Surveillance technicity: affect,retention and videogame analytics

Introduction

We are situated within a surveillance culture, writes Lyon (2018). Monitoring ourselves, monitoring others and being monitored by others – often through digital or electronic technologies – are active and central parts of contemporary life. Articulating some bleak prognostications about the convergence of contemporary surveillance culture and capitalism, Zuboff (2015, 2019) has recently argued that the data derived from contemporary forms of technological surveillance have the capacity to yield significant knowledge about people’s behaviour, and the potential to profoundly shape that behaviour in future, for the benefit of large corporations that traffic in data (such as Google or Amazon). As other writers like Pasquale (2015) and Fourcade and Healy (2017) suggest, the plethora of digital sensors sedimented in everyday life operate as informational ‘dragnets’ (Fourcade and Healy, 2017: 12), shaping the structure of a range of different markets, and in particular how profits are made by large corporations in the finance and insurance sectors in command of a wealth of granular data about individuals’ tastes, risk and worth. Zuboff’s dire assessment of surveillance culture also dovetails with other perspectives on the political economy of data and surveillance, where data are understood as ‘infrastructure’ (Srnicek, 2017), or as a form of ‘capital’ (Sadowski, 2019). From such perspectives, data accumulation facilitates the continued use and development of technologies, enhancing their capacities as to create subjects more amenable to capitalism (e.g. the constant feeding of data into algorithms on platforms like Facebook or Instagram, as to make interfaces capable of accurately predicting what content we will likely engage with. See Bucher, 2018). In short, the power and control exerted through logics of accumulating and circulating data as a form of 21st-century capital (Sadowski, 2019; Srnicek, 2017), backed by technologies and techniques of surveillance, denote what Zuboff (2015, 2019) pithily terms ‘surveillance capitalism’.

Absent, however, from accounts of surveillance capitalism (in both Zuboff’s work and within the wider literature on political economy of data and surveillance) is a sense of how the technologies in deploy of surveillance capitalists actually shape processes of individuation – or the incremental, situated becoming (see Mackenzie, 2002; Simondon, 2017; Stiegler, 2010) – of those who use or are subject to them. While we know that most aspects of our everyday lives have become sites for the ever more efficient extraction of profit through data and surveillance capitalism, we lack sustained and critical attention to questions like ‘what does life amidst the technics of surveillance capitalism feel like?’. ¹

While a range of writers have theorised how human subjects individuate between their corporeality and data (Lupton, 2017; Smith, 2016), and others have more explicitly pointed out the relations between affective capitalism and surveillance (Andrejevic, 2011), we generally lack critical engagement with the biopolitical stakes of modern surveillance capitalism within the context of software interfaces and what we might call its technicity – that is, how the technologies and techniques of surveillance capitalism, and the circulation of data, are central components of our affects, embodiments and feelings of using software. My call for further attention to the technicity of the technologies of software’s surveillance capitalism resonates with more general calls to focus on the technicity of digital platforms such as social media (Bucher, 2012) and videogames (Crogan and Kennedy, 2009), which have productively oriented attention to the constitutive nature of media and technology interactions in contemporary life. Perspectives about surveillance capitalism and people’s phenomenological experience of software is pertinent in the present moment, where software reliant on surveillance is a ubiquitous part of our everyday lives, mediating our capacity to think, act and feel about the world.

To address the ‘technicity’ of surveillance capitalism within the context of software, this article is focused on the specific example of videogaming. Videogames are a globally popular activity, a central part of mediated life and a site of significant economic value. They are also a rich site for thinking about surveillance capitalism, particularly, for thinking about how the surveillant capture of user data conditions how users affect and are affected by the game software. To investigate this, I concentrate on a case study of the commercially available data analytics tool ‘DotaPlus’ for the popular multiplayer game Dota 2. This proprietary, subscription-based tool provides statistical information about players’ ingame performance, as well as dynamically updated guides for player activity directed by machine intelligence. While this article focuses on a single case study, it is representative of broader shifts playing out in the design of videogame interfaces (as we see in a range of other proprietary and third-party data applications available and in development ² ).

In interrogating the phenomenological implications of DotaPlus’ surveillance aspects, this article considers three different forms of surveillance that arise from different features of DotaPlus, each of which creating specific relations between player and game. These are (1) self-surveillance, (2) lateral surveillance and (3) autonomous machine surveillance. I argue that these different systems of surveillance produce what Slaby et al. (2019) call ‘affective arrangements’ – a non-subjective take on affect as situated relationality or ‘intensity’ (Massumi, 2002) between various human and nonhuman forces and conditions. Drawing from Slaby et al., I argue that particular systems of surveillance create particular relations between the various human and nonhuman actors involved in playing Dota 2, from which affective states, feelings or emotions felt by humans may emerge.

