If You Can’t Win, Cheat
Hello!
A longer piece today on people cheating at what I call "coercive gamification". Before that, a quick roundup of my recent appearances:
-
Clive Thompson interviewed me for his Wired column this month: QAnon Is Like a Game—a Most Dangerous Game
-
Noah Feldman interviewed me for his podcast Deep Background: The Allure of QAnon. It's amazing how just a bit of editing can make me sound much smarter than I feel.
- Radio Browser based its Surveillance vs Privacy episode around my Digital Sight Management piece, although it's not really about AR or digital sight management at all, but way richer and more joyful and from an African perspective on funerals in Covid; sex & masturbation; teaching old people how to have safe sex with HIV+ men; and documenting disappearing rituals.
I've also been posting daily on Metafilter for the last two months. It's been a really interesting and unexpectedly rewarding exercise, and I plan to do a longer piece about this soon.
Lots of people have been asking me to help make anti-QAnon ARGs. It ain't gonna happen, folks.
If You Can’t Win, Cheat
Coercive gamification turns all of us into cheaters. Should we feel bad about cheating our workplaces and health insurers?
I confess: I cheat at my Apple Watch.
In the Before Times, I used to travel from my home in Edinburgh to Six to Start’s office in London every fortnight. The train took four and a half hours, but I managed to fill it most ways with a mixture of work, reading, and Game of Thrones. The main irritation came from my watch, which nagged me ten minutes before the hour if I’ve been sitting too much.
By default, the Apple Watch displays an “Activity Rings” icon showing how much you’ve moved, exercised, and stood up. You can change your “Move” target, but the Watch insists you do 30 minutes of exercise and stand for at least a minute during 12 hours each day. Miss your goals (which are really Apple’s goals) and you’ll receive passive-aggressive reminders to do better. Hit them, and your Watch will reward you with a shiny achievement badge – and there’s plenty more gamification woven throughout the Watch, like competitions against friends and time-limited challenges.
I know it’s meaningless, but I see those rings every time I lift my wrist and it’s dagger to heart to see them incomplete. Most of the time it’s easy to fill them, but when I’m on a four hour train, there are fewer opportunities to stand. So here’s what I do: at one minute before the hour, I stand up for two minutes so I get “standing” credit for both hours. I can then sit easy for another 118 minutes before rousing myself once more.
Is this cheating? Yes – and no. Who am I really cheating – and when did I even choose to play this game?
Wherever there’s a game, there’s cheating. When I moved to London in the mid-2000s, I discovered I could get a fancy gym membership on the cheap via Prudential Health Insurance (now known as Vitality). Because I was a young and I didn’t smoke, my premium was low, but more importantly, if I earned enough Vitality points I got a discount on the gym membership such that I was practically paying nothing for both the health insurance and the gym.
You could earn Vitality points with self-reported information (“how much do you drink?”) but reaching the higher levels required more work, like buying vegetable from the supermarket, walking more steps during the day, or visiting the gym. I know for a fact that at my gym, several members would regularly tap in, sit in the cafe for fifteen minutes, and then leave – long enough to register as a proper visit. It sounds ridiculous but you could save hundreds of pounds a year by gaming the system. It might have done it once or twice myself. What can I say? It was fun.
These days, health insurers track your movement via your phone, which has led to a thriving trade in “swinging cradles” that make my Apple Watch hack look like child’s play.
It’s funny, but it’s also incredibly sad. Something has gone desperately wrong if insurers and employers are driving people to cheat at being “healthy”. If someone’s bone-tired after clocking off from their low-paid job, no amount of cheery gamification will convince them to walk 10,000 steps rather than drop their phone into a swinging cradle while they collapse on the sofa.
That people would cheat at games shouldn’t surprise us. Not long ago, a friend of mine told me about how his mother and mother-in-law were so obsessed with Draw Something that they’d cheat in order to continue their chain of correct guesses by writing the word “rose” instead of drawing a rose.
To be clear: the point of Draw Something is for one player to draw something and for the other player to correctly guess what it is. That’s how they started out playing, but they became fixated on seeing a particular score going up that when my friend exclaimed, “That’s cheating!” his mum strenuously disagreed. Harmless enough, but it shows our natural human aptitude for cheating.
