- RT @TDaytonPM: Riding the inevitable curve from analytics to deep learning theplatform.net/2015/04/30/rid… @Google #hyperscale #bigdata @adataoinc 2 days ago
- Check out my answer to 'Is Concur as good as they market themselves?' on @founderdating bit.ly/1EIM7dy 3 days ago
biz | tech | psych | law
The Limits of the Game Layer
April 20, 2011Posted by on
A shorter version of this post appeared on BusinessInsider on May 13, 2011.
Seth Priebatsch’s “Game Layer” TEDxBoston speech is a wonderful introduction to what seems to be a relatively new idea: applying game mechanics to all sorts of situations to make people do what you want them to do. Seth believes that a properly-constructed “game layer” (a term coined by Seth) can solve really difficult problems—like how to improve failing schools—and even says that game mechanics can solve global warming.
Seth’s vision of the game layer has captured the attention of many at respected institutions, including Gartner (“By 2015, more than 50% of organizations that manage innovation processes will gamify those processes“), Ad Age (“the potential for the application of games and game mechanics is undeniable“), and the Huffington Post (“When unrelated game dynamics and incentives are added to social good actions, they can condition people to ‘do good’ for extrinsic rewards“), to name a few.
Despite Seth’s ambition, however, the game layer does not have limitless possibilities—in fact, the game layer is simply a twist on a number of different long-standing ideas in social psychology, and is limited in a similar way. This post defines the “game layer” in a rigorous way, explains how it fits into a larger body of experiments in social psychology, and outlines the limits that the game layer will eventually encounter.
The Game Layer: A Definition
Seth Priebatsch does not appear to have offered a rigorous definition of the game layer (his 10-second on-the-spot answer at SXSW 2011 defined it as taking game mechanics from computers and putting them into the “real world”), but his speeches (and examples) are good enough that I think I can define it. The game layer is a set of artificial limitations on and targeted feedback to peoples’ actions in a semi-public space designed to influence behavior. There are three key components that create the game: artificial limitations, targeted feedback, and semi-public space; and one core goal of the layer: influencing behavior. The next section of this post will go into detail on research into influencing behavior and each of the three components, but before doing so, I want to tie this definition back to the aforementioned TEDxBoston speech to show how my definition captures the essence of Seth’s vision of the game layer.
Seth implements his version of the game layer by using “game dynamics” (i.e., “game mechanics”), which are ways in which games are constructed to make them fun (and addictive). Seth describes four of these game dynamics in his TEDxBoston speech; I will describe each of them briefly and show how they fit into the above definition. The appointment dynamic is an artificial limitation that seeks to make someone do something somewhere at some specific time; Seth gives Happy Hour and crop-watering times in FarmVille as examples. Influence and status (Seth’s examples: AMEX black card, Valedictorian) are the results of targeted feedback in response to specific behavior presented in a semi-public place for others to see. Leveling up (Seth’s example: LinkedIN profile status) is targeted feedback in response to specific behavior, but is not in a semi-public place for others to see. Finally, communal discovery is targeted feedback in response to specific behavior with others in a semi-public space; Seth gives the DARPA balloon challenge as an example.
The Game Layer: A History
While the term “game layer” may be less than a year old, the concepts that underlie the game layer have been around for decades. If we want to understand where the game layer is going, our first stop should be to look backward to see how the game layer fits into existing academic research—specifically in social psychology. Social psychologists have been studying how to influence behavior for over one hundred years (see Norman Triplett’s social facilitation research, which found that bicyclists performed better in competition with each other at the same time rather than alone). Ultimately, the game layer is applied social psychology.
Kurt Lewin, a founder of social psychology, famously defined behavior as a function of a person and an environment. In other words, according to Lewin, what someone does is completely determined by who that person is and the situation in which that person finds him or herself. Thus, to influence someone’s behavior, it is easiest to alter that person’s environment because altering the person is much more difficult and time-consuming (and potentially impossible). The field of social psychology is essentially the investigation of what people do differently in different environments.
Social psychologists tend to prefer testing theories on how different environments will produce different results through experiments. Some of the more famous examples include Stanley Milgram’s Obedience to Authority experiment (if an authority figure takes responsibility for a person’s actions and gives strict orders, that person is likely to do anything) and Phil Zimbardo’s Stanford Prison Experiment (with enough environmental reinforcement, people will take on artificial circumstances as if they were real). And even though the main goal of most social psychologists is to understand human behavior better, some have written guides on how one might use social psychology to influence behavior. Perhaps the best-known, publicly-accessible reference on this subject is Influence: Science and Practice by Robert Cialdini. Influence has become one of the most widely-read books by marketers, and turned Cialdini into a high-priced speaker. Influence outlines six different ways that people can be influenced against their own rational interests, from “social proof” (e.g., agreeing to a known wrong answer) to scarcity (e.g., “for a limited time only”).
