What the Goldilocks principle can teach Canadian gamification researchers

Yaniv Corem is the CEO & Founder of Playful Meme, a gamification design & consultancy firm with offices in Boston & Israel. Corem is also a co-author of the research paper, Got Skillz? Player Matching, Mastery, and Engagement in Skill-Based Games, which was featured at the recent Gamification 2013 conference in Stratford, Ont. Corem, a former architect and MIT graduate, says “MIT instilled in me a deep respect and curiosity for data-driven exploration while my architectural background gave me the tools to design a great user experience. Gamification allows me to combine the two.”

CommerceLab recently interviewed Corem (via e-mail) about his work and his approach to gamification design.

Can you briefly describe your research in layman’s terms around player ratings and matching?

YC:  Much of the ongoing discussion about what is gamification touches on where a game ends and where gamification starts. It’s very common to hear people in the field saying, “Gamification is not about turning everything into a game” in a somewhat apologetic-defensive tone. And that’s true, gamification is not a game, but it does borrow many things from games, for example, the visual language, mechanics, et cetera. In our research, we try to push the envelope and explore what are the deeper constructs, models, and techniques of games, and how can we leverage them in gamification. In that regard, we decided to focus on the player and some of the techniques we could use to increase engagement.

Gamification is very much a skilled-based activity and we wanted to know how skill affects a player’s experience of the game. We based our hypothesis on Mihaly Csikszentmihalyi’s notion of Flow saying that in order for players to remain engaged, we must quantify their skill level (= player rating) and use this information to match them with a fair challenge (one that is not too easy nor too hard). Being in a state of flow (Csikszentmihalyi’s Flow Channel) results in players feeling that they are improving in the game, or in other words, on the path to mastery.

What inspired your research in this area? Why the Elo rating system?

YC: Most gamified systems today are a one trick pony, meaning, every user goes through the same experience (perhaps there’s some variance in the order or magnitude) but that result is ultimately very boring…once a user “gets it,” it stops being interesting. Using data to personalize the gamified experience is what inspires our research (and us). We decided to use the Elo rating system because it’s one of the oldest, most tested, and highly reliable ways of quantifying a player’s skill level in skill-based games. Also, Naor, one of my co-authors on the paper, is a Chess champion (Elo’s system was initially developed for Chess)…so that didn’t hurt.

Recently on CommerceLab: An Alberta researcher offers object lessons in the gamification of learning

Recently on CommerceLab: Dr. Who, Gamification pioneer

Explain the Goldilocks principle and why it may point to some of the problems in gamification today.

YC: The Goldilocks principle is essentially Csikszentmihalyi’s notion of a Flow Channel. It’s about keeping the player engaged by making sure challenges are neither too easy nor too hard. Different people use different ways to describe this principle but ultimately you don’t need a fancy word to explain something so intuitive. Understanding this principle and incorporating it into the design of your gamified system (for example, by using player ratings and matching) will determine if you end up with an epic win…or fail. There are so many gamified systems out there that are too easy…almost trivial…causing players to quickly lose interest. On the other hand, there’s also the risk of creating a gamified system with elaborate tasks that confuse and overwhelm the player.

How do you hope your findings will be incorporated or commercialized into gamification projects or initiatives?

YC:  First, I hope it’ll get people thinking and asking questions about how can they make their gamified systems more personalized and responsive. Second, we need to develop a gamification toolkit that includes a stack of methods necessary for creating and analyzing a great gamified experience…I’m pretty sure player rating and matching should be one of the first methods to go into that stack.

Is there any further research you’re doing now that will build upon what was presented in this paper?

YC: Absolutely, we’re seriously looking into player types, and, more specifically, rethinking Bartle’s player types. This is another area where industry has taken some piece of research conducted in an entirely different context (MUDs), and without properly investigating it, is making use of it in almost every gamification talk, presentation, or project. We’re interested in taking data from gamified systems and extracting new (and more scientifically-tested) profiles which, again, can then be used to personalize the user’s gamified experience.