Specifically, I argue that DotaPlus works to make probable affective states that might impel continued consumption of Dota (and minimise the likelihood of negative affective states). To understand how DotaPlus mediates positive affective states, I focus on how data interfaces mediate the temporality of play (and the temporal consciousness of the player) – in doing so, changing up the affective arrangement of the gameplay ‘event’ (as assemblage of bodies and game technology interfacing with one another). To think human perception in terms of time is of course not without precedent and can be traced to 19th-century phenomenologists like Husserl, for whom human experience can be understood as segmented into distinguishable moments. Indeed, perception in relation to technology more generally has been long understood in terms of time (see, for example, Ash, 2017; Bucher, 2012; Dieter and Gauthier, 2019; Kinsley, 2015; Langlois, 2014; Mackenzie, 2002; Stiegler, 2009, 2011). Likewise, the experience of videogame play has been widely conceived in terms of time (Ash, 2015; Jayemanne, 2017; Keogh, 2018), characterised by its temporal ‘rhythms’ (Apperley, 2009) which emerge between player and game. Thinking about perception as temporal has also been productive in studies investigating how user attention is captured and maintained by software designers (see Ash, 2013, 2015; Dieter and Gauthier, 2019).

To develop the concept of surveillance technicity, to think about the mediation of temporal perception and affect via the surveillant technics of DotaPlus, I draw on the work of philosopher of technology Bernard Stiegler. Specifically, I mobilise his concepts of ‘retention’ and ‘grammatisation’. For Stiegler, human experience can be taken as temporal perception. This is understood in terms of the interplay between retention (or memory) and protention (or anticipation) which is conditioned by nonhuman technics. For Stiegler, technics contain traces of memory via a process of grammatisation; that is, of spatialising temporally constituted phenomena (such as actions, words, sounds and so on – performed or composed through memory – discretised as letters, pictures, numbers, etc.). This materiality of temporal perception is a dynamic he refers to as ‘technicity’ (Stielger, 1998). Stielger (2009, 2010, 2011) takes this premise to develop an innovative re-thinking of political economy around the coordinates of technics and the body, concerned with the industrial manipulation of our perception by the media and culture industries. He argues that our temporal consciousness is channelled and shaped through systems of technics, our technicity put in service of a hyperactive (and deeply corrosive) system of capitalism.

In this way, Stiegler offers a post-phenomenological perspective on how technics – such as those of the automated, ubiquitous and real-time tools of surveillance capitalism – might go about conditioning behaviour. By reading the case study of DotaPlus through the concept of surveillance capitalism and Stiegler’s retention and grammatisation, not only do we get a sense of how the most mundane aspects of everyday life are increasingly mined for profit but also a sense of the techniques and technologies used to do so, as well as their effect – producing certain experiential, affective qualities by transforming how time is spatialised through data-based game interfaces.^3,4 In summary, my foremost question in this article is how do DotaPlus’ technics of surveillance mediate economically desirable, positive affective states, making users more amenable to continued play and consumption?

My account of surveillance technicity, as a way to think through the relationship between surveillance and user experience of videogames as a contemporary and massively popular form of digital interface, is expected to be of interest to scholars researching software and mediation – contributing to an interest in interrogating how ‘software conditions our very existence’ (Kitchin and Dodge, 2011: ix); our everyday lives a process of individuating between our own individual somatic habits, memories, psychological perceptions and collective social and cultural milieu, but also through an exterior, technological milieu – which increasingly includes technologies reliant on the accumulation of data (Kitchin and Dodge, 2011; Smith, 2016). I also believe that the argument developed in this article will be pertinent to those studying the relation between software interfaces and people’s experiences of space-time (see Ash, 2012; Dieter and Gauthier, 2019), and more generally, researchers interested in examining the deployment and effects (and affects) of software as industrially produced digital media (Ash, 2015; Bucher, 2012, 2018).

In addition, this article makes contributions to scholars’ understanding of the relation between our voluntarily and involuntarily data emissions and the performance of human activity, with a range of research published on the subject in recent years (see Lupton, 2016; Smith, 2016). Despite videogame play being an extremely popular mediated activity, the relationship between user data analytics tools for the tracking and sorting of data about gameplay has received relatively little attention to date (see Egliston, 2019a). In this way, this article contributes to understandings of what I have elsewhere termed the turn towards ‘quantified play’ and an intensification of the role played by data in videogaming (Egliston, 2019a).

