detail the nature of danmaku records and propose an analytic framework for understanding this form of competition. This will be achieved via first exploring the contrasts and comparisons we can draw with both eSports and speedrunning, before drawing on both for the theoretical grounding needed to understand danmaku competition, and examining the resulting epistemological u e tai t of the o ept of the o ld e o d . High Scores as eSports? ESports is a term used to describe organized and sponsored competition in multiplayer games – first-person shooters (FPS), real-time strategy games (RTS), multiplayer online battle arenas (MOBA), fighting games – where players directly compete against one another in zero-sum games, and the fi al esult al a s e eal[s] lea i e s a d lose s Witko ski, 2013, 161). Competition is directly against other players in the same game instance, not specific stages or levels in the game, and as such, competition in eSports is structured in the form of tournaments where competitors (winners and losers from previous maches in the tournament) are pitched against one another. Such a form of competition means there are multiple winners and losers throughout the tournament in any given match (qualifiers, quarter-finals, semi-fi als, a d fi als , ut o l o e e e tual i e e e ges to take the title of the ha pio . In this regard we see a potential similarity to danmaku competition – in eSports only one pla e o tea i s the tou a e t, e e if i di idual at hes along the way may be won by others. In danmaku games we could readily compare this to defeating individual bosses or levels hi h a pla e s a a hie e, o eati g o e’s pe so al est s o es, hilst still k o i g that the e is ulti atel o l o e o e all i e at a gi e ti e: the player who holds the world record. However, the problem with applying this to danmaku games is the concept of the champion. The champion (either an individual or team) is essential to eSports – Da id “i li ’s o petiti e ga e ’s i le “i li , is su titled Becoming the Champion, whilst a range of other work (Rambusch, 2011; Seo & Jung, 2014, 2) either implicitly or explicitly assumes a champion-based epistemology for defining the most skilled players or teams who emerge at the apex of a tournament structure. Similarly, a tou a e ts o pete to defi e the sel es as the o ld championship of computer ga es Ra us h et al, 2007, 162, my emphasis). The concept of the champion therefore appears integrally tied to tournaments, which – excluding extremely rare historical danmaku tournaments which only crowned champions within the context of that tournament, not in the context of the game as a whole – does not apply to danmaku games. The model of eSports therefore has some value to our understanding of danmaku records because it allows for both interim and lesser victory and a single eventual victor, but a champion exists within a specific tournament in a specific time and place, not the gradual competition over many years (and in many different homes and arcades) indicative of the pursuit of high scores. This partly helps us conceptualize the world record, but is clearly incomplete and not entirely transferrable from eSports to danmaku. However, there is something else valuable we can take from eSports: the importance of physical competence and skill, and the attendant comparison with ph si al spo ts Wag e , ; “eo & Ju g, , . Although s hola s i spo ts studies te d to poli e the defi itio of the te spo t Fe a i, , a d the de ate o e the o pa iso between competitive gaming and physical sports is beyond the scope of this chapter (cf. Wagner, 2006; Witkowski, 2009; Taylor 2012), it is worth repeating an oft-noted observation. By a standard definition of sport – for example, that by Boxill (2003, 2-3), who suggests that sport must be a freely chosen activity, governed by rules, be physically challenging, and involve competition – all of these apply ready to competitive gaming. Focusing on the third of those requirements, the physical challenge, shows us the ready translatability of this definition of sport into competitive gaming (Taylor, 2012, 37): a player must be able to carry out la e upo la e of ph si all de a di g a tio i o de to e o petiti e Witko ski, 2012, 369) in any challenging game, danmaku included. Performing these actions well and reliably requires what Witkowski (2009) calls training and Ferrari (2013, 5) calls drilling, hi h i da aku ga es p i a il ea s p a ti i g a pla e ’s patte e og itio a d memorization abilities (McMillan, 2010, 275). At the same time, like many competitive games, danmaku ga es do ot o l e a d this ki d of stati spatial k o ledge Ree es et al, 2008, 207), but rather the ability to immediately respond and react to the bullet patterns the player is faced with (whilst many danmaku patterns are fixed, many are also quasi- random). Navigating expected patterns, and reacting to unexpected ones, are therefore both embodied physical abilities (Taylor, 2012, 59) very distinct from scoring optimization; these types of actions are emphasized in the eSports literature and have obvious applicability to the acquisition of danmaku world records as well. High Scores and Speedrunning? We can see that eSports literature acknowledges the physical dimension of gaming skill – essential to danmaku games – but it is clear that danmaku competition is nevertheless distinct from eSports, due to the lack of organizing bodies and tournaments, the indirect nature of this competition, and