rules of order policy-making as
game design? Sebastian Deterding (@dingstweets) Hans Bredow Institute for Media Research January 23, 2013, CPDP, Brussels cb
Overview 1. Premise: Gamification for
policy outcomes 2. Promise: Designing the polis for collective well-being 3. Perils: Governmentality masked as playful empowermentThe question the organisers of this panel kindly put to us is: Can we use elements of games and game design to improve policyoutcomes, e.g. in health or the environment? And what I would like to do as the opening talk is to ground and widen the debate a bit:First I will explore the premise of the question, to then argue that the promise of game design for politics writ large far extends »double-plussing« good citizen behaviour, and finally outline some issues one might run into in the course – including privacy implications.
1 premise Gamification for policy
outcomesSo on to the premise: Can we use game design elements to improve regulatory outcomes like healthier citizens, cleaner air, a morepolitically engaged public, and the like.
strategic trainingThe first thing to
note is that states using games is not a new phenomenon: The military has been using strategy games like Go orChess for strategic training for millennia.
planning & simulationWhich is echoed
in the 20th and 21st century in the use of strategic games and simulations for military and political planning andsimulation, notably during the Cold War.
active regulationWhat we see today
with »gamification« is that maybe for the first time games are being used as a regulatory tool to actively and directlysteer citizen behaviour.
HOW ALGORITHMS SHAPE OUR Code
is everywhereThe underlying technological enabler of this is the »robot-readable world«, »code/space« or »everyware« we live in today: A worldpervaded by ubiquitous sensors, processors, networks and actuators. More and more of our life world is run by software on a pervasiveand invisible digitally networked infrastructure.
Code is politicalAs we know
from Science and Technology Studies, this code/space has always been eminently political: Values, biases, powers getinscribed into and reproduced by our technological infrastructures.
code is used to shape
conductBut what we are seeing in the last ten-or-so years is that designers, psychologists, economists and politicians are trying to use thiscode/space – specifically web and mobile applications – to actively steer, seduce, or »nudge« behaviour.
»Code is law.« * *
and law enforcement is increasingly being encoded Lawrence Lessig code, version 2.0 (2006)That holds true for governments as well: Not only is code law, as Lessig famously put it – increasingly, law enforcement is beingencoded, offloaded into software and the infrastructure it operates on.
»Gamification« is but the most
recent outgrowth of this trend of the »codification of conduct«. The idea is that we can use this code/space, this world of ubiquitous sensors and algorithms and networks and actuators we already live in to put a »game layer« on top ofreality, to track our everyday activities in order to weave goals, rules and feedback systems into them ...
»What if we decided to
use everything we know about game design to fix what‘s wrong with reality?« Jane McGonigal reality is broken (2011: 7)… and by doing so, make these activities more fun, motivating, enjoyable. That way, we may »fix« all kinds of social issues and brokensocial systems.
healthLike health, as in the
case of Zamzee, which equips (primarily) kids with activity sensors, tracks how active they are, sets missions toachieve, and gives points, badges and virtual items for virtual characters.
2 promise Game design as
21st century policy-makingSo much for the premise. However, to really grasp the full promise of game design for policy-making, I would argue that we need toextend this standard story twofold: from »nudging« to systems design as a practice, and from regulatory outcomes to politics as awhole.
#1 ion ns teex For
policy-making and game design have a deep structural similarity: Both are essentially about »How to Do Things with Rules«. Both design rule systems to organise our coexistence for well-being – on the large as life scale with laws and regulations, on the scale of a shared afternoon with friends in the case of games.
politics, Sociology law Governance Social
order Public Policy Institutionalization Interpretation Scripts (STS) Economics computer science Game Theory Algorithms Behavioural economics Modeling, abstraction, Market/mechanism design automation, simulation Game design Design, Dynamics, ExperiencePolitics and game design are not the only fields interested in how to do things with rules, of course. Sociology tells us a lot about theworkings of implicit social rules (inscribed in tech), economics about the design of efficient market rules and mechanisms, computerscience about the practicalities of automating rules. But arguably, game design is special in that it potentially integrates all theseperspectives in an applied design practice that designs for intended behavioural dynamics and experiences.
filibusteringTake the phenomenon of filibustering:
It is just one of countless instances of »regulatory failure«, of unintended negative consequencesof a political rule system. For a game design, filibustering is only to be expected, because it understands rules not as a deterministicmechanism, but as part of a holistic system of rules and humans whose interaction will lead to emergent behaviours and experiences.
emergent behaviour and experienceTake Speed
Chess, for example. By »just adding« a rule of time constraint, Chess doesn’t become Chess, but only more (excitingly) so:The strategies of how to play drastically change, as does the experience: Speed Chess feels very different from regular Chess. http://www.flickr.com/photos/8147452@N05/2913356030/sizes/o/
Prototype, Playtest, iterateWhat game design
does to accommodate for this emergent quality is to build a real prototype of the system in question, playtest it withreal participants, learn from what works and what doesn’t, and based on that, tweak the system again, in a rapid cycle of iterations.
Mechanics Dynamics Aesthetics Marc LeBlanc
et al. mda: a formal approach to game design (2005)A good way of framing this is the MDA framework by Marc LeBlanc and others. Simply, it states that a game’s mechanics (rules) ininteraction with the players give non-deterministically rise to emergent behavioural dynamics, which again give non-deterministicallyrise to experiential aesthetics.
mechanics dynamics aesthetics +$ +
Slow poverty Frustrating -$ - gap End game MonopolyIn Monopoly, for instance, the rules state that if you buy streets, other people have to pay you when they land on them, increasing yourability to buy more, and if they haven’t enough cash, they have to sell their streets, reducing their chances of buying. In game play, thisleads to a slowly widening poverty gap of haves and have-nots. And because this gap opens only slowly but predictably, the long,protracted end game becomes very frustrating to the players who can see that they are going to lose.
game design policy-making mechanics dynamics
aesthetics +$ + Slow poverty Frustrating -$ - gap End game MonopolyNow map policy-making on this model, and you will see that it by-and-large is only concerned with the mechanics, guesstimating theirpresumed dynamics. Game design, on the other hands, looks at the full picture of mechanics, actual dynamics, and actual (motivatingor demotivating) experiences.
evidence-based policy-making?Now you may say:
Wait a minute, what about evidence-based policy-making (EBP)? Well, EBP simply uses whatever scientific evidenceit can find as input for the mechanics design, and sets up a (protracted and seldom working) »monitoring and learning framework« forafter a policy has been established. There is no rapid trying out and learning in the actual mechanics design process itself.
nudging?And nudging? True, »nudging« brings
in insights of behavioural economics in the mechanics design, but again, this mechanics designis thought of and done very linearly and deterministically, the black box of participants experiences is never really opened, and there is norapid testing out and iterating actual designs in their messy interaction with real-life participants and contexts.
market/mechanism design?But surely, economic market
and mechanism design can help us? Yes, both are eminently useful tools for solving tough distributionand other problems efficiently and fairly (like deciding who gets donated organs), but their view of the human being is still veryrationalist, and what they design is mechanics on paper, not living systems on the ground.
policy-making Rationalist hypothesising of linear
effects of »oughts« game design Holistic prototyping of systems for emergent behaviour and experienceSo in summary, whereas current policy-making and its tools are mostly engaged in a rationalist hypothesising of what effects might flowfrom the oughts they write into laws and regulations, game design takes a holistic, messy, and systemic view of humans and theirinteractions with systems, therefore iteratively prototyping and testing rule systems for emergent behaviours and experiences.
#2 ion ns teex institutional
regulation & public service non-institutional Such a broader view of the meaning of game design for politics also comes with a broader view of the possible application of game design in politic life, both institutional and non-institutionalised. The standard account of »driving policy outcomes« focuses exclusively on regulation and public service provision.