The article proceeds as follows. First, I outline my methodology. Then, I introduce and explain Dota 2 and DotaPlus. From there, I develop the concept of surveillance technicity. In doing so, I draw from Stiegler to theorise the relationship between surveillance capitalism, some of its associated technologies and techniques, and temporal perception (via Stiegler’s ideas to do with mediated memory). Further to this, I discuss how mediated existence can be considered in terms of affect – with technology conditioning particular affective states that matter to humans, and that may be profitable to the culture industries, by altering the temporal structure of experience. Following on, I examine DotaPlus through the concept of surveillance technicity. To take seriously surveillance’s often heterogenous nature, particularly within the context of new media, I present three discussion sections examining three main kinds of surveillance: self-surveillance, lateral surveillance and machine surveillance. Finally, I append the article with some brief conclusions.

Methodology

Consistent with previous work on the effects and affects of data and software, which has made use of ‘autoethnographic’ methods (see Ash, 2013; Bucher, 2012), my discussion of DotaPlus in this article derives from my own experience as a user of DotaPlus and a player of Dota 2. I draw upon reflections on my own play since the release of DotaPlus in March 2018 (until April 2019). During this time, I played for approximately 600 hours using DotaPlus. Throughout my play, I made written memos of my experiences as a technique of revisiting significant events for analysis. Such an approach – centred around close reflections upon my use of DotaPlus – is consistent with my aim to examine the relationship between technology, perception and affect. Particularly key in this project was my sensitivity towards the always more-than-human nature of experience, requiring close and constant reflection upon how I was always imbricated in a human–technical assemblage. While reflections upon my own experience present an illustrative view of the relationship between play and systems of surveillance, the insights emerging from a single researcher’s experience are of course limited in their generalisability. It is my hope that the meditations offered in this exploratory article encourage further empirical work on this burgeoning area of videogaming.

Dota 2 and DotaPlus

Dota 2 is a free-to-play, PC-based ‘Multiplayer-Online-Battle-Arena’ (MOBA) game, popular in most regions of the world. Taking on elements of strategy games like StarCraft and role-playing-games like Warcraft, gameplay involves two five-player teams attempting to destroy the opponents’ base, or ‘ancient’ (a win condition analogous to taking one’s ‘King’ in Chess). Each player adopts the role of one of 115 characters, called heroes, based on their preferred mode of play (each game features some combination of 10 unique heroes). Each of these heroes has distinctive play mechanics (usually exclusive to that hero). Due to the large number of unique heroes, the game is considered incredibly difficult.

Introduced by the game’s developer Valve Corporation in March 2018, DotaPlus is a paid subscription service for Dota 2 (priced at US$4/month). DotaPlus has a number of features characteristic of the ‘battle pass’ model for monetising games that has become popular in recent years (originating in Dota 2, but becoming popular in popular multiplayer games like Fortnite, Rocket League and PlayerUnknown’s Battlegrounds). In battle passes, users pay a subscription fee to access additional content not included in the free version of the game. Released 5 years after the initial launch of Dota 2, DotaPlus appeared at a point in Dota’s lifespan where the game was no longer significantly growing. As such, the move for Valve with DotaPlus was to more effectively monetise its existing playerbase.

The main selling point for DotaPlus, as marketed by Valve, is its analytics features. These features provide quasi-statistical readouts (numerically and visually) of gameplay. This covers both ingame information (e.g. the DotaPlus interface has a statistical overlay present in each ongoing match) as well as out of game (e.g. users can reflect on broad playerbase trends, broken down into tiers of skill). This includes information about both one’s own and others’ gameplay. As noted elsewhere (Egliston, 2019a), DotaPlus is marketed heavily as a tool for improving gameplay performance.

Through its various surveillance features, examined over the following sections, I discuss how DotaPlus represents a rich site for thinking about surveillance technicity, showing how the spatialisation of gameplay activity as data shapes player experience in a way that may encourage continued play and consumption.

Retention, affect and theorising the technicity of surveillance capitalism

In understanding how DotaPlus’ surveillant technics profitably reshape player experience qua temporality, I draw on Stiegler’s theory of retention. Retention features prominently in Stiegler’s (1998, 2009, 2011) work and is a crucial part of his broader theory of ‘epiphylogenesis’, developed over the Technics and Time series of books – doing so via the renovation of numerous continental thinkers (chiefly, Heidegger, Simondon, Husserl, Derrida and Leroi-Gourhan. Explaining exactly how he engages with each of these thinkers, however, is beyond the scope of this article). The crux of Stiegler’s theory is that human individuation (our ongoing ‘becoming’, after Simondon) is in large part about a relation to time. How we relate to time, however, is characterised as an ongoing exchange, or transduction, between humans and nonhuman technics. As Stiegler sees it, humans have always related to time through a complicity of cognitive faculties and equipment, tools and material environments (what he refers to as a condition of ‘technicity’).