the problematic concept of the champion which does not seem to fully translate into danmaku competition. Speedrunning – another competitive community based around singleplayer games and (like danmaku the o ept of the o ld e o d athe tha the o ld ha pio – may be more illustrative. Speedrunning is the act of completing a game as rapidly as possible (Franklin, 2009; Menotti, 2014), and is the efo e a e t aoffi ial fo of o petitio Me otti, , e e pla ed fo i game development. The speedrunning community uses the concept of the world record (Scully-Blaker, 2014) rather than that of the champion as in eSports. The shortest proven completion time for a given game in a given category is understood as the record, just as the highest danmaku score in a given game is categorized as the record; similarly, speedrunning does not involve a direct competition against opponents in the same physical room or the same virtual game lobby as the player, but rather records may be pursued at any time and place (as with danmaku). However, there is a major difference between the community of speedrunners and the community of danmaku players: in most speedrunning competitions – except those which pursue what Scully-Blake te s fi esse u s – glitches, bugs and in-game exploits a e all allo ed i the pu suit of the sho test ti e F a kli , ; Me otti, , . The Legend of Zelda: The Ocarina of Time, for example, can be completed in under twenty minutes by using a significant glitch which warps the player from the cha e of the ga e’s fi st oss, to outside the ha e of the ga e’s last oss, i sta tl . Usi g this glit h a d all other glitches) is considered entirely acceptable – each time a new glitch is uncovered, therefore, it becomes possible to beat the previous world record not through superior physical ability nor strategic planning, but by using a new glitch; this means that the goalposts of what is considered a record, and what in-game possibilities exist for a player pursuing a record, change every time a new glitch is uncovered. By contrast, danmaku ga es’ goalposts e ai statio a ; ugs, e ploits a d glit hes a e ot allo ed the pla e community, meaning that from the day the game is released until the game where the last person interested in competing for the record gives up, the goalposts are one and the same: the highest s o e, usi g the ga e’s esta lished ules a d s o i g s ste . We a see this reflected in some speedrunning games where there are multiple world records, using diffe e t ulesets based on which glitches are allowed and which are not (Scully-Blaker, 2014); by contrast, any danmaku game with different record categories will have those categories hard-coded into the game (different difficulty levels, different stages, etc) rather than created by the fan community (cf. Taylor, 2012, 54) as a response to growing knowledge a out the ga e’s s ste s i.e. diffe e t glit hes a e u /a epta le . Co side i g speedrunning allows us to see that these, like danmaku, are competitions in singleplayer games with no need for physical proximity or direct competition between players, but the nature of a record shifts constantly as new glitches are found – and danmaku is a offi ial competition written into the game, not a community-created form of competition – meaning that (like eSports) speedrunning is also not a perfect framework for understanding high score world records. The Epistemology of the World Record High score competition, therefore, appears to be something distinct. We can see there is similarity to eSports in danmaku competition in the extremely high levels of physical skill a d the a k o ledge e t of i te i i to ies hi h uild to a e e tual i to – but danmaku competition lacks the official governing bodies and does not, for the most part, use a o ept of a o ld ha pio . Alte ati el , if e o side speed u i g, e a see a similarity in the pursuit of high-level competition in a singleplayer game, but glitches are allowed in speedrunning whilst they are never valid in danmaku competition, and whilst speedrunning is an extraofficial form of competition, high-score lists have been a part of danmaku games since Space Invaders. To fully understand the nature of a world record, we should return to the discussion earlier in the chapter of score forums, DVDs and arcade high scores: these are all submitted, and created, by players (cf. Taylor, 2012, 228-9), not a body associated with the game which manages scores and records, nor a system programmed into danmaku games which automatically updates a score list on a server (a very small number of modern PC danmaku games do possess this functionality, although such systems are not always reliable and cheat-proof). The lack of automatic recognition of victory – of the sort one would get from playing an eSports game in a public league or tournament system – and the requirement for p oof leads to a le el of episte ologi al a iguit i the defi itio of the o ld e o d (for both danmaku and speedrunning). To take an example: the game Warning Forever (2001), a very unusual danmaku game which eschews the traditional level-boss structure for a game composed entirely of bosses designed by the ga e’s AI i espo se to the pla e ’s actions, supposedly has a world record of around ninety-six million points. Although a number of people claim to have seen this video, the only previously extant copy (from back when forum members discussed the record) is no longer available from its original source, and the author