#2 ion politics ns teex
policies institutional regulation & public service civic participation well-being civic education civic monitoring non-institutional organising collective civic life action & voice Yet game design may help improve and transform many more processes in our political life.
regulation OutcomesThat does not mean
that we cannot (and should not) try to use it to drive regulatory outcomes, like improving the speediness of taxreturns by turning them into lottery tickets (if filed promptly).
public service experienceLikewise, we should
experiment with improving the experience of our public services through gameful design. Imagine your online taxform walking you through the process step by step like a good game tutorial.
Politics: process qualitiesYet we could
also use game design to improve politics itself. A central feature of politics is that it aims for two kinds of qualities at thesame time: outcome qualities (a good law), but also process qualities (a good democratic process). So: How might we design rulesystems that are less prone to filibustering or radicalisation, and more inviting to inclusion, diversity, deliberation, consensus-building?
Politics: process qualitiesI am thinking
here of examples like the gamified online deliberation platform Opinion Space. By placing your position in the total spaceof existing positions, and asking you to rate other positions by their insightfulness (rather than as right/wrong), it gently nudges youinto a more deliberative, self-relativising mindset.
civic educationWe can use gameful
and playful designs in civic education, for instance – why not? – to teach about online privacy, as in Six to Start’sonline game »Smokescreen«.
civic lifeWe might use them
to encourage people to engage in basic acts of civility, as supported by applications like Acts of Kindness.
civic lifeThough personally, I’d prefer
interventions in the style of former mayor of Bogota Antanas Mockus, who let mimes control traffic – andridicule rather than punish inattentive drivers. Instead of rote compliance and execution, such interventions create experiences thatinvite reflection, remind us of and reassert civic values (see the work by Karen Greiner and Arvind Singhal).
collective actionWe can use gameful
structures to aggregate and organise collective interests, voice, and action. An (admittedly crude) example was theGreenpeace anti-Volkswagen online campaign »Dark Side« that used points, badges and levels to organise and motivate participants.
civic participationWe can improve civic
participation in institutional political processes, like CommunityPlanit by Eric Gordon and colleagues, whichstructures traditional public hearings on city planning with missions and other game elements to not only make them more fun, butactually better-working, e.g. equalising the representation of louder and more introvert participants.
well-beingFinally, if the state (following
Aristotle) is there to enable us to survive and flourish, we have to talk about gameful interventions inpublic life to just improve our general well-being, like guerilla gardening games that motivate participants to green their cityenvironments.
+ + designing the code/space
of our polis for well-beingSo in summary, the promise entailed game design for politics is to design the code/space we live in together to improve both processesand outcomes for collective well-being.
uei ss #1 structuring participationThe
first issue is political science 101: You may celebrate systems like CommunityPlanit as increasing participation. But you also have toask: Who gets to partake? Who brings the necessary access and literacies? How is power distributed via the participation process? Andmost importantly: Who gets to decide about all this, and is she democratically legitimated in doing so?
uei ss #1 »The critical
problem raised by simulations is the black-box nature of the models.« Paul Starr, the seductions of sim (1994)(This entails an older argument in game studies: The rule systems in digital games are usually black-boxed. Hence, they quickly comeacross as a natural fiat – or rationality itself. Yet they are always built by someone with some interests and biases.)
uei ss #2 gaming the
systemThen there is gaming the system. For instance, when Bevan and Hood analysed the introduction of metrics and targets in the UK publichealth care system, they found that hospital workers gamed them as resistance. If the (unrealistic) target was to provide a bed to apatient 12 hours after admittance, trolleys standing on hospital floors were simply rechristened as hospital beds.
