Wednesday, September 12, 2012

Cutting Cookies with a Metal Press

A Deeper Look at Richard Bartle's Player Types, Part III
Screw Press
Cookie Cutter Overkill
"Don’t use a metal press to cut cookies!"
- Richard Bartle, Multi.Player Conference Keynote, July 21, 2011
Richard Bartle does not mince words when he describes how his landmark player type model has been misapplied by others. He used the metal press analogy above to describe how the gamification industry often applies Bartle's model without adapting it to fit the context.

If you're not familiar with Richard Bartle's Player Type model, part one of this series delved into the definitions of each player type, how players of different types interacted with each other and amongst themselves, and how multi-user games need to achieve a balance between types.  In part two of this series, we looked at Bartle's full model and how players move between types.  We also investgated how to design applications that support the entire user life cycle.

In part three, we delve deeper into Bartle's own critiques about the misuses of his theory and discuss a framework for how players types might be properly used outside of the narrow confines within which Bartle's model was originally conceived.

What Bartle Says
Again we will start with what Bartle himself has siad about the subject.  At the Gamification Summit in June 2012, Bartle addressed the topic of how game designers view gamification by saying that they would find gamification weird and would be appalled if their games were so bad that they had to bride people to play them. He also said that player type theory's main contribution to gamification should be that application designers now think hard about who they want to engage and how to account for different types of users. 

At the Multi.Player conference in July 2011, Bartle also addressed the misuses of player types.  In this instance, Bartle's main target was the barely social games played on social networks but his comments apply equally to gamification.

Bartle explained that his player type model was specifically designed to describe the behavior of people playing in virtual worlds for fun, and that he could not justify applying his theory beyond these bounds. The theoretical underpinnings of player type theory is the ability for virtual world players to undertake a hero's journey. These journeys take place in deeply immersive worlds where each player can develop a virtual identity over the course of playing the game.

Many gamification attempts do not have a persistent, automated, real-time, shared world that Bartle presupposes and usually don't have much opportunities for character development either. Gamification rarely involves virtual worlds, and in the case of enterprise applications, it very much deals with the real world. Many people do not view how they engage with work, a mobile app or shopping as necessarily fun, although we all know some who do and maybe more of us should.

Nonetheless, player type theory strikes a chord that other user typologies do not for the very reason that "fun is what gamification wishes to mine".  Bartle lists several alternative profiling systems that have failed to resonate with the gamification industry including the Minnesota Multiphasic Personality Inventory, the Five Factor Model, the Myers-Briggs Type Indicator, and the Reiss Motivation Profile. Formal approaches like the ones above tend to be too broad brush while informal approaches lack predictive power and rely too heavily on folk wisdom. Most of all, other approaches don't resonate because they aren't fun-centric.

Outside the Comfort Zone
Thinking of Venturing Outside of the Box
So, what are we to do? Bartle's player types should properly be limited to fun virtual worlds, and other profiling systems are inadequate. While Bartle himself is careful to avoid making applicability claims that aren't justified, he does approved of the well considered application of player type theory outside of its comfort zone. Fortunately, we've covered our bases well in the first two articles in this series and have developed a good understanding of what Bartle's model is and how and why it works.

With this knowledge, we can avoid rookie mistakes like rewarding explorers with achievements since we know that explorers really want more knowledge. Furthermore, we know how to tailor the way users progresses through an application to provide a differentiated experience as they go from novice to learner to veteran to elder.

Taking this approach still risks applying Bartle's types in contexts that don't make sense. So, if users are too far removed from Bartle's Killers, Achievers, Socializers, and Explorers, we can use a more abstract rending of Bartle's model to inform our application designs. In general terms, Bartle's theory teaches us that:
  • Different users will behave differently
  • Stable systems often require a balance of different user types,
    (So avoid designing systems that only appeal to users like yourself)
  • Shifts in individual user behavior will have predictable changes over their life cycle
Using these lessons, we can dispense with the specifics of Bartle's model and develop our own typology to suit the unique contexts of our applications.

Psychological Underpinnings of Bartle's Model
This leads us to our next question.  Are there any models that might help us develop our own user types?  Fortunately there is at least one.  In 2009, PhD student Monica Mayer independently verified Bartle's player types and explained them using Dietrich Dörner’s Psi-Theory of action regulation.  Mayer was able to describe Bartle's player types in psychological terms. Mayer found that each of Bartle's player types represents a preferred strategy of need satisfaction, and resulted from differing levels of need for affiliation, certainty and competence.

Stefan Strohschneider from the University of Bamberg summarizes the Psi-theory system using a metaphor of tanks filled with "motivational liquid" as depicted in Figure 1 below. 
Figure 1: Psi-Theory as a System of Tanks
According to Strohschneider:
"Each container has a set point for its level. Liquid gets lost either automatically (as is the case with hunger) or through environmental conditions - for instance, when one gets hurt physically or by harsh words. An intention to increase the container's level is generated as soon as the level drops below the set point. The satisfaction of a need then means that liquid is added and the container is again filled beyond its set point."
To see how this might work, let's examine the affiliation motive. If players' actual level of affiliation is below their set point, they will behave like Socializers and seek affirmative signals of social approval from other players. If they are over saturated with affiliation, they will not have the same need for social interaction and if fact may seek solo activities, very much like achievers and explorers. Finally, according to Mayer, killers are typically near the alarm level for affiliation (and other needs) and will often resort to anti-social behavior to gain attention.

Similar to Bartle, Mayer was investigating the motivations for behavior displayed while playing digital games for fun. However, since Psi-Theory underpins Mayer's findings, all of Dörner's needs may, separately or in combination, provide a useful framework for analyzing user behavior in non-game contexts.

Bartle's player type model is a powerful and succinct description of how people behave when they play in virtual worlds; how they interact with each other, and how their behavior may change. Joseph Campbell's heroic journey and Psi-theory both provide compelling insights in to the underlying reasons for why player types seem to work in so many contexts.

It is easy to see why the Bartle's model has become so popular with designers of not only virtual worlds but also of games in general and of non-game systems. Now that we have armed with a deeper understanding of the complexities and subtleties of Bartle's player types, I hope we will see many more effective applications of his groundbreaking ideas.  Now let's go out there and make the world more engaging and fun for everyone! I'm sure Richard Bartle would approve.

[This blog was also posted on]

No comments: