This is the text of the talk I gave at this morning’s Festival Of New MR session. I’ll follow it up with more posts and a reading list, but here’s the basic content. Of course, in this form you don’t get the various screenshots etc from retro videogames, but they’re not essential: the charts were intended as a kind of “hashtag” add-on to the talk - if you got what a game was and why it was on a slide, that’s a nice Easter Egg, but it’s not important for understanding the content.
SURVEYS ARE VIDEOGAMES
I started to become interested in research and games a year or so ago when an idea popped into my head: surveys are videogames. It made me smile, so I started looking for ways it might be true, then turned them into a short blog post. It got a few tweets, a couple of people mentioned it to me at conferences, but it didn’t seem 100% germane to the actual work I was doing, which has mostly been thinking about social media and communities.
But the idea wouldn’t leave my head. And I kept seeing social and mobile media blogs and articles talking about gaming, and game design, and something called “gamification” – making services and activities more like games. This concept had “buzz”. The most well-known example of gamification was a social media site called Foursquare, where by “checking in” to certain locations a user could unlock badges – achievements within the service, like a badge for checking in in several different countries. The user who checked in to a particular location most became the “mayor” of that place on Foursquare. Foursquare began life as a social site but the gaming elements proved to be its ‘killer app’ – it turned cities into gameboards.
The Foursquare concept has been much imitated. Site after site now offers badges and achievements to their users. As we’ll see later, there’s a backlash against this type of gamification, but still the buzz continues to build. One very enthusiastic blog post claimed that “game” will be to the 2010s what “social” was to the 2000s – an all-consuming buzzword, a kind of marketing holy grail.
Of course the danger of such buzzwords is that they become so ubiquitous as to be meaningless. The concept of gamification is that you can use “game mechanics” to change people’s behaviour – if you gamify something, goes the thinking, you raise the motivation to participate. But the flipside of that is that almost any marketing activity can end up being interpreted as a “game mechanic”, from loyalty cards to crowd-sourcing.
This presentation will look at game mechanics in a research context, and the first point I want to make is that in many ways we are already experts – our industry’s accumulated wisdom on incentivisation and survey design gives us a head start on ‘gamification’ if indeed that’s a path we want to go down.
So we come back to that initial idea: surveys are videogames. But how, exactly? In my view there are a bunch of things game designers do which we researchers do too. Not always well! But we do them.
BRAVE NEW WORLDS
The basis of game design is building artificial environments for participants to play in – from the 8x8 gameboard of chess to the complex virtual worlds of multiplayer online games. And this is also one way of thinking of a survey or community – an artificial environment you need your participants to progress through.
To do this successfully we need our ‘players’ not to get bored, frustrated, or drop out. We have to sustain their interest – as much as the topic allows. Beyond any rewards or achievements, a successful game or research project has to stimulate curiosity – what will happen if I get to the next level? Which question will I be asked next?
In qualitative research, and increasingly in online methods, we do this partly by encouraging interaction between players – setting up multi-participant environments like groups or communities. This creates its own emergent mechanics and behaviours – those operating in qual groups are well known, and the varieties of behaviour in online communities are becoming much better known to researchers.
Games further motivate their players by establishing goals which create the intrinsic motivation to succeed. Research projects also have goals – task and survey completion – and associated rewards, even if they’re not always intrinsic ones.
But simple in-game achievements aren’t always enough. The best games are designed to reward imagination and creativity in players, and we’re seeing new MR methods emerge – like crowdsourcing and similar social idea generation techniques – which aspire to doing the same.
And finally we share with games a relationship to storytelling. Games don’t always tell stories – instead they’re what designer Tom Armitage calls narrative exoskeletons: vehicles which empower players to create and live their own stories. I think research projects do something similar – we create structures from which stories emerge in collaboration with our clients and participants.
Of course in some ways we want research projects to be LESS like games. Competition between participants can be healthy and motivating but can also distort research’s primary role as a data collection tool. More serious is cheating. In research terms there are some activities which blatantly fall outside the core rules we establish – faking identities, for example.
But every set of rules also includes grey areas between obedience and cheating – activities which the rule set allows but which perhaps go against the spirit of the game. In surveys, an obvious example would be straightlining – moving rapidly through a survey in order to complete it more quickly. And here we run into the key difference between surveys and games – games build worlds and encourage the player to develop a strategy for dealing with those worlds effectively. Research projects build worlds but actively discourage strategy. Should this always be the case? After all, the feeling of mastering the environment is one of the most engaging and motivating in games.