Stiegler’s use of retention to describe temporality (after Husserl) refers specifically to the phenomena of how humans anticipate their futures and remember the past. In summary terms, how we remember the past and anticipate our futures is always mediated – implicitly or explicitly – by the presence of some sort of tool or technology. To elaborate, for Stiegler, we experience time as an interplay between our immediate perceptions (primary memory), recall or imagination (secondary memory) and our forward-leaning anticipations (or protention) – all of which are intimately related to, and moreover conditioned by technics (tertiary retention) – which work as exteriorised traces of memory or past action and that potentialise all further human activity. Technology as tertiary retention enables the movement of memory across individuals and groups over time (e.g. the transmission of somatic skills through tools which imply particular modes of use). This is a dynamic that Stiegler maintains as central to all human becoming – from cave paintings in the Upper-Palaeolithic era, to modern life amidst computers and mobile phones (with different epochs of technics constituting our experience of time differently).

How memory is materially retained in objects or environments operates through what Stiegler calls ‘grammatisation’, a process of spatialising or ‘discretising’ temporal or mental phenomena as ‘gramme’ (after Derrida). The gramme – an object of Derrida’s project of deconstruction, referring to the written mark – represents an un-opposing of the historically enduring opposition between thought and writing (and technics more generally). This is a dynamic that Stiegler radicalises in his theorisation of originary technicity, whereby gramme are the condition for human thought and action. Writing can be taken as a culturally foundational example of grammatisation. The discrete marks of orthographic writing represent grammatisations of the flows of speech, which in turn, has had clear epochal implications (e.g. the enduring influence of philosophical texts on modern thought).

While Stiegler’s (2009) initial objective in developing a philosophy of technics is to address a neglect of technics in Western philosophical thought, he expands his scope from Technics and Time, 2 onwards to develop a more politically orientated philosophy. Here, Stiegler begins to advance a critique of capitalism, the corrosive effects of which he suggests are amplified by modern technologies which aim to massively condition retentions, eliciting palpable shifts in how people negotiate their lives (see Stiegler, 2010: 73).

Media as a specific form of technics, Stiegler writes, create specific arrangements of retentions. As Stiegler sees it, this moulding of experience and cognition results in a standardisation of our experiential and cognitive faculties (which he argues has dire implications for the quality of individuation), a process that is often economically lucrative for the media industries. Simply put, Stiegler is interested in theorising that contemporary media technologies put the technical composition of consciousness in service of global capitalism and consumption. Recently, a range of writers have used a Stieglarian perspective to think about the way that software mediate temporality as to order various dimensions of everyday life in economically desirable ways (see Ash, 2015; Bucher, 2012; Kinsley, 2015; Langlois, 2014).

In mediating humans’ awareness of time, technics have clear implications for channelling affective states in humans – that is, the emotional content of our lives or how we think or feel about other human and nonhuman things within our environment (see Slaby et al., 2019). Affect and affective states (or emotion) are, however, widely understood as distinct. ⁵ Before proceeding further, I offer some definitional clarity.

Following its popular definition via Spinoza and Deleuze, affect refers to the intensities that characterise the process of affecting and being affected by the world (Siegworth and Gregg, 2010) – what can be understood in terms of becoming or individuation (to use the terminology of Simondon. See Simondon, 2017; Slaby et al., 2019). In this way, affect is not the property of human subjects – it denotes an ongoing exchange, or transduction, ⁶ between individual and milieu, a process that characterises being. In this way, affect is understood by Slaby et al. (2019) as an ‘arrangement’. Derived from situated arrangements of affect are ‘affective states’ (Slaby et al., 2019: 5) – named or ‘personalised’ affects (Massumi, 2002), or emotional content felt by humans (e.g. affect could denote the bodily sensations – such as the tingling of one’s body – provoked by an unsettling environment, whereas the named emotion might be fear). As Slaby et al. (2019) point out in their relational account of affect, affect is often materially (and increasingly technologically) conditioned. As Slaby (2018) suggests, material settings are often configured by humans as to make particular forms of affect more or less likely. To demonstrate this dynamic, Slaby et al. (2019) discuss the affective properties of the modern workspace and how its spatial and material characteristics can operate to capture and stabilise certain affective dispositions which might lead to productive labour. The design of consumer technologies is another good example – as others have shown – with specific technologies shaping particular kinds of affective states (see Ash, 2015). Through affective arrangements, such as technology, affective states are not straightforwardly generated. Rather, affects, that are not reducible to an intentional human subject are produced, arrangements which generate the potential for subjective affective states to emerge (potentially serving a particular end, for example, making people more amenable to using a product or service).