Seth Priebatsch describes his own work in a way that is similar to the research of social psychologists—he has studied which elements of games have the most impact on influencing behavior, and he has distilled them into a discrete set of “game dynamics” with clear definitions and examples so that they can be reused in other circumstances. Ultimately, however, the world of games is much smaller than the world examined by social psychology, and the world of games is much less focused on driving human behavior than games are. In other words, if the goal of the game layer is influence all kinds of human behavior, then it is very limiting to only study what is addicting and fun about games. It is perhaps even more important to study the ways that social psychologists have found to influence behavior.
The artificial limitations within the game layer operate in much the same way as artificial scarcity imposed through social psychological experiments. For example, social psychologists have shown that people will value cookbooks in short supply more than more-available cookbooks, and creative entrepreneurs have found that by etching a fly into urinals as a “target”, they can reduce 80% of “spillage”. More generically, the concept of “choice architecture,” coined by social scientists Cass Sunstein and Richard Thaler in their book Nudge, is that by setting up particular restrictions, we can retain each person’s sense of free will while still directing them toward the choices we wish them to make.
Perhaps “nudge” is also the best word to describe the goal of artificial limitations in the game layer: happy hour is at 5pm to nudge more people into the bar during a slow time and FarmVille crops need watering on a regular basis to nudge more people back into the game at constant intervals. And perhaps the best reason for the success of artificial limitations is described by Barry Schwartz’s The Paradox of Choice: people are happier when they are given a limited number of choices; too many choices and no choices both breed discontent. Games are historically limited (and often linear) in nature, which has not hurt their popularity.
The research shows that artificial limitations will make people happy and will “nudge” them toward doing specific behavior. From a gaming standpoint, the former has been a facet of gaming for years, but recently has been exploited by companies like Zynga to make games like FarmVille that have been criticized for “making people worry about [the game] when they’re away from their computer and drain[ing] attention from their everyday life.” The latter idea of “nudging” people toward specific behavior is what the game layer is adding, and it can only be helped by a thorough understanding of the related research in social psychology.
The vast majority of successful games provide constant and detailed feedback to players: points, levels, and money all make regular appearances in games to provide incentive to the player to continue. The related area of research in social psychology here is “rewards”. Many experiments in social psychology show that simplistic extrinsic rewards (like money and grades) produce inferior work to intrinsic rewards or no rewards at all (Alfie Kohn, although not an academic, has written the most comprehensive popular books arguing against the effectiveness of extrinsic rewards, including the best-seller Punished by Rewards).
Consider the following experiment conducted in a nursery school in 1973. Children who appeared to enjoy drawing were randomly divided into three groups: those who were told they would get a “Good Player” certificate for drawing (and then did), those who were surprised with “Good Player” certificates after drawing, and those who drew and did not get anything. Two weeks after the initial division, those children in the first group were less interested in drawing than before, but the children in the other two groups were just as interested in drawing as before. (See “Undermining Children’s Intrinsic Interest with Extrinsic Rewards: A Test of the ‘Over Justification’ Hypothesis” by M. Lepper, D. Green, and R. Nisbett, in the 1973 Volume of Journal of Personality of Social Psychology).
Seth addresses many of these same concerns in his speeches when he describes school as a “game that sucks.” Seth believes that a strategy of “leveling up” in school, building from Level 1 Newt to a Level 20 Paladin, would be much more successful at motivating students to care about doing schoolwork than the current grading system. I agree.
Successful games do not use simplistic, predictable rewards or punishments. One example that Seth holds up concerning targeted feedback, Call of Duty: Modern Warfare 2, has a reward system called “Perks” that get unlocked fairly unpredictably that give players different, unexpected talents (e.g., the “Sleight of Hand” perk lets players reload faster). Successful game feedback can also be negative, such as the Nintendo 64 Goldeneye 007 superlative of “Mostly Harmless”, given after a multi-player game to the least deadly combatant.
Regardless of whether the reward or punishment is in an academic experiment or in a video game, the same general principles apply in designing feedback to motivate behavior. The critical move that the game layer is making here, though, is from a game for fun to a game to drive specific behavior. The above and related research in social psychology will be just as important to designers in the game layer as examination of games and game mechanics.