I didn’t turn on the Activity Rings and its attendant gamification features when I set up my Apple Watch – they’re baked in, just as gamification is baked in to so many other tools and services we use today, willingly or not. What happens when you turn the world into a game, when you coerce people into playing? What happens when you raise the stakes so high people’s livelihoods depend on winning?
You train people to cheat.
Before we explore how people cheat at gamification, we have to take a brief detour to consider what counts as gamification, if only to address the inevitable objections that I left something out or included something that I shouldn’t.
There are two ways to draw the boundaries. The first way is to only consider cases where game mechanics have been explicitly or intentionally grafted on to existing systems in order to motivate people into behaving differently, like adding a leaderboard to a running app or points in a teaching app. This interpretation is pleasingly simple and covers pretty much all widely-known cases but has the downside of acting as if this kind of manipulation only sprang into being since the term was coined in the 2000s.
The second way solves that problem by including all systems constructed to manipulate behaviour that resemble games in the way that they operate and can themselves be gamed. This would include school league tables, frequent flyer programs, and political loyalty tests, which long predate gamification, videogames, and in some cases, even boardgames. This is admittedly stretching “gamification” – a transitive verb – to the breaking point because no-one is intentionally gamifying anything when they design these systems, but since the end result is the same, it’s still instructive to consider how these systems work and how people cheat at them alongside newer systems.
Taking this broader perspective turns gamification from being a novel curiosity to the latest fully-digitised, glossy version of very old systems of reward and punishment; systems where participants (or players) often have no way to opt-out, and where the prizes and penalties can be so big that some form of cheating is necessary to survive.
One of the best thinkers on this subject is Alfie Kohn, author of Punished by Rewards: The Trouble with Gold Stars, Incentive Plans, A’s, Praise and Other Bribes. Kohn eviscerates the behaviourist practice of using extrinsic rewards and punishments to motivate behaviour, especially when it comes to parenting and education, since it leads to rampant cheating and gaming:
Grades dilute the pleasure that a student experiences on successfully completing a task. They encourage cheating and strain the relationship between teacher and student. They reduce a student’s sense of control over his own fate and can induce a blind conformity to others’ wishes—sometimes to the point that students are alienated from their own preferences and don’t even know who they are.
This happens at the level of institutions, too. An ongoing battle in the UK has seen Ofsted, the schools regulators, claim that some schools are removing vulnerable pupils and forcing others to take “sham qualifications” in order to maximise results and improve league table positions, presumably in hope of attracting more funding and “better” students. The National Education Union, for its part, blames Ofsted for putting so much emphasis on testing and grades.
Kohn also criticises rewards and punishments in the workplace:
[Incentive plans lead to…] deliberate restriction of output, hidden jigs and fixtures, hidden production, fudged records … antagonism toward those who administer the plan, cynicism with respect to management’s integrity and fairness, indifference to the importance of collaboration with other parts of the organization (except for collusive efforts to defeat the incentive system.
Employees may do what is necessary for scoring higher on the evaluation form rather than what is genuinely needed for doing the job correctly,” one pair of analysts has pointed out. This often involves manipulating the schedule for completing one’s tasks—“gaming”—in order to get the reward. Salespeople may delay processing an order, or conversely, promise it sooner than it can be delivered, in order to qualify for an incentive based on sales recorded in a given period of time. They may also while away the hours monitoring the numbers that determine how much they will earn. And they may even engage in patently unethical and illegal behavior as a result of the pressures generated by incentive programs
If you agree with Goodhart’s Law, “When a measure becomes a target, it ceases to be a good measure,” Kohn tells you what happens when teaching, work, and life become dominated by those targets.
Punished by Rewards was first published in 1993, too early to comment on gamification. Thankfully, he treats us to his opinion in the latest 2018 edition, where he pronounces gamification:
a heavy-handed dystopian satire of Skinnerian behaviorism … It’s the equivalent of popping a treat into Rover’s mouth when he rolls over, and it’s built in to all sorts of apps, as well as being used by managers to measure and manipulate their subordinates. (But it’s, you know, a fun sort of manipulation.) Moreover, “most examples of workplace gamification are competitive in nature,” which means that one worker’s success comes at the price of a colleague’s failure.
Rather than a high tech innovation, Kohn sees gamification as a literal afternote on very old systems of quantified rewards and punishments, dating back centuries if not millennia. As such, discussions of gamification are incomplete without considering its ancestors and their flaws. And like those ancestors, gamification’s vulnerability to cheating highlights the chasm between its goals and its execution.