GAMES AS INSPIRATION
This tension between what the designer intends people to do, and what they can do, is one of the ideas which comes into sharper relief when we start thinking about research projects as games. Once you make this comparison – once you start seeing research projects as environments for people to engage with – a whole vocabulary and set of concepts opens up. So this next section is about the inspiration we might draw from games as we explore them more.
For instance, one of the basic dynamics in games is between linear games, where the designer gives players specific paths to follow through a game, and the idea of the “sandbox” – an environment where players have more leeway to explore, play freely and carve out their own path. Platform games like Sonic The Hedgehog are linear, while multiplayer online games and games like the Grand Theft Auto franchise are sandboxes. Research has generally been a linear activity. With the advent of longer-term research communities we’re seeing sandbox projects become a reality.
But we can still learn a lot from their balance of compulsory and optional tasks. Most sandbox games give players the opportunity to explore within a loose structure of missions. Completing these missions is not the same as completing the world. Is this a model we can imagine working in a research community, as opposed to the rigid task-based model we often see used now?
Games can also teach us about the mechanics of intrinsic rewards – something we often take for granted, taking an “if we build it, they will come” approach which can be fatal for research communities. Behind apparently obvious reward mechanisms are complex processes which take into account different playing styles and different degrees of engagement. Take, for instance, the multiple meanings of the idea of “levels” in games – the word can mean levels of hierarchy and skill, levels of difficulty, or areas within games which a committed player can unlock. This isn’t just a semantic confusion! These represent three very different ideas of progress, ones likely to appeal to different participants.
Similarly, most games allow for different levels of completion - finishing the core game, but then also completing side activities, or achieving a certain degree of skill. This flexibility means they appeal to different levels of engagement. In current mobile hit Cut The Rope, for instance, my four year old son is motivated simply by winning each level – feeding a piece of candy to a hungry monster. I find myself wanting to find the solution that also lets me collect the maximum three gold stars – a very different challenge. This is something community designers can certainly learn from – creating tasks with different definitions of completion built in, for instance.
But the most important lesson we can learn from game design is this: game mechanics aren’t in themselves interesting.
What’s interesting, instead, is gameplay – what people do in the environment you’ve set up. The mechanics shape the play but they don’t always predict it. Chess, for instance, has relatively simple mechanics but a vast library of playing styles and tactics.
LOVE THE PLAYER, NOT THE GAME
And here’s the problem for world-builders like us: we don’t control the gameplay, we only control the mechanics. Which means that the mechanics we introduce can have unintended consequences. When social network Tumblr, for instance, introduced a mechanic for measuring bloggers’ popularity – called Tumblarity – it seemed like a classic application of game mechanics to community building. But it was hugely unpopular – it rewarded the mechanical posting of junk content rather than the genuine spreading of interesting material. The mechanic was quietly scrapped.
From our own research within Kantar, here’s a small example of this kind of unintended consequence. The humble progress bar has become a fixture of online surveys. From a respondent satisfaction perspective it seems a no-brainer – let people know how far they’ve got and they’ll be more likely to finish. But from a gaming perspective the continuous presence of the progress bar is a turn-off: players tune it out, and it doesn’t seem to represent any meaningful achievement. So we’re looking at progress bars which will appear occasionally, giving a respondent a minor sense of achievement at key stages in the task rather than continuous, but somewhat empty, reassurance.
So it’s ever more important to keep our focus on the users and respond to their real behaviour – and this goes for research respondents as much as game players.
This is where the hype around gamification and the more sober business of research may find themselves at cross purposes. The point of gamification, after all, is to influence behaviour – in many ways it’s a subsection of behavioural economics. As I’ve tried to demonstrate, it’s a powerful way of thinking about research, but influencing respondent behaviour has obvious implications for the data we collect, and we’re still in the infancy of understanding these fields.
For example, it’s pretty obvious that people in different lifestages will respond to game mechanics in different ways. Take Foursquare, for instance – the most popular example of gamification, with its systems of badges and mayorships to reward check-ins. It’s a service built by and for the young – once you have kids, your movements become more circumscribed. Who wants to be the mayor of their child’s nursery?