Computer games have become significantly more addictive over the past 10 years as they have grown to include interactive play with friends and strangers. World of Warcraft has been called “more addictive than crack cocaine,” and it is no surprise that a “FarmVille”-type game needed the enormous user base of Facebook to become one of the most popular games of all time—FarmVille integrated into Facebook’s social features in order to spread like a virus. I use the term “semi-public space” to define the places where these games exist. They are rarely wholly public spaces, usually requiring the ownership of a computer or smart phone and access to fast internet, and also the leisure time to engage in playing. However, those requirements are not so exclusive as to prohibit those who are highly motivated to take part. Hence the term “semi-public space”.
Human interactions put the “social” in “social psychology”, and add significantly to one’s toolkit in creating environments to change behavior. Of Robert Cialdini’s six “weapons of influence”, only two (commitment/consistency and scarcity) don’t rely upon two or more people interacting. Or, as Seth might point out, the AMEX black card is only an effective status symbol if you can pull it out for others to see, and the DARPA balloon challenge wouldn’t have been solved without lots and lots of people interacting.
Those building the game layer will likely find better guidance about how to influence behavior using social techniques from social psychology than from existing games, because massively social games simply haven’t been around very long. For example, I have yet to see at least two of Cialdini’s six “weapons of influence” fully exploited in game design: reciprocity and social proof. There is a wealth of information in social psychology just waiting for the game layer.
Seth Preibatsch has said, “with 7 game dynamics, you can get anyone to do anything“. Unfortunately, that’s not true. Although the tools of the game layer are powerful, they are not omnipotent. At the end of the day, the game layer’s artificial limitations and targeted feedback in a semi-public space is a passive existence (that is, you must choose to play of your own free will), waiting for people to play and be influenced.
One could argue that if Facebook has added over 600 million users also leading a passive existence, why not the game layer? (Especially if the game layer is built on top of Facebook?) And I think it is reasonable to assume that all of the users of Facebook will have access to the game layer. However, the essential characteristic of the game layer that drastically limits its effectiveness is the desire Seth has for the game layer to influence behavior.
Social psychology has investigated the limits of our ability to influence behavior through things like artificial limitations and targeted feedback in a semi-public place, and has some fairly clear answers that the nascent industry around game dynamics and the game layer has not yet discovered. The main finding that social psychology has to offer here can be summed up by an old joke: How many psychiatrists does it take to change a light bulb? Answer: One, but it has to want to change.
Yale Psychologist John Bargh, in a fairly dense paper, discusses the extensive work that social psychologists have done to try and eliminate the effect of stereotypes on people. He shows convincingly that even in experiments where people are being closely scrutinized, stereotypes can only be countered if the subject has the time to prepare him or herself, the inclination to ignore stereotypes, and the mental ability to apply the relevant techniques. And even if all of those things are true, the impact of the anti-stereotype technique is usually fleeting—no more than an hour—and the standard stereotype effect returns.
The game layer may be a much more effective way to deliver things like anti-stereotype techniques, but it’s not going to be effective on people who reject that as a goal, or who don’t have the mental ability to achieve that goal. For example, if someone had wanted to use the game layer in 2007 to make people more motivated to push to get U.S. troops out of Iraq (remember, Seth argues that with seven game dynamics, “you can get anyone to do anything”), one would most likely use pictures or videos or other types of information. However, social psychologists have shown that individuals faced with an issue they view as political will not change their opinion on the issue, regardless of whether confounding facts are presented (see “Same Facts, Different Interpretations: Partisan Motivation and Opinion on Iraq“). In fact—and even more depressingly—confounding facts make partisans more confident of their original position, not less.
Recall Kurt Lewin’s formula: behavior is a function of a person and an environment. Ultimately, the game layer can only directly influence the environment. The biases that the person brings to the game cannot be controlled by the game layer, and so the game layer can never completely control behavior. Whatever good can be done through the game layer is ultimately limited by who plays the games.
The Game Layer: The Future
Seth Priebatsch is building his game layer by studying games, not social psychology. On one hand, this method is a bit like re-inventing the wheel, since social psychology can give much guidance. On the other hand, by studying games (and reading hundreds of pages on the effects of games each month), Seth remains at the forefront of innovation; a place where social psychologists rarely go. By studying which games are effective and why, Seth is trying to make boring or rote experiences (like getting discounts) more fun and engaging. I look forward to several years down the line, when more and more of my interactions with corporations can take place through the game layer.
However, the problem with the game layer is that it fundamentally relies upon volunteers. At the time of Seth’s TEDxBoston speech, his LinkedIN profile was at 85% (not 100%), and he was only at level 4 (out of 70) in Modern Warfare 2. My LinkedIN profile is at 75%, and I have never played Modern Warfare 2. And therein lies the critical limitation of the game layer: you have to have the time, ability, and inclination to play.