»The more a quantitative social
indicator is used for social decision-making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor.« Donald T. Campbell assessing the impact of planned social change (1976)This is not a new observation, of course. Already in the 1970s, sociologist Donald T. Campbell observed this internal contradiction ofgame-like systems of control. The very fact that policy values and goals are explicated into rules, quantitative indicators, targets, andattached consequences invites their corruption.
Reframing as strategic instrumental actionOnce
you create metrics and rules and targets, you communicate that the activity is not worth doing for its own sake, or some civicvalue. You invite people to become »rational actors«, strategic decision-makers focused on maximising their payout – the kind ofstrange creature that otherwise only lived in the Prisoners‘ Dilemmas of mathematic game theory and economics. http://www.rasmusen.org/x/images/pd.jpg
In games, we even have
a word for people afflicted by this: We call them »Munchkins«. To quote Wikipedia, »a munchkin seeks withinthe context of the game to amass the greatest power, score the most ‘kills’, and grab the most loot, no matter how deleterious theiractions are to the other players’ fun«. The munchkin forgets that he is not just a rational actor within the rule system, but also a socialactor within the larger social mesh of values and relations in which the rule system is embedded.
<Insert Dilbert cartoon here> fixating
thinking inside the systemManagement consultant James Rieley observed that munchkindom is pervasive in organisations: Once they introduce procedures andmetrics and KPIs, people start to focus on following procedure and meeting KPIs, not asking whether doing so is always actuallybeneficial for the organisation at large. (And I am sure you can think of many examples you have personally met in your work life.)
creating »negative externalities«The flipside of
this »thinking inside the box« are »negative externalities«, becoming externalities by the fact that they are not included inthe rule system. When BMW for instance introduced dashboards with competitions around fuel-efficient driving, people indeed diddrive much more fuel-efficiently. But they also did other things ...
So you also played EcoChallengeTM?…
like not braking at red lights. Because stopping and restarting would have used up fuel, you see? Because safety was not»internalised« explicitly in the system, it became a neglected »externality« for some participants.
undermining intrinsic motivationFurthermore, we know
from decades of psychological research that adding external incentives to activities people are alreadyintrinsically motivated to do may actually decrease that intrinsic motivation – among other reasons because people feel diminishedautonomy in their action, coerced and micro-managed by others.
detraining autonomous regulationOne documented long
term effect of this is that people do not learn and build their skills in autonomous self-regulation, becoming evenmore reliant on outer systems of control – arguably the opposite of the kind of citizen we would want to see in our democracies. http://www.ﬂickr.com/photos/courosa/4955407599/sizes/l/in/photostream/
crowding out civic valuesIncentives also
crowd out civic values: When researchers polled the citizens of the village Wolfenschießen whether they would accept anuclear waste repository built next to them if the parliament decided so, 51% agreed. When a financial incentive was added, supportdropped to 25%, even at 8,700US$/year and person: incentives re-framed the act as an economic exchange rather than civic virtue.
uei ss #3 face-value empowermentThe
third issue: Most of today’s gamified applications come across as huge empowerments of the individual: We help you becomefitter, happier, more productive – for instance by tracking what applications and website you use over the day, allowing you to setyourself goals and understand your own behaviour, as in the RescueTime app. But click on this little link here ...
… is covert surveillance… and
you will be lead to the »Manager Features« of the app, outlining how it can be used for collective time tracking, monitoring ofemployee activity, and controlling data access. The very fact that the empowerment of individuals relies on the tracking and storing ofdata on their behaviour on the servers of the provider opens the individuals to surveillance by third parties.
It doesn’t matter how much
rhetoric the app providers spin around this (RescueTime provides managers with a handy crib sheet how tosell the app to their team): Without these apps and their data storage on the app provider’s servers, there wouldn’t be the need for allthis privacy control and rhetoric to begin with. And many actors are not as conscientious as RescueTime seems to be.
… in healthThis issue –
face-value empowerment de facto opens the individual to behaviour tracking it would have never accepted as a sheerdemand from other parties otherwise – applies across all the areas we just discussed. Zamzee users open themselves to the possibilityemployers, health insurances, journalists might get interested in the honest picture of their health behaviours.
Marx, Das Kapital (Bd. III)
Congrats,Comrade! You just unlocked the Marx Badge for reading all volumes of Capital. That is ...interesting. Let me make a note. … everywhereIn essence, whereever you use a gamified app to improve yourself, you create a data trace of unprecedented consequentiality (actualbehaviours!).
… now with a playful
smileThe genuinely new thing about gamification here is that it frames these services as (a) self-empowerment and (b) non-serious play. It isan essential (and empirically open) question whether this reframing makes people more careless.
uei ss #4 Technologies of
the selfEven more insidiously – to move to our fourth issue – gamified self-help apps potentially do away with the need of surveillance byothers to begin with. To use the helpful framing by Michel Foucault, we can see them as »technologies of the self« that empower theindividual to determine and shape themselves.
… are technologies of dominationThe
catch, as Foucault points out, is that they all are technologies of domination at the same time. For they not only enable you todetermine yourself, they invite and eventually (once they become normal) expect you to. Because modern liberal democracies and post-industrial economies require and demand us to monitor, control, and optimise ourselves in lieu of direct outer force.
governmentality installation kitsSo gamified self-help
not only opens you to surveillance and control by others – they outsource that precise job back to you. Why spendmoney having others monitor you if we can get you to do that job for us yourself with the tools we provide you, even paying for theprivilege of willing self-subjugation?
uei ss #5 individualist rhetoric
aschange Implicit theory of social decoyThis logic of self-control pay onto a strong and problematic rhetoric in today’s society. Gamified apps implicitly communicate that all oftoday’s social issues – obesity, diabetes, broken education, global warming – are due to individuals not behaving themselves – usuallybecause they lack the willpower and determination to do so. So gamification offers them the tools to extend their willpower.
uei ss #6 virtualpolitikAnd this
leads us to the final issue, nicely illustrated by the Playpump, a contraption for water supply in developing nations popularisedby retired advertising executive Trevor Field. The pump replaces traditional pumping mechanics with a roundabout for kids to play with.The water reservoir is surrounded by billboards. Water would be pumped easy as child’s play, and advertising would pay for the pumps.
Sounds like a brilliant solution,
right? Indeed, the images of happy African kids playing on a roundabout made exceptionally good press,and it can still be found highlighted as »good design thinking« across all kinds of design blogs. Thanks to a very favourably PBSdocumentary aired in 2006, Field managed to get a commitment of 60 million US$ in aid for installing PlayPumps. But in 2009,problems started to surface: The pump was more costly and less efficient than existing solutions. It required maintenance by speciallytrained and approved PlayPump mechanics, such that many were left defunct once they broke. Advertisers interested in rural Africanpopulations mysteriously did not materialise. One calculation showed that children would have to operate the pump 27 continuoushours to pump the daily water demand of an average rural African village. Thus, women ended up working on the inefficientroundabouts, resulting in strained backs because they had to constantly bow down to operate a child-sized roundabout. In a word, themain purpose and success of the Playpump was media attention and good conscience in the developed nations, while the pump wasan utter failure for the actual people having to use it in developing nations. (See Ralph Borland’s excellent dissertation on the matter.)
If we follow Elizabeth Losh’s
recent rhetorical analysis of government new media use (including matters game-related), much of itfollows the same logic: Far from achieving actual impact, their main purpose is to signal to the public that government is cool,innovative, up to speed, and getting things done. Or in another word:
in summary1. IF (code =
law) THEN (regulation = coding)2. Gamification = coding motivation3. Gameful policy-making > motivating good behaviour through code4. Policy-making ≈ game design5. At best, gameful policy-making = designing the coded polis for collective well-being6.At worst, gameful policy-making = symbolic politics & governmentality masked as playful empowerment