Cheat or Die
There’s a particular fascination in the west with Chinese people cheating at gamification. Maybe it’s because of the erroneous belief that “social credit scores” govern all life in China (in reality, it’s more a patchwork of multiple feuding, and largely incompatible, systems), or maybe it’s because China retains a quality of being just a few minutes in the future compared to the US and Europe.
In any case, last year the media turned its attention to “Study the Great Nation”, one of the most downloaded iPhone apps in China, reportedly with over 100 million registered users. It is basically gamified propaganda:
The app allows users to earn points for staying on top of news about Mr. Xi. Watching a video about his recent visit to France, for example, earns one point. Getting a perfect score on a quiz about his economic policies earns 10.
Worse:
The government is forcing people to use it and punishing those who cheat or fall behind.
Still, the threat of punishment hasn’t dissuaded enterprising users of Study the Great Nation:
Many reportedly asked their kids to help them download and answer questions so they could rank up on the leaderboard. And like leveling guides for video games, there are plenty of popular articles online guiding busy officials on how to earn points most efficiently. Tech-savvy users have even attempted to develop some “aimbot” programs to earn points automatically and top the leaderboard, but the app prevents that by giving points only if users spend a certain amount of time on each section.
From a distance, it’s easy to believe the Chinese government’s control over its citizens is monolithic, futuristic, and highly effective. The Great Firewall, social credit scores, ubiquitous facial recognition, and now gamification in the form of Study the Great Nation, they’re all assumed to work perfectly.
But they don’t. The fact it’s so easy to cheat at Study the Great Nation belies the narrative of complete control we’re told by both Chinese and western governments. Gamification can be intrusive and ineffective and worrisome, all at once.
China is interesting from a gamification perspective due to the sheer number of people its apps affect, but the US isn’t far behind. Amazon, Uber, Instacart, Microsoft, United Airlines, and Barclays also deploy gamification on millions of workers and they, too, punish those who fall behind:
You might race one-on-one as flying dragons against a nearby worker or compete as a floor against another floor for your Amazon mascot to run faster around the course. In another game, you attempt to complete missions that have you work faster and faster for longer periods of time. They also provide incentives that boost your score for returning from break faster. The programs have finely tuned tricks to tap into your mind so that we work harder, faster, and longer.
There’s another way to describe making workers labour longer for the same amount of pay: cutting their wages. So it’s no surprise that like in China, like everywhere else in the world, Amazon workers cheat:
The [Amazon] warehouse workers I encountered play games, against themselves or their coworkers. They cheat to artificially boost their productivity numbers. They pass these tricks around in coded language.
When workers cheat at games they’re coerced into playing, it’s not the same as financial traders using inside information to pad their millions. It’s literally a question of survival.
But the problem goes deeper: gamification doesn’t merely result in cheating but actively encourages it, and the reason is because what gamification considers cheating, most people would consider life. The fate of US truck drivers is an instructive tale.
Drivers are required track their hours in order to ensure compliance to strict “hours of service” rules set by the government. Until fairly recently, they used paper log books to account for every 15 minute segment of their time, including rest. As with all self-reporting, some drivers fudged their numbers. Sometimes the reasons were comparatively innocent, like fitting an unexpectedly long drive home into their hours of service. Other times, drivers have falsified their hours simply to make more money.
Starting from the 1980s, the US government began exploring the possibility of moving to electronic logging devices (ELDs) that would replace drivers’ paper log books with an automated, unforgeable system, so as to improve compliance with hours of service and thus increase road safety. These efforts repeatedly failed until 2012, when a rule was introduced to phase in ELD from December 2017 to December 2019.
To cut a long (and quite interesting, but beyond our current scope) story short, the fleet-owners who employ the vast majority of truckers are now equipping their ELDs with gamification features “to motivate individual drivers and teams to compete for better scores, badges, prizes and bonuses.”
In contrast, independent owner-operators and small-fleet drivers hate ELDs; they’ve always bristled at the rigidity of hours of service, but at least paper logs gave them flexibility to deal with what they say as the realities of driving. As one driver notes, “on paper logs we have 7.5 minutes of wiggle room, because they are broken down to 15-minute increments. So if in the morning I must log 15 minutes for fueling, but it only takes me 7 minutes to actually do it, then at night I can make up that time if it takes me a little longer to find a parking space.”