has been unable to identify another copy of this record on any video sharing site, or to get in touch with the player who supposedly achieved this record. The best that can be identified is seventy million (still an extremely impressive score), currently available for all to watch on Youtube. Therefore: is ninety-six million still the world record? Naturally a e od i a o petiti e e dea ou a go a k a ds if it e o es lea that a ew e o d as ea ed disho estl , i hi h ase the p e ious e o d o e o e e o es the e o d , ut hat if all e ide e of a o ld e o d is lost? To consider this in terms of physical sports, we might equally ask: if someone beat the 100m sprint record, but then all visual records of that achievement were lost, would we still consider that to be the world record, when we only have the statements of those who saw it to go on? This is difficult to answer, and naturally the analogy is not perfect (physical sports have larger numbers of people interested in them and therefore we could reasonably assume a larger number of people would have witnessed the achievement), but it is nevertheless an illustrative comparison about this unusual question. The establishment of any kind of record in any competitive endeavour is according to community norms: communities of competition i pli itl ask the uestio hat is suffi ie t p oof of a e o d? hilst e“po ts o u ities ask hat ust a pla e do to e o ed ha pio ? , a d o e ti e a answer to that question emerges, which in the danmaku community means a video (for less- played games a screenshot will sometimes suffice, but for games where the competition is fierce, a full video is essential proof). Without a video – is that still the record? Simply put, players differ: once again highlighting the lack of any central or official coordination for these issues and the distributed nature of danmaku play. Some danmaku players would say that the previous record remains the record, but many would not, and thus acknowledging the inherent uncertainty and ambiguity in a world record appears to be an integral part of such a conception. World records in danmaku games are the site of fierce competition, but that competition is community-managed and player-led – a bottom-up approach, rather than a top-down one – and this leads to the possibility of uncertainty, debate, and competing claims to a single title. A world champion is defined, and whoever meets that definition is crowned; a world record is something ever-changing, and is only defined in relative terms to previous scores. We can therefore see that the concept of a danmaku world record now exists in a discursive space defi ed hat ideos a d p oofs have been submitted (and how believable/reliable these are deemed to be by an unofficial jury of peers) rather than relying on the cultural milieu of the arcade, whilst a world champion appears to require the existence of a structured competition designed to define who possesses such a status. In turn, although the score-keeping system in danmaku games is structured by game designers, the competition is not, being grass -roots. We can therefore appreciate a spectrum: eSports games most often have both the method of keeping score and the structure of the competition coded into the game, danmaku games have a hard- coded method for keeping score but the competition remains grassroots, whilst speedruns look to the grassroots community for both. Newly global Danmaku games therefore exist somewhere in the space between the two, resulting in a competitive situation which is highly distinctive. Cooperation, Competition, and Sportsmanship As we have seen in the above section, danmaku game world record competition exists in the space between our understandings of eSports and speedrunning, although it appears significantly closer to the latter than the former. Having explored this nature of world records, danmaku competition and the frameworks we might use to understand them – acknowledging the inevitable vagueness that competitor-reported world records bring with them – this chapter will now explore some of the intriguing impacts and aspects of danmaku competition, which cannot be fully articulated if relying solely upon work on eSports nor speedrunning. These impacts fall into two categories: the interplay between cooperation and competition, and the lack of sportsmanship in danmaku competition. Examining these two factors will strengthen our understanding of this competition and the community of players involved in it, and further the body of game studies work into the highest levels of competitive gaming. Let us begin with the first of these – the interplay between cooperation and competition – in order to further understand the experience of being a danmaku competitor and the cultural norms one is expected to adhere to. Earlier in this chapter we discussed the sharing of strategies via the videos of high-level danmaku play released to the public, and it is to this point we now return. In eSports, competing teams or individuals would never exchange information about the game, and in some cases top players have bee k o to hold a k new tactics or moves in preparation for an important tournament, to then unleash them on unsuspecting opponents (Taylor, 2012, 95). In speedrunning, however, it is expected that all players will work together on deducing the best strategy for the game (given current knowledge of glitches and exploits). Players cooperate in their theorycrafting – the analysis of