malvasia bianca

In the January VGHVI Symposium, we discussed some of Roger’s thoughts on teaching. Which was a very interesting conversation, and I’d like to follow it up more. Unfortunately, I’m hampered for a couple of reasons:

• I haven’t been in a classroom at all for a couple of years, I haven’t been the primary instructor in a classroom for almost nine years, I haven’t seriously experimented with new ways of structuring courses for about eleven years.
• The symposium in question took place three weeks ago, I don’t trust myself to remember the details of Roger’s position, and he didn’t actually put a concrete position statement on the symposium blog post. (See the Pericles Group website for some information about his approach, though.)

So, in other words: what I’m about to do is talk about a woeful misrepresentation of somebody else’s point of view based on knowledge and experiences of my own that are equally woefully ill-informed and/or out of date. (Alternatively: I’m about to write a blog post! *rimshot*)

Roger sees a close tie between games and teaching, and had some sort of pithy phrase that he used to express that tie. I can’t remember what the phrase was, but I believe its gist was that classrooms are always a game, and that students are going to perform according to the rules of that game: so make active, conscious use of that fact, designing as good a game as possible and one where success in the game is as closely tied to your learning objectives as possible. And, as far as I can tell, he and his co-conspirators are extremely successful in this—I can’t imagine reading some of Kevin Ballestrini’s posts from last school year and not getting the feeling that something special is going on there. So I’d like to understand it, to relate to my own experiences and philosophical predispositions, and see what I can learn.

On which note: my philosophical predispositions towards teaching are strongly shaped by reading Alfie Kohn. His book No Contest had a huge effect on how I structured my classroom time; his book Punished by Rewards had a fairly strong effect on how I structured my assignments and grading, contributing to my feeling that I wasn’t a misfit in academia solely for research reasons, I ultimately was probably more of a misfit for teaching reasons, even though (because?) I cared about the latter more than the former.

And certainly there are many ways in which Kohn agrees with (my interpretation of) Roger’s point. For example, Kohn rails at length against standardized tests, and one of his main points is that standardized tests encourage students, teachers, entire school systems to do well on those tests even if that comes at the expense of learning; to me, this dovetails quite nicely with Roger seeing classes as games, because you’d better make sure that the rules of the game enforce the behavior that you want! Standardized tests are, of course, a lousy game with lousy goals; Roger does much better on that end, and I’m sure that Kohn agrees that the sort of richer feedback mechanisms that Roger’s methods provide are a huge improvement.

Where I suspect the two would disagree (or, more concretely: my reading of Kohn gives me pause) is on the nature of the motivators that are involved. The point of Punished by Rewards is that intrinsic motivation is much more powerful than extrinsic motivation, and that the latter drives out the former. Now, classes are already chock-full of extrinsic motivators (grades in particular); if you accept that as the basis that you’re starting from, then sure, craft your extrinsic motivators to promote learning in the areas that you’d like, and overlaying role-playing game mechanics may help with that. But if you start from an environment that’s trying to work with and nurture intrinsic motivators, then while role-playing sounds good, I get nervous about game mechanics: it’s hard to do that without bringing extrinsic motivators into play.

Looking at this from a slightly different angle: I like learning. I think feedback is inextricably bound to learning. But I’m a lot more dubious about certification: its coupling of feedback with extrinsic motivation can be actively counterproductive. And that coupling is often very strong, and is expressed as a refusal to give feedback without submitting to those extrinsic motivators: e.g. most colleges will kick students out of school if they refuse to engage in actions that lead towards them getting graded.

(Tangent: in my last year and a half in academia, I taught calculus. Those courses were full of pre-meds; as far as I can tell, the course served much more of a weeding out role than a thoughtful attempt to ensure that those students learned mathematical concepts that would help them be more effective doctors. Most of the students put in a decent effort to learn the material—you generally don’t get into Stanford without such habits—but not all were particularly interested; from my point of view, not being interested was a perfectly reasonable possible choice, indeed one that probably more of the class should make, and I did not enjoy working within a system with strong forces pushing against students making that choice, or even being aware of the possibility.)