While Stiegler does not explicitly discuss affect, work adopting and building upon Stiegler’s philosophy has considered how the mediation of temporal awareness has the capacity to shape affective states (see Ash, 2015; Egliston, 2018). Following a retentional approach to think about affect is useful methodologically, inasmuch that it allows us to trace out some of the modes of relation comprising how individuation happens in a specific situation. By focusing on some of the specific technologies and techniques within an affective arrangement, we can map out how particular feelings or qualities come to matter for human subjects in an environment.

To consolidate the theoretical perspectives and concepts laid out in this section, my argument, developed through the example of DotaPlus, is that systems of data and surveillance (as a digital grammatisations of gameplay activity) contribute to a particular affective arrangement – of ways players can affect the game and be affected by it – which have the potential to lead to the emergence of affective states that matter to a human user. Specifically, I focus on the creation of affective states that might be deemed positive within the context of a competitive and difficult multiplayer game. Drawing from the concept of retention, the grammatised data interface changes up how users draw on their memory of how to do, and to perceive and anticipate, and ultimately act/play, which alters the affect of gameplay. Specifically, DotaPlus works to limit the range of affective states that might emerge from play for human subjects to those that might be seen as positive (e.g. pride from improving as a player, elation from winning more games), making these states more likely by changing up the retentional dynamics of play.

Keeping with Stiegler’s (2009) view that we ought to consider the materially specific dimensions of technology in evaluating how technologies condition perception (p. 6), and surveillance studies perspectives on the varied and materially specific nature of surveillance (Monahan and Murakami Wood, 2018), I focus on three different sites of surveillance within the DotaPlus tool – which differently achieve the common objective of manipulating users’ temporal perception as to create desirable gameplay experiences. These are self-surveillance, lateral surveillance and machine surveillance.

Self-surveillance

The first kind of surveillance at play in DotaPlus is what can be called self-surveillance (for a more in-depth account of self-surveillance in games, see Egliston, 2019a). This is not dissimilar to the kind of surveillance happening around quantified self-style technologies and practices (FitBits, Apple Watches, etc.) – where technologies are used for tracking of the practices and processes of everyday life, with the goal often to generate new forms of self-knowledge (e.g. for self-improvement or self-reflection) and to form new habits (see Lupton, 2016).

In this way, self-tracking applications do not simply quantify human activity. Rather, they produce new qualities and ways of experiencing the world, of feeling through our data (see Egliston, 2019a; Lupton, 2017) – a point resonating closely with Stiegler’s understanding of grammatisations – which condition ways of thinking, feeling or doing.

In shaping the affective experience of play through our own grammatised data, in using the DotaPlus interface from March 2018 to April 2019, I noted that it appeared to operate according to two different, opposed logics: of creating feelings of spatiotemporal proximity to and distance from one’s past performance. My framing of data in these terms is influenced here by literatures about data from sociology, media studies, geography and surveillance studies. Writers like Lupton (2017) and Stark (2016) discuss the ‘visceral’, sensory and embodied affective states provoked through data being part of our everyday lives. Others like Ball et al. (2016) adopt the language of ‘proximity’ to think about how our lives amidst ‘big data’ may regulate normative relations between body subject and world. Conversely, writers like Ash (2015) have discussed how data about user activity can offset negative affective states associated with a performance, by allowing users to help contextualise each activity as part of a broader trajectory of performance – creating a feeling of ‘distance’ from the affective states associated with any individual performance. To lay out this dynamic of oscillating between proximity and distance to one’s data, within the context of DotaPlus I give several examples.

I will first give a basic example of one of DotaPlus’ features – the ingame ‘Head Up Display’ (HUD). The HUD is a real-time monitor of gameplay performance which makes various forms of information an instantly retrievable modality of the present. The interface displays the user’s current performance relative to the average performance of a similarly skilled player at that particular moment of time in the game. These are shown in a small window at the top left-hand corner of the screen. What is tracked ranges from the straightforward (e.g. the number of opponents one has killed) to the much more complicated (e.g. net worth, the total accumulated resources of one’s character, which is very difficult to work out in the moment of play).

Players’ performance is also grammatised and relayed back to them in the moment of play through the ‘relic’ system. Relics are virtual items, exclusive to DotaPlus users, which capture statistics about a users’ activity with a particular hero. For example, a relic might display how many times a user has performed a particular style of kill on an opponent. This is displayed in glowing text above the player character’s head. In my previous research about Reddit DotaPlus users’ perceptions of self-tracking (Egliston, 2019a), some users suggested that the real-time display of past performance allowed them to retrieve affectively positive past memories, paralleling existing scholarly accounts of users’ affective connections with achievement points (Ash, 2012) and self-tracking data (Lupton, 2016: 73). In this way, by exteriorising secondary memory, users’ positive affective states in play are amplified.

