I attended the excellent Digital Surrey Event last week with guest speaker Michael Wu on the Science of Gamification. Digital Surrey is a vibrant not -for-profit business community of like-minded professionals who arrange regular events to “keep up-to-date with the ever changing digital landscape”. Although sponsored by various companies including thebluedoor and Dell, there was no overt marketing or promotion in evidence just a genuinely engaged crowd. The networking before and after Michael Wu’s talk was excellent and I was left being greatly impressed by the evening and the organisation behind it.
For an in-depth overview of Michael Wu’s talk have a look at Mark Wilsons’ blog here.
From the slides above and Mark Wilson’s blog you will see that Michael explained several models used to explain the Science – I have listed them below with a few of my reservations.
Fogg’s Behaviour Model
I found Michael’s explanation of this excellent; it undoubtably gives an excellent model to explain how motivation, ability and trigger converge to prompt action. However on it’s own I question whether FBM is somewhat limited in terms of gaining long term and sustained engagement.
Maslow’s Hierarchy of Means
Michael drew some very interesting parallels between Maslow’s deficiency and meta-needs with specific game mechanics and dynamics, however this hierarchy was drawn up in 1943 at a time when scientific thinking believed human beings to be a race of rational deciders – or homo economicus if you prefer. According to the science of Behavioral Economics, rather than being predictably rational, we are actually predictably irrational and often make poor decisions which negatively impact our access to Maslow’s needs. Therefore relying on this model is, to me, somewhat questionable. I covered this issue in some detail, relating gamifiction to Dan Pink’s Drive as well, in an earlier post here.
Watson and Skinner – Learning and Conditioning
Though Operant Conditioning has become something of a “dirty word” in gamification circles Michael provided a good overview of the use of Reward Schedules to sustain engagement.
As I have stated in several earlier posts I am a fan of Dan Pink’s work and clearly when one looks for an environment that sustains engagement and motivation by providing autonomy, mastery and purpose the Watson and Skinner model falls short by some degree as it provides none of these elements.
Michael provided an excellent description of this model, especially on how people move in and out of flow over time.
Flow is the optimum work state, the challenge for gamifiers is to understand how (and if) gamification can be used to achieve prolonged periods of Flow. In my opinion gamification and Flow run parallel but to some degree mutually exclusive courses – when a user is in a state of flow they are experiencing a high degree of engagement and are driven by purely intrinsic rewards – to then add extrinsic rewards is most likely to be counter-productive and risk devaluing the user’s sense of autonomy thus jeopardising the sustained enjoyment of Flow. In these terms gamification is best suited to monotonous tasks that do not prompt intrinsic rewards – in this circumstance gamification create a sense of autonomy, mastery and purpose which can then lead to a state of Flow.
Bartle’s Player Types
Michael used Bartles’ player types throughout the presentation however I found his understanding and interpretation of this model fundamentally inaccurate. Slide 26 of the presentation is unfortunately blank (I believe this is caused as it is an animated slide which slideshare does not support), however on this slide Michael listed the player types:
Where I believe Michael to be innaccurate is that he explained player types in terrms of demographics as in the population as a whole breaks down into people who fall in to each of the types; so a certain percentage of us are Socialisers, a percentage Achievers, Explorers or Killers, with the suggestion being that individuals fall in to only one type.
In fact the model actually says these roles are not mutually exclusive, the average player of social games is comprised of something like 80% Socialiser, 55% Explorer, 45% Achiever and 20% Killer.
When looked at in this way it is also fair to assume that a player’s type make-up changes dependent on circumstances and their perception of ability which creates a far more fluid model.
I covered this in a prior post here, which I have subsequently asked Richard Bartle to check and he has confirmed my understanding to be correct.
I will stress that I am a major advocate for Gamification and do agree with a lot of Michael Wu’s theories; I also found him to be a thoroughly nice chap in the brief conversation I had with him in the bar afterwards!