Do paper logs enable cheating and result in more road accidents from overtired drivers, compared to ELDs? No, according to a study in 2019. Compliance with hours of service increased following the introduction of ELDs, but there was no evidence the number of accidents decreased. In fact, ”drivers for small carriers appear to have increased their frequency of unsafe driving (e.g., speeding) in response to the productivity losses caused by the mandate.”
In other words, ELDs reduced one kind of cheating (hours of service) but increased a far more dangerous kind of cheating (speeding). They presumably also made drivers’ lives more miserable by reducing the wiggle room in how they worked and lived, and I can’t imagine the shiny gamification badges and prizes made up for it.
If you’re thinking, “I don’t work in a warehouse or drive for a living, so I’m all set,” think again. It’s tempting to assume that coercive gamification will be limited to lower-paid workers who don’t have the option to walk out the door. Surely the intrusiveness of surveillance technology would give managers pause before they impose it on less-replaceable workers? Thanks to one major bank, we know the answer is: no.
Earlier this year, Barclays introduced a pilot of Sapience’s computer surveillance system throughout the product control department of the investment banking division at their London headquarters. This fully-automated system monitored employees’ computers in real-time, instructing perceived slackers to spend more time “in the Zone” and “mute the phone, disable email/chat pop-ups, avoid breaks for 20+ minutes, 2–3 times a day.”
Drawing a direct link to Digital Taylorism, Sapience claims it has “the most cost effective and accurate way of doing time and motion studies,” time and motion studies being a ‘scientific management’ technique related to Taylor’s work with stopwatch timers, along with Lillian and Frank Gilbreth’s practice of filming workers’ motions to examine and improve repetitive physical work. Instead of filming workers’ movements, Sapience collects metadata about their computer activities, such as the websites they visit, their time on websites, and use of corporate software. Sapience uses this metadata as fuel to “improve employee engagement through customized games … with results aimed at achieving business goals.”
And as COVID-19 grinds on, you can expect this kind of surveillance and gamification to extend to your own home. Naturally, people are coming up with creative cheats, like hooking their mouse up to a very slow ceiling fan to simulate “activity”, but the fact this is even necessary is deeply depressing.
You’re doing it wrong
Even if you love gamification, there’ll be one part that annoys you: when the game fails to recognise your accomplishments. Since upgrading to watchOS 7, I’ve found myself ostentatiously washing my hands right under the faucet in order to get that sweet “20 seconds handwashed” Apple Watch reward, because I’ve found that its gyroscope, accelerometer, and microphone sensors often misses the first ten seconds of my washing.
In those cases, it feels like the game’s cheating me – and if you’re going to turn life into a game, at least get it right!
But maybe we should be careful what we wish for. If we had perfect sensors, if we had ethical companies and good government, couldn’t we use gamification for good? Wouldn’t that lead to a fairer, more just society?
And since there would still be cheaters, we could punish them based on how badly behaved they were – but it’s OK, since they could balance out any negative points out by good acts, and…
Well, if you’ve watched The Good Place, you know where this line of thinking ends – dystopian surveillance state where no-one is ever good enough, and where the most virtuous people are those who weigh up every tiny interaction with the world for its potential plusses and minuses.
But where did it start? Here’s Mike Schur, creator of the show, on his realisation:
… The nightmare for a person like me, who’s obsessed with rules, is that … nothing you actually did had any meaning. That you cut the line, and you cheated on your taxes, and you treated people badly, and it didn’t matter, you just got away with it. People getting away with it is really the thing I’m fighting against with this show, because the entire premise of the show is every tiny little thing you did mattered. It had a point value, those points accrued, and at the end of the day it was a videogame, you got a score, the top scorers get to, you know, [positive sound] and everybody else [negative sound] and that to me is just a way to try to enact a post-life world that I hope exists in some way, like, I hope it matters…
…As soon I invented that [consequentialist points-based system] system which I thought was, for the sake of the show, an absolute objective system, it was very clear that that system also sucks. There’s no way to do this that doesn’t have its massive problems, and so the journey of the show for the characters has largely been about them trying to untangle the specific ways that this system they’re in is bad and explaining to the audience why it’s bad and trying to fix it.
God knows I feel the same righteous anger Schur does when I see someone breaking the rules. But The Good Place, along with being fantastically funny and imagination, demonstrates the injustice that’d result when you invest so much oversight and power into a system that’s meant to evaluate and reward and punish all human behaviour. It cannot be done.