game mechanics designed to optimize and improve player decisions and strategies (Wenz, 2013) – and then compete to be the player with the best possible reflexes and execution ability (cf. Witkowski, 2009) who can carry out these strategies in the game itself. Now that danmaku ga es a e the do i a t su ge e of sh up, the skill level required to play [these games] has become dau ti g C a fo d, a d thus olla o ati g o the st ategi le el reduces the pressure and the requirement on any given player to master strategy, as well as mastering execution. All players therefore benefit from this shared knowledge (cf. Boxill, 2003, 108). There is an interesting contrast with physical sports here. Boxill (Ibid, 5) notes that each pla e i a ph si al spo t ust de elop st ategies to ou te a o petito ’s skills a d st ategies ; o t ast, e ha e see that fo danmaku games (and speedruns) the strategizing aspect is a generally collaborative endeavour, even between opposing players who will later compete for a world record. One must only develop skills to outmatch the skills of all opponents (McMillan, 2010, 370-371), ot st ategies to ou te a oppo e t’s strategies (since all high-level players are assumed to be attempting to execute, in essence, the same optimal strategies). And if one player does create an entirely new strategy and intentionally does not share it with the world, as soon as they achieve a world record with that strategy and upload the video, the strategy becomes public knowledge, and the competition thereby once more becomes about the skill of execution rather than the quality of the (now public) strateg . This ea s a pla e de uti g a e st ateg ill a t to get the best score possible with that strategy – si e the a t of eleasi g the st ateg ele ates the ability of all other players – and from this point, it is extremely interesting to examine the ole of spo ts a ship i danmaku competition. Sportsmanship is a contested term, although it is generally agreed to mean some combination of fairness, ethics and generosity in competition (Keating, 1964; Boxill, 2002; de Koven, 2013; etc). One proposal of what sportsmanship should entail is a particular scenario k o as the a ti- lo out AB thesis. Dixon (2003, 84) describes the AB thesis thus: if a pa ti ula ga e appea s to e tu i g i to a lo out , ea i g a ga e he e o e tea is obviously going to achieve an immense and potentially humiliating margin of victory over the other team, the clearly-dominant team or player should pull back and stop trying their est to a i ize the a gi of i to i a o e-sided o test . Feezell , a gues i support of the AB Thesis, proposing that continuing to rack up points after victory is obvious fails to espe t t ue o eal o petitio , e ause su h o petitio e ui es ei g halle ged o tested o th oppo e ts . It is this defe e of the AB thesis hi h is interesting to explore: in a danmaku context it could reasonably be take to mean that one should not achieve an unnecessarily high score, because that implies one was not battling worthy opponents, but rather that one should only achieve a score necessary to secure victory. To see why this is inappropriate and how danmaku is therefore distinct and noteworthy, consider two competitive matches of physical sports, where two teams or two individuals are directly pitted against one another. A score of 2-1 in one match has no relevance to a score of 10-1 in another; one cannot say with certainly that the team that scored ten points in the second match would have defeated the team that scored only two points the first match, for those matches are both entirely different contexts, and one cannot assume the two losing teams are equally skilled (or unskilled) despite each achieving but a single point. However, in danmaku high score competitions, one can make judgements of this sort: each score stands on its own and therefore every single score can be meaningfully compared to every other score, for they all fight against the same in-game foes, levels, scoring systems and bosses. We can therefore see that the AB thesis is in essence only relevant to what those of us in game studies would consider player-vs-player (PvP) competition, such as that found in eSports – both or all players are competing within a given match, and there will always be a winner and a loser from that immediate contest. It does not apply to danmaku (or speedrunning). We can therefore identify the competition for high scores as being what McMillan (2010) alls i di e t P P , ut hi h I elie e is ette e e plified hat I shall te asymptotic competition. Players are not competing against each other (Crawford, 2013) to achieve the highest score in a given day or week, for example, and nor are they competing within the same instance of the game at the same time and trying to defeat bosses faster than each other, or clear the entire game before their opponent. A skilled player therefore does want to maximize the margin of victory, for the higher that margin is, the trickier it will be for any subsequent player to come along and trump it. Equally, almost all danmaku games are finite (Bailey, 2013) and will therefore ha e a a i u possi le s o e – no world records are pe fe t i that ega d, ut if e alte ati el o side the o petitio fo o ld e o ds as a o petitio to get lose a d lose to the pe fe t ga e , he e e e useful pi kup is collected, and every enemy defeated, in a perfect order which maximizes the scoring mechanics of a given game, we could see this as