And even beyond this, different players have different styles. Some people are individualists, others prefer co-operative play. Some want to finish the game, others want to push against the rules. My wife loves a game called Zoo Tycoon, where you build a virtual zoo and populate it with animals. But when my brother played it, the first thing he did was to weaken the fences around the lion and polar bear cages so as to cause the animals to escape and devour visitors.
Different players will also have very different attitudes to risk. We see this in something as relatively simple as Monopoly – let alone the far more complex interactive environments we build as online communities, where the risks might involve exposing oneself or ones opinions in uncomfortable ways.
And here’s another tensions. In the worlds we build, we’re trying to encourage people to express their authentic selves – but in a game environment, designers are encouraging expression by means of an avatar – a virtual self. Resolving the tension between these will be a key challenge for researchers looking to use game mechanics in their work.
So taking an “if we game it, they will play” attitude – whether as marketer or researcher – is very risky. You’ll be rewarding some people and styles of play more than others, which means you’ll be rewarding certain types of people more than others. In a community, the more articulate and committed individuals rise to the top of the internal hierarchy – but are these hardcore players of the community game representative?
GAMES IN FLUX
And this is all happening at a time when ideas of games and their place in our lives are changing rapidly. I’ve talked about game mechanics and research in terms of building worlds, but it’s not just about that.
For one thing the worlds we do build are becoming sophisticated sources of data in their own right. Kantar company Red Dot Square was founded by game designers. They did the golf-type games you find in pubs and airports, like Golden Tee. They saw the opportunity for creating virtual shopping environments which participants could explore like actual stores. Most of the analytical richness the company provides comes from passively generated paradata: research via analysis of behaviour in an environment, rather than direct interaction with a respondent.
The wider shift in the nature of gaming, though, is being driven by mobile and casual games. Mobile devices allow games to interrupt the real world – people no longer have to set aside time for gaming, and so the game experience can become more casual and shorter.
Mobile is also the key to games’ new ability to transform the world – in fact this is at the heart of gamification. The idea is that a “game layer” can sit on top of the real world, transforming everyday activities into games. At the moment the mechanics for doing this are fairly crude – simple systems of action and reward - but augmented reality promises more sophisticated game-like interactions with real life.
This leaves “gamification” at something of a crossroads. Many people hate the word – they feel the current focus on badge and achievement based game mechanics, a la Foursquare and its many imitators, should simply be called “pointsification”, a virtual equivalent of the age-old reward card system which owes little to the rich and addictive experiences of real games. Others see these admittedly crude systems as the tip of an iceberg of behavioural change.
RESEARCH AT PLAY
So to close, some advice for researchers thinking of incorporating game mechanics into their work. First, mind the expectation gap – once you start calling things “games” or using game-like language such as “play” or “fun” to describe research tasks you inevitably raise expectations.
Second, don’t simply equate gamification with simple stimulus-response mechanics. Pushing buttons to gain predictable rewards – as in the famous Skinner Box mechanism researchers use in animal behaviour experiments - does seem to work as a short term lever of human activity. Just ask the millions who’ve found themselves hooked on Farmville! But ultimately it’s rewarding exactly the kind of routinised behaviour we’ve tried to eradicate from our panels.
Third, the area where the biggest advances in relating game mechanics to the real world are likely to be made – and where the most interesting approaches might be tried – is mobile. After all, we know a great deal about building environments, but a lot less about transforming them or interrupting them – the areas where mobile gaming is beginning to excel.
Similarly, we should learn from the games which have already been successful on mobile – simple, modular activities which can be completed quickly: very different to the typical survey or community task.
And even if research doesn’t adopt game mechanics, there are opportunities for it to work in partnership with games. Peanut Labs, for instance, offers players of Zynga’s immensely popular Facebook games the opportunity to win in-game currency and rewards in return for completing surveys.
And there will be opportunities too for researchers to contribute to assessing the effectiveness of game mechanics and gamification as a marketing tool – especially given, as I’ve said, the wealth of knowledge we already have in the industry around incentivisation and engagement.
I end with a quote, from an essay about the culture of gaming by Frank Lantz and Eric Zimmerman: “The game designer becomes a model for the architect, educator, scientist, artist, scholar, and others who invent and investigate the world.” As investigators of the consumer world, we should approach gaming with curiosity, excitement and caution. I hope this short introduction has inspired at least one of those emotions.