Taken together, through the in-match interfaces, past performance is exteriorised and made present. Temporal perception, then, emerges in part out of relations between the game and data interface. Each moment I am negotiating the game is ‘fixed’ through the tertiary retention of the data interface, constantly updating as the game progresses. In this way, memories of past play are made constantly visible and more palpable, or ‘visceral’ (Lupton, 2017; Stark, 2016) a part of how we negotiate the game. Specifically, this may work to induce positive affects, such as the intensification of feelings of accomplishment or achievement in exceeding a performance goal.

Self-surveillance in DotaPlus features outside of the ingame interface – working to create spatiotemporal distance between the player and their previous performance. Users can navigate through the home menu of Dota 2 in order to view statistics about their career performance. In doing so, they can see their performance with each hero and are shown statistics such as the number of games played, win–loss ratio and so on. As previous work on analytics has pointed out, games can use statistical presentations of one’s ‘career performance’ in order to minimise players’ feelings of discouragement when they lose (see Ash, 2015: 109; Nansen et al., 2019).

Through the performance history interface, my affective state in playing Dota 2 is not just shaped through my gameplay – but rather through statistical exteriorisation of my prior gameplay performance. In this way, in looking at my performance summary of Dota 2 play, each match I play is framed as part of a broader career. Losing 11 games in a row, for example, may typically be a very frustrating experience. I may experience such anger or dejection that I no longer want to play the game. Through the performance summary, however, I can frame this poor performance within the wider context of my overall performance (which may be more positive).

Taking these two examples together, the recording and display of data to do with players’ own gameplay activities works, in DotaPlus, according to two different logics – maintaining proximity (by making our past performance a pronounced quality of the present moment), and also creating distance (positioning our current performance as part of a broader gameplay history) between the player and their gameplay data. These various ways of mediating retention have the potential to generate positive affects and feelings about gameplay experience.

Lateral surveillance

DotaPlus also affords users the ability to participate in what Andrejevic (2004) has termed ‘lateral surveillance’. Lateral surveillance denotes a mode of ‘peer’ surveillance; a mode of seeing others and also a mode of ourselves being seen (a dynamic of vulnerability that often works to shape how individuals act). Inflections of lateral surveillance have been offered around new media, such as social media (see Marwick’s (2012) account of ‘social surveillance’). Within the context of gaming, scholars have addressed lateral surveillance. We might think here of the work of Paul (2018) and Taylor (2006) on World of Warcraft and the governance of player activity through user-created mods, which in essence, quantify and rank one’s performance (according to various metrics – for example, the amount of healing or damage one has done). Despite the useful perspectives on surveillance afforded by this existing work, there is a lack of attention to the relationship between lateral surveillance and the phenomenological experience of gameplay.

One way that lateral surveillance happens in DotaPlus is through the ‘global hero trends’ feature. Here, users can view the efficacy and popularity of playstyles across any of the game’s seven separate skill brackets. For instance, looking at the game’s top skill brackets, I see that difficult to play heroes like ‘Templar Assassin’ and ‘Visage’ are very popular and effective, yet much less so in my current, lower skill bracket. From this information, I am able to draw from and habituate some of these techniques – familiarising myself with a wider range of gameplay approaches. ⁷

How I draw on my memories to develop forward-leaning expectations might then be mediated by familiarity with the top gameplay trends in a higher skill bracket, rather than simply drawing on my past experiences of gameplay. In this way, data interfaces that allow us to watch (but also be watched in forming part of the data set) provoke what Stiegler (2010) would call an organological ‘mutation’ – that is, the dynamic whereby new technical forms mediate new kinds of collective and individual memory. In this case, we see such a mutation in the interplay between collective playstyles, strategies and so on (shown via DotaPlus’ playerbase-wide statistics), and people’s own individual skills (e.g. somatic skills), ultimately changing up the technicity of playing Dota 2 – creating the potential for efficient negotiation of the game and the potential for positive affective states (e.g. from winning in using such techniques or from feeling a heightened sense of mastery in utilising a wider variety of heroes, etc.).

Another example of lateral surveillance in DotaPlus is the performance summary feature, giving subscribers a summary of player performance every time a teammate is killed. Users can post these data into the game’s chat window. For instance, in one match I played, a teammate makes the infelicitous choice to aggressively fight against an opponent who dealt heavy magic damage, without possessing any kind of item to mitigate this enormous damage. Unsurprisingly, my teammate was killed and removed from play for 120 seconds. I shift click the damage received pane which appears on the centre-left of my screen, which allows me to relay a detailed ‘log’ of gameplay information to my team. My teammate has taken 1400 magical damage from the enemy ‘Lion’ hero’s ‘Finger of Death’ ability, prompting them to follow an item build that mitigates some of this devastating damage.