the pursuit of excellence (cf. Witkowski, 2013), rather than the pursuit of victory. It is the pursuit of an ideal perfect playthrough of Ikaruga (or whatever other game), and each time a new record is achieved, human achievement moves closer and closer to that ideal. There is always a winner at a gi e time, but in many cases this is a temporary winner until a superior playthrough is achi eved. I therefore select the term asymptotic to describe this kind of competition: just as in mathematics an asymptotic curve approaches a line but never reaches it, in danmaku high score competition scores approach the perfect game but never achieves true perfection (such a playthrough would require constant pixel-perfect and frame-perfect movements and inputs for upwards of half an hour, which is immeasurably beyond the skill of any human game player). Such competition is better understood as being more akin to individual sports – running, jumping, swimming – than team sports, and is therefore significantly closer to the o ept of speed u i g tha the o ept of e“po ts; all o petito s oope ate i o de to identify and comprehend the scoring systems in the danmaku game in question and to exchange strategies, and the competition rests on who has the technical skills to execute the riskiest but more rewarding manoeuvres, and carry out the most optimal scoring path through the game. Whilst this might seem like a fundamentally modern form of competition enabled and mediated through the internet, this is effectively the model of competition used in track and field events throughout the ages (especially since the inventions of photography and video) – there may be sometimes competitions constrained to within a single day or event, but it is the wider global competitions which concern the highest-level competitors, and the attempts to reach ever more impressive physical feats and world fi rsts. Asymoptotic competition exists both against the game or the physical task itself, which must be mastered and improved, and against all those who are attempting the same. Therefore, crucially, in a danmaku game one is not just attempting to defeat opponents in the present – both those present in the game world, and the world record – but also to pre- emptively defeat opponents in the future. The higher a score a player attains, and the more distant that is from the prior world record, the more challenging we can reasonably assume it will be for anyone else to subsequently seize the world record. We can therefore understand all competition in danmaku games as being asynchronous, both in terms of players competing at different times and having no knowledge of each other (until a new record is achieved), but this asynchronicity also extends into the future, for each new world record exists with an implicit assumption of future competition, and therefore the higher any given record, the longer it is likely to stand for. This is, of course, the same model as individual athletics holds to, although as noted above, these are still primarily raced against others in the specific context of a given competition, giving a level of temporal specificity to when (and a spatial consistency to where) competition is carried out, unlike the spatially and temporally distributed competition of danmaku games. This is therefore a development of M Milla ’s , i po ta t o ept of i di e t P P , ut odelli g this as as mptotic o petitio a k o ledges oth the st i i g to a ds a pote tial pe fe t ga e as ell as against other players, and the projection into the future which necessitates maximizing your best possible score, as well the comparison with older high scores distinctive to indirect PvP. To return to the earlier arguments in this chapter, this has become even more relevant to the danmaku community now that competition is spread globally and virtually, rather than physically and confined to Japan. Asynchronous competition is simpler when the games can e pla ed f o o e’s home rather than in an arcade, and when scores can be immediately updated and uploaded online rather than having to check in with scores obtained a t other physical arcade venues. The Winner and the Million Losers We have now identified that danmaku games are focused for many players on a spacio- temporally distributed form of competition towards the highest possible score; that the pursuit of a high score rather than victory within the delineated space of a tournament venue leads to asymptotic competition, which sees players competing for every greater individual feats within a given competition no matter when or where these feats are accomplished; and that this means traditional concepts of spo ts a ship do ot eadil appl , a d e gagi g i su h p a ti e ould o l edu e o e’s a ilit to o pete at a high level. However, one question still remains, in this case about the role of all the players who are, simultaneously, not the world record holder. If there is only one world record holder at any given time, who – due to the irrelevance of traditional notions of sportsmanship in this context – seeks to optimize their score to the greatest extent possible, what can be said of the other players? Is every other player (as we saw earlier in this chapter, there are more such players these days than ever before) who competes at danmaku games but does ’t ea the o ld e o d a lose , hilst the pla e ith the e o d is the o l i e ?A d what, as Witkowski (2013, 161) asks, does an overt focus on winning as the sole drive