So the question that that raises is: are games simply feedback mechanisms that can be used in a variety of ways, or are they certification mechanisms? I was going to say that, whenever you bring in scoring, you’re already moving in a certification direction, but upon reflection that’s too strong: if a game really is about itself (go or, I assume, Starcraft), then the scoring mechanism is feedback pure and simple.

But if the game is about something else (as classroom-based games always are, though Roger’s approach works at narrowing that gap), then scores make me very nervous. For one thing, if the score is tied to something else (e.g. a course grade that is necessary for getting a degree) then it’s certification, not simply feedback; for another thing, the distance between the score and the broader topic means that you aren’t getting feedback about aspects of the topic that aren’t covered by the scoring mechanism. I see both of these all the time in video game RPGs: if you don’t fight and level up, RPGs will refuse to give you access to the game’s content, and even if you are willing to go along with that, that focus on combat and leveling encourages you to neglect other aspects of role-playing. (Fortunately, there are people whose drive is strong enough to withstand such discouragement.)

If you’re sensitive to these issues (as I’m sure Roger is), you can design your games to open up as wide a space as possible for learning. Take Rock Band as an example; in this context, we’ll think of it as a tool to learn about music, e.g. by introducing you to a range of music, to help you pick out the different parts of a piece of music (Paul McCartney’s bass lines), even to teach you concrete physical and mental skills involved in playing music. The first iteration of the series was relatively prescriptive: it wouldn’t even let you try to play harder songs until you’d performed adequately (according to the game’s criteria, not your own!) on the easier songs. I suspect no-fail mode existed in the first game, but I felt that its use was discouraged; in contrast, the second game turned no-fail mode on by default if you’re playing in easy mode, so if you want to listen to music with a bit of guidance from the game as to the shape of one of the parts, you can do that without having the game punish you if you don’t conform properly.

By the third game, the amount and range of possible feedback has expanded enormously; because of that feedback, I’m finding the experience much more powerful as a teaching tool, with my actions being much less driven by the scoring mechanisms of the game. I almost always have no-fail mode turned on (and I wish there were a way to turn off the missed note sound: frequently I find that sound to be useful feedback, but in some circumstances it’s actively counterproductive to my learning goals), and while the game’s scoring system (and other metrics, e.g. streak length) can be a useful feedback mechanism (e.g. breaking a streak while playing Outer Space last weekend pointed out that I was missing a bass line transition), the extrinsic motivation aspects of that feedback, while still relevant to me, is no longer as dominant as it once was.

And with Rock Band 3 in particular, there’s feedback that’s provided outside of the game context, that your ears and hands give you. That game is, admittedly, a quite special case, but its nature may make it particularly well suited to provide examples for how to design games to work in a classroom situation.

Returning to what I said earlier: I’m convinced that Roger’s methods are effective, but I’m not sure I really understand the sources of that effectiveness. Continuing the theme of talking about areas that I’m ignorant of: how much of the effectiveness of these methods is due to a magic circle effect? Bringing in an explicit game mechanic (instead of the implicit mechanic that’s provided by grades and testing) may serve as an inoculation against extrinsic motivators, as an explicit acknowledgement of those motivators coupled with a refusal to give them undue power. And role-playing mechanisms in particular may be a particularly strong inoculation, with the dual role allowing for one of those roles to be motivated by intrinsic motivation while the other role goes along with the more certification-y aspects of the feedback systems.

Which, in turn, raises the question: what would a classroom look like with magic circle effects but without game mechanics? That puts an unexpected light on some of my own teaching experiences. One of the most powerful such experiences that I had was in the very first course I taught at Stanford: it was a differential equations course, and I’d spent a lot of time thinking about how I wanted to design the course. I balanced student work and lecturing in a very different way than in courses I’d been in as a student, and had a quite unusual homework / exam policy. I continued feeling this out as the quarter went along