Beyond the distractions, beyond the gamified toothbrushes that inure us to self-monitoring, the most influential games are those we’re forced to play by our employers and by our governments. These games are about manipulation and enforcing a certain kind of society and behaviour. As such, they can’t abide cheating, even as they generate cheating through the act of measurement itself.
Dealing punishment to the punishable
In Debt: The First 5000 Years, David Graeber writes about the phenomenon of “moral confusion,” noting that throughout human history, most people seem to have agreed both that paying back one’s debts is the essence of morality and that moneylenders are evil. There’s an interesting parallel with gamification, where I think most would agree both that cheating is wrong and that forcing people to play a rigged game is evil. Something’s got to give.
In this case, I’m not talking about a specific game, although there are plenty of objectionable games employed by Amazon, Uber, McDonalds, United Airlines, recruitment agencies, trucking firms, customer support portals, call centres, and the like. I’m addressing any instance of gamification that connects “performance” with monetary reward or punishment. This type of coercive gamification gaslights us into thinking our poor performance – and correspondingly poor compensation – is our own fault, not that of those who have power over us.
It’s that power differential – born of hypercapitalism and inequality and surveillance technology – that enables coercive gamification to be wielded on so many. It reminds me of a line from The Eloquent Peasant, an Ancient Egyptian story: “Deal punishment to the punishable.” Four thousand years on, so many of us are still punishable, but hey, at least gamification cloaks that punishment in a language of reward and empowerment!
It’s insidious, really. We’re so used to believing that boardgames and videogames are designed to be fair, that every player has a decent chance of winning. But in many workplace gamification systems, if you lose, it’s because you were meant to lose.
Are workers are stupid for participating in a rigged game? No. They have no choice but to play. The companies I’ve mentioned are the dominant employer in some cities, attaining a monopsony status – the mirror image of a monopoly, where a single buyer, or in our case employer, can control the market as the major purchaser of goods or labour. In an economy where employment has become increasingly precarious, it’s hardly surprising that McDonalds and Amazon can dictate terms to workers.
As it becomes possible to track more and more of our behaviour via environment and wearable sensors, companies and governments can feed more metrics into how they dole out rewards and punishments. That also gives more opportunity and more justification to cheat, since any single infraction might be the one that tips you over the edge, and since any failure of the sensors that feed into the game might unjustly affect your score.
Ultimately, this kind of coercive gamification (which, to be clear, is not hypothetical) creates a cycle of distrust. The more you control people, the more you encourage them to rebel – so gamification may backfire because the control it exerts denies the autonomy people desperately desire.
But until gamification collapses in on itself and everyone has the choice to opt-out, it’ll inflict misery on billions, and turn people who like to think of themselves as honest players into habitual cheaters.
#notallgamification
I’m not a pollyanna. There will always be status and rewards, even in utopia; Iain M. Banks’ The Player of Games has a memorable example. But we shouldn’t turn basic human needs like a living wage and healthcare into a “reward” that needs to be won, where cheating will be endemic.
So yes, people cheat at their taxes (well, unless it’s automated) but you don’t want to create a situation where people have to cheat at taxes to survive. And for my own part, I try to make games where cheating is pointless. In Zombies, Run!, we eschew outsize rewards and punishments, and where we have leaderboards, there’s no prize or even a shout-out for coming first.
Instead, the game is enjoyable to play regardless of how fast you run or how many supplies you collect. That makes it harder to design, since you can’t rely on the crutch of shiny badges and points, but it also makes the game more fun and rewarding – which, believe it or not, happens to be good business.
But this is just dabbling at the shorefront of a vast ocean of horror. There is so much more to confront, and gamification is merely the tool of those who treat humans as things to be used rather than people deserving of respect. We need to excise every instance of coercive gamification in our lives, from school leaderboards and entrance exams that ration access to the best (or merely adequate) schools, exams that encourage families to cheat the system with tutors and donations; to fleet management software that encourages unhealthy and unsafe driving behaviour.
All too often, gamification is a warped and ineffective version of a fun, freeing activity, but coercive gamification is a step beyond the pale. We need make every school good and every job good. And if people are cheating at coercive gamification, we need to ask ourselves whether it’s their moral failing, or our failure as designers of the system.
Cheating is the symptom, not the disease. Cure the disease of coercive gamification, and you’ll make millions of lives that bit more worth living.