for gameplay do for the experiences and cultures or those engaged in such practice? This section will explore these questions to conclude our understanding of the particular form of competition offered by (and culturally created around) danmaku games, arguing that despite the clear focus on winning, winning does not mean the same to all players, and danmaku games are often structured in a way to engage lower-skilled players in ways which do not necessary draw recourse to those competing at the highest level in the same games. Firstly, we must acknowledge that many players compete against their own personal bests rather than the global bests when playing (Crawford, 2013). There is still a vast range of possible skills and achievements which are not at world-record level: beating a danmaku game, beating it on higher difficulties, beating it without using a continue, and maximizing o e’s pe so al s o e. Whe a pla e loses – which is to say they lose their last life, their score is tallied up, and the game ends – a player generally perceives themselves as having lost to a certain challenge in the game, not the abstract score of the world record. This might be a particular boss, or a particular level, or a particularly challenging enemy. There is generally no built-in player-versus-player system in danmaku games, but rather in-game challenges which push the player to improve their skill solely in the context of their own ability, whether or not this subsequently expands to improving their skill in the context of other globally-dispersed players of the same game. Victory is seen as victory over those same in-game challenges, hi h a e i to ies the ga e ill t a k o a pe so al high-score list; only high-level players understand victory as relative to the scores achieved by other players and the global high-score list, and even then, one would not feel that one had been defeated the e o d holde if a pla e loses thei last life a d has to esta t the ga e. Even if competing for the record, an unsuccessful record attempt still feels as if one has simply not yet beaten the current record holder, rather than feeling a sense of defeat at the hands of the record holder. This brings us once more to asymptotic competition, showing that feelings of victory or defeat are therefore not resigned to a specific spa tial or temporal moments, and that competition can take place at any place and time. Such a model is somewhat distinct from individual sport competitions: most of those take place in situations (the Olympics, etc) where there are multiple individuals in a single arena competing for the same record (although such competitions also take place in less formalized contexts as well). Such a real-world meeting is extremely rare in danmaku games; most often the record is so ethi g o tologi all out the e a d a given player knows nothing of anyone else competing for it at a given time (and the current record holder will presumably have no idea that another specific player is competing until that unknown player beats their record). There are also other interim benefits for non-world-class players which suggest that winning and losing in danmaku are not clear cut. The seemingly impossible mass of bullets on the screen, once found to be navigable, yields a significant sense of achievement and mastery Ga eO e, ; i tu , this suppo ts M Milla ’s lai , that a o e dau ti g game makes players feel less upset when they are unskilled, as they can rationalize the perceived difficulty of the game from what they see on screen. We can therefore suggest that danmaku games are able to simultaneously offer significant subjective rewards to skilled players, whilst also giving less-skilled pla e s a get-out lause to e use comparatively low achievements. This is also particularly important when we consider the length of time and effort it takes any player to reach the highest levels of play. A top Japanese danmaku player once estimated that finishing the highest difficulty of one of the DoDonPachi games would take a new player four hours a day for a period of ten years Ke ps, , , a st iki g a ou t of the outi ised pe fo a es of the p a ti e of game-playing (Seo & Jung, 2014, 9) necessary to compete at the highest levels. As one Japanese danmaku pla e i C a fo d’s 013) documentary 100 Yen: The Japanese Arcade Experience puts it: it’s ot fu – it’s ou life to e the est . This a ot e see to appl to all danmaku players, however, and relegating everyone except the record holder to the status of a lose i the game seems indefensible; a more nuanced understanding is that o l the est pla e s o side i i g a d losi g i te s of the o ld e o d, hilst all other players hold different (but no more or less valid) alternate metrics for defining their successes and their failures. This is not to suggest that lower-level players do not engage with the world records at all, for many watch world record videos for their aesthetic and visual qualities rather than to educate themselves on strategy, but rather that there is a correlation nevertheless present between player skill and player understanding of the higher levels of interpersonal competition. The question of the world record still suffuses all danmaku games as it perhaps must suffuse any game with a score counter, but is only directly and immediately relevant to the gameplay of a smaller group of highly skilled competitors. In order to now come full- i le a k to the hapte ’s fi st e plo atio of danmaku in the