We might then think of lateral surveillance as providing the opportunity for players to recognise past error and change the way that they engage the game in the future. In this way, the grammatisation of a player’s mistake in numerical or visual format shapes their ongoing experience of the match and their forward-leaning anticipations, in a way that might be positive or capacity building – potentially minimising frustrations associated with continual mistakes.

I also observed that lateral surveillance – through features like the ‘hero levelling system’ (where players accumulate and can display their skill with a particular hero via levels, from 1 to 25) – operates as a less ‘representational’ form of tertiary retention (e.g. traces of memory implicit in an environment, such as the traces left on the land via agricultural technologies and practices, which in turn imply ways of doing and being for future occupants). The levelling system might similarly be read as a kind of ‘trace’ of human activity within the virtual environment of Dota 2, implying particular forms of action for those within that space. A key example was of how the simple mechanism of player levels impelled more cohesive, team-orientated gameplay. This is particularly important in Dota 2, which is mostly played with ad hoc teams, sorted by an automated ‘matchmaking’ system (because of this, there is a tendency to focus on individualistic styles of play, which often leads to uncoordinated and dissatisfying gameplay experience). For example, in one match I played, a player opted to use a particular strategy that synergised with a high-level player on the team, recognising their skill with a particular hero due to the level system (and therefore wanting to maximise the impact of the high-level player as to more likely win the game). As they remark – partially in jest: ‘Holy shit, is that RTZ? [pro player] I’ll support’. In this way, the capacity to be visible, and also see others, is useful in creating a more cohesive team dynamic – encouraging particular ways of affecting the game which have potential positive outcomes.

This performance summary – as a kind of exteriorised memory – also transduces a transpersonal, atmospheric arrangement of affect (Anderson and Ash, 2015; Slaby et al., 2019). I observed that the lateral surveillance in Dota 2, and the dynamic of vulnerability, of always being tracked and visible by DotaPlus and other players, created an atmosphere of tension, impelling players to perform carefully and not make mistakes. ⁸ While vulnerability might commonly be thought of as an undesirable state, it can also yield positive affects (see Ash (2013) on the way that popular videogames demand a kind of corporeal vulnerability). As we see through lateral surveillance in DotaPlus, the creation of vulnerability is a key technique in intensifying the agonistic dimensions of Dota 2 play.

Taken together, lateral surveillance can transform the technicity of playing Dota 2, allowing users to reflect on their own and others’ past performance – exteriorised as data – in order potentialise more effective action in future. In addition, the lateral surveillance at play through DotaPlus, and the capacity to see and be seen, creates feelings of vulnerability and transduces particular transpersonal affective states among a team of five players.

Machine surveillance

The final kind of surveillance key to DotaPlus is the automated, machine surveillance that enables the machine-learning ‘PlusAssistant’ tool – a form of technical memory that directly informs gameplay performance. As Valve’s advertising copy describes it,

The Plus Assistant uses machine learning to suggest items and abilities for your current game. It uses the heroes that your teammates and enemies have selected, along with the items that you’ve purchased and abilities you’ve skilled up, to adaptively suggest popular options among winning teams. (Dota Team, 2018)

In this way, as I have argued elsewhere (Egliston, 2019b), DotaPlus can be described as what Stiegler terms a ‘mnemotechnic’ – a specific kind of tertiary retention that works expressly to shape memory, and moreover, individuation (often in service of capitalism, see Stiegler, 2011). The technological capture and relay of user data, much like Stiegler’s reading of audio–visual media (television and cinema), are key in the mnemotechical production of perceptual experience and individuation: ‘they bear witness for me . . . of a past that I enforcedly did not myself live’ (Stiegler, 2011: 28), he writes.

In line with this view of mnemotechnics, and in much the same vein of machine-learning programming (see, for example, Bucher, 2018; Mackenzie, 2013) and algorithmic culture, social software and prediction more generally (e.g. Kitchin and Dodge, 2011), the ways that individuals and collective playerbases negotiate – and specifically perceive and anticipate Dota 2 gameplay – is ordered by an already-there of statistical data. Specifically, I refer to the rendering of data about playerbase-wide gameplay activity as spatially discrete numbers, such as statistically weighted suggestions for what players should do in particular ingame scenarios (like choosing which ‘spell’ (or ability) to skill up on one’s hero). Each ingame ‘level’ a player accrues, the more powerful they are able to make a particular spell. The most appropriate decision (based on the percentage of players who have done so) is highlighted by a glowing gold square and a percentage figure (with the highest percentage corresponding to the ‘best’ choice). This creates the feeling of more fluid, competent, yet still relatively autonomous negotiation of the game.

The way that the game provides directions for the player to potentially follow resonates with ‘nudge’ techniques often used in interface design – understood as a form of choice architecture that alters people’s behaviour without outright forcing or preventing any options, increasingly prompted by the capture of user data at an individual and collective scale. ⁹ Within the context of DotaPlus, the soft power of nudging the player towards what is deemed to be the statistically optimal move does not completely militate against the possibility of autonomously negotiating the game – rather it represents, as Yeung (2017) puts it, ‘a “soft” mechanism of surveillant control’ (p. 129). By making suggestions about what to do, the interface creates a feeling that the player still has an active hand in negotiating the game, while de-emphasising the importance of accruing large amounts of secondary memory of how to do (potentially easing attentional load, and moreover, minimising the potential of negative affect – such as frustration associated with a steep learning curve). In this way, the potentially forbidding territory of Dota 2 play is made more user-friendly through orienting ‘nudges’.

Through machine-learning-based guides, the technicity of Dota 2 is altered significantly; specifically, shifting it towards an experience about pre-empting experience (rather than being an experience reliant on the accumulation of secondary memory of how to do, as is most videogame play). ¹⁰ To illustrate the relationship between anticipation and analytics as technical memory in the case of DotaPlus, the DotaPlus software provides recommendations on what has been statistically determined as the most appropriate hero, item or skill build given the circumstances of a particular game. By way of a simple example, in one game I played, an enemy picked the ‘Riki’ hero, a stealthy assassin who has the power of invisibility. The game then recommends the ‘Bounty Hunter’ hero – whose main gameplay mechanic is tracking heroes under the effect of invisibility. The game represents a technical exteriorisation of the knowledge one would accrue through playing Dota 2 for many hours, understanding the game’s many mechanics and the appropriate contexts in which to use certain abilities, heroes or items. Through my belief in the veracity of the system, as a user, I develop forward-leaning expectations and remember appropriate strategies through the statistical interface of DotaPlus.

According to Valve, such platforms allow players to reshape – in a positive, empowering, or capacity building way – the cognitive and somatic styles involved in play. This is done – as they see it – by giving players access to the most effective and context-appropriate techniques, reducing the need to draw on accumulated memories of how to play (based on extensive firsthand experience). Moreover, they might minimise risks of making poor decisions (and performing poorly), and work to create a sense of pre-emption, creating a constant awareness of the future as a predictable outcome from data about gameplay, enabling careful, timelier and more fluid responses to ingame events.

In summary, the machine surveillance used by Valve in order to develop a dynamic and responsive machine intelligence tool, with the objective of assisting players in gameplay, works as a system of technical memory which transforms the technicity of playing Dota 2. Put in more concrete terms, the conditions for performing skilfully are shifted through real-time presentation of grammatised information derived from nonhuman systems of surveillance – making more probable relations between player and game that might result in positive affective states for a human player (e.g. winning, not becoming frustrated by having limited knowledge).

Conclusion

Situated around a case study of the analytics tool DotaPlus in the game Dota 2, in this article, I have read the concept of surveillance capitalism (taken as the commercially motivated governance of behaviour via mass accumulation of data) with Stiegler’s notions of retention and grammatisation in order to think about how technical systems of surveillance operate as a form of tertiary memory. In doing so, I have offered the concept of surveillance technicity, a way to address some of the biopolitical stakes of surveillance capitalism, with particular focus on the effects and affects of data’s capture and circulation in the use of videogame software. Drawing from Stiegler’s account of the technicity of time, I addressed these affects by examining how the technologies and techniques of surveillance at play in Dota 2 – and the grammatised feeding of data to users – mediates the temporal perception of play. Specifically, I argue that data shape users’ capacities to remember or imagine, and their horizons of expectation (Stiegler, 1998); operating as a technique to profitably order users’ somatic actions, cognition and affective intensities. Mindful of surveillance’s heterogeneity and material specificity, I attended to three different forms of surveillance enabled by (and enabling) DotaPlus, all of which variously work to maximise positive affect and create captivated, attentive subjects.

Through developing the concept of surveillance technicity via the example of DotaPlus, this article has shown some of the increasingly prevalent techniques used by game developers to channel desirable forms of affect – capitalising on the actions that have already emerged from relations between player and game (grammatised as data) in order to create affective arrangements likely to yield positive affective states for players. This is expected to be of relevance to scholars studying videogames, attention capture and the mediated conditioning of player perception (see, for example, Ash, 2015). More broadly, the concept of surveillance technicity is expected to be pertinent to scholars thinking about how ubiquitous, real-time technologies and techniques of ‘data-veillance’ are deployed by surveillance capitalists in other aspects of everyday life, how they might regulate lived experience, and constitute a taken-for-granted part of our individuation.