Japanese context, an obvious question presents itself: has the skill at danmaku games spread out into the rest of the world along with interest in this genre? In a word, the answer to this question is no. Despite the last decade or two and the expansion of the genre into the West, a large percentage of all records (perhaps at high as ninety percent, although the previously-mentioned epistemological ambiguity over some records makes it challenging to make this number exact) remain held by Japanese players, and only a small number of danmaku games have their records held by non-Japanese competitors. One such record is held by the author (who is based in Europe); a number are held by several other strong Western players (primarily from North America); but that is where it ends. It is also worth noting that few of these non-Japanese records are held in the most well-known games; all Western records the author has been able to identify (including his own) are for danmaku ga es i the se o d tie of the ge e – with large playerbases and extremely high levels of competition, but not necessarily the required near-perfection of records in, for example, the DoDonPachi or Mushihimesama series of games. This is comparable to an observation from Rambusch et al (2007, 162) in their examination of CounterStrike, noting that it is not sufficient to merely be a top CS pla e , ut that o e also has to pla i a good tea that wins tournaments e phases i e – or, in this case, to earn records in the most ote o th ga es. The glo of i i g shi es ighte in proportion to the magnitude of the halle ge “ ith, , , a d hilst e e danmaku record is a feat which requires overcoming the immense challenge of these games discussed earlier in this chapter (Bailey, 2013; Davison, 2013b), not all records or tournaments (in all fields, not just gaming) are created equal. The extent to which the highest level of skill will globalize to the same extent as the games themselves therefore remains to be seen. Conclusion This chapter has examined danmaku games, and advanced several interrelated arguments. It firstly argued for several connected shifts in the genre in recent years – from a Japanese genre to a world genre (albeit one where dominant players still remain in Japan), from a genre played in physical arcades to one played primarily in the home, and from a model of record keeping based on arcade high-score lists towards record keeping which emphasizes online repositories of scores and their attendant videos. It first showed that from an origin in the Japanese arcade, a significant interest in games of this sort has emerged outside Japan, even if few non-Japanese players are (yet) able to beat their Japanese cou te pa ts at thei o ga e . This expansion was fuelled by a number of drivers, including the decisions of danmaku developers to experiment with porting their games to consoles, the development of Steam and other games platforms, and the iconic Ikaruga which brought the visuals and gameplay of the genre to an entirely new demographic. The chapter then examined in detail the distinctive form of competition which exists in danmaku, analyzing several related points. It examined previous work on professional, competitive or high-level gaming, primarily that into eSports and speedrunning, identifying points of commonality and difference with each, and analyzing the epistemology of the world record as opposed to the world champion, proposing the concept of asymptotic competition as a useful theoretical framework for understanding competition which strives towards excellence and perfection in play against the game as much as other players, and where competition is indirect between participants. It then considered the complex relationship between collaboration and competition in danmaku, and the absence of any clear notion of sportsmanship (not out of ethical lacks on the parts of the players, but due to the asymptotic structure of this competition which renders notions of sportsmanship non-applicable). It then examined the different rationales players use when engaging in danmaku play – since so few players are skilled enough to compete at the highest levels – noting the ability to compete against the ga e, agai st o e’s pe so al est, a d the i pli it desig hoi es of these ga es hi h discourages frustration and feelings of inadequacy amongst less skilled players. The chapter lastly observed that danmaku games remain dominated by Japanese players, although this does not discount the fact that many new players have found danmaku to be an enjoyable and engaging form of competition which avoids the moving goalposts of speedrunning and the public contests of eSports, whilst still demanding gaming skill every bit as significant as both these other two forms, and it remains to be seen whether the strongest competitors in this form of competition will spread beyond Japan in much the same way as the games themselves. Figure : “ ree shot of Ikaruga . Accessed on 21/5/2015, available from http://store.steampowered.com/app/253750/ Figure : “ ree shot of Cri zo Clover . A essed o /5/ 5, availa le fro http://store.steampowered.com/app/285440/ Bibliography Ao a a, Yuko & Izushi, H. . Ha d a e Gi i k o Cultu al I o atio ? Te h ologi al, Cultu al a d “o ial Fou datio s of the Japa ese Video Ga e I dust . Research Policy 23(3): 423-444. Baile , Tho as B. W. . The Da aku Ga e as a Ne Opti al A t, Pa t . Rhizome. Accessed on May 19 2015, available from http://www.gamestudies.org/0302/lee/ Betts, To . . A Histo of “h ups . Nullpointer. Accessed on May 19 2015, available from http://www.nullpointer.co.uk/content/endless-fire-a-history-of-the-shmup/ Boxill, Jan. 2002. Sports Ethics: An Anthology. Malden USA: John Wiley & Sons. Clea ate , Da id. . What Defi es Video Ga e Ge e? Thi ki g a out Ge e “tud afte the G eat Di ide . Loadi g…, 5(8), 28-49. Col ell, Joh & Cato, Makiko. . Video ga e pla i B itish a d Japa ese adoles e ts . Simulation and Gaming 36(4): 518-530 C a fo d, B ad. . Ye : The Japa ese A ade E pe ie e . A essed o Ma 2015, available from https://www.youtube.com/watch?v=2PfK2v02Pnk Daviso , Pete. a. Cu tai s fo ou: The Histo of Bullet Hell . USGamer.net. Accessed on May 19 2015, available from http://www.usgamer.net/articles/ten-great-bullet-hell- shooters De Koven, Bernard. 2013. The well-played game: A player's philosophy. MIT Press. Davison, Pete. 2013b. “Ten Great Bullet Hell Shooters”. USGamer.net. Accessed on May 19 2015, available from http://www.usgamer.net/articles/ten- great-bullet- hell-shooters De a, Ch ist . . E e gi g Pa ti ipato Cultu e P a ti es: Pla e -Created Tiers in Alte ate Realit Ga es . I Convergence, 14(1): 41-57. Fe a i, “i o . . e“po t a d the Hu a Bod : Fou datio fo a Popula Aestheti s . I Proceedings of DiGRA 2013: DeFragging Game Studies. Accessed on May 19 2015, available from http://www.digra.org/wp-content/uploads/digital-library/paper_387.pdf F a kli , “e . . O Ga e A t, Ci uit Be di g a d “peed u i g as Cou te -Practice: Ha d a d “oft No e iste e . Resetting Theory. Accessed on May 19 2015, available from http://www.ctheory.net/articles.aspx?id=609 Ga eO e. . Histo of “hooti g Ga e . A essed o Ma , a aila le f o https://vimeo.com/11393132 Hock-koo , “é astie . . Affo da es of Ellipti al Lea i g i A ade Video Ga es . I Proceedings of DiGRA Nordic 2012 Conference. Accessed on May 19 2015, available from http://www.digra.org/dl/db/12168.59440.pdf Keati g, J. W. “po ts a ship as a Mo al Catego . I Ethics, 75(1):25-35. Ke ps, H. . Re o t e a e des supe pla e s . IG Magazine, 16:165-179. M Milla , Luke. All of You Base a e Belo g to Us? “h ups as a “ou e fo Bette Ga e Desig . PhD Thesis. 2010. Accessed on May 19 2015, available from https://www120.secure.griffith.edu.au/rch/file/075ff367-71c0-00fd-f5f4- 5d4b221fc9b6/1/McMillan_2011_02Thesis.pdf M Milla , Luke. . The O igi of the “h up Ge e: A Histo i al “tud . Gamasutra. Accessed on May 19 2015, available from http://www.gamasutra.com/blogs/LukeMcMillan/20130206/186184/The_Origin_of_The_Sh mup_Genre_A_Historical_Study.php Me otti, Ga iel. . Video e as ga epla : ecording playthroughs and video game e gage e t . I Game: The Italian Journal of Game Studies, 3(1): 81-95. Ng, Natha ael. . UPLIFT: Desig i g fo Flo i A tio Ga es . A essed o Ma 2015, available from http://steel.lcc.gatech.edu/~nng/uplift/nng_uplift_designdoc_final1.pdf Novak, Jeannie. Game Development Essentials: An Introduction. Clifton Park, NY: Thomson Delmar Learning. Ra us h, Ja a. . Mi d Ga es E te ded: U de sta di g Ga epla as “ituated A ti it . PhD Thesis. Accessed on May 19 2015, available from http://www.diva- portal.org/smash/get/diva2:375941/FULLTEXT01.pdf Ra us h, Ja a; Jako sso , Pete a d Pa g a , Da iel. . E plo ing E-Sports: A Case Study of Gameplay in Counter-“t ike . I Situated Play, Proceedings of DiGRA 2007 Conference. Accessed on May 19 2015, available from http://www.digra.org/wp- content/uploads/digital-library/07313.16293.pdf Ree es, “tua t; B o , Ba a d Lau ie , E. . E pe ts at Pla : U de sta di g “killed E pe tise . I Games and Culture 4(3): 205-227. Scully-Blake , Rai fo est. . A P a ti ed P a ti e: “peedrunning Through Space with de Ce teau a d Vi ilio . I Game Studies 14(1). Accessed on May 19 2015, available from http://gamestudies.org/1401/articles/scullyblaker Seo, Yuri and Jung, Sang-Uk. . Be o d solita pla i o pute ga es: The so ial p a ti es of e“po ts . Journal of Consumer Culture. Accessed on May 19 2015, available from http://joc.sagepub.com/content/early/2015/05/11/1469540514553711 Sirlin, David. 2006. Playing to Win: Becoming the Champion. Accessed on May 19 2015, available from http://www.sirlin.net/ptw/ Smith, Steven G. 2011. Worth Doing. New York: State University of New York Press. Taylor, T. L. 2012. Raising the Stakes: E-Sports and the Professionalization of Computer Gaming. Cambridge, MA: MIT Press. Wag e , Mi hael. . O the “ ie tifi Rel a e of e“po t . I Proceedings of the 2006 International Conference on Internet Computing and Conference on Computer Game Development, Las Vegas, Nevada: CSREA Press. Wag e , Mi hael. . Co peti g i Metaga e Ga espa e: e“po t as the fi st professionalized computer metagames . I Space Time Play: Synergies between Computer Games, Architecture and Urbanism, edited by Friedrich von Borries, Steffen P Walz and Matthias Böttger, 182-186. Germany: Birkhäuser Architecture. We z, Ka i . . Theo afti g: K o ledge P odu tio a d “u eilla e . I Information, Communication and Society, 16(2): 178-193. Witko ski, E a. . P o i g the “po ti ess of e“po ts . I eSports Yearbook 2009, edited by Julia Christophers and Tobias Scholz, 53-56. Norderstedt, Germany: Books on Demand. Witko ski, E a. . O the digital pla i g field a d ho e do spo t ith et o ked o pute ga es . I Games and Culture 7(5): 349-374. Witko ski, E a. . Co petitio a d Coope atio . I The Routledge Companion to Video Game Studies, edited by Mark J. P. Wolf and Bernard Perron, 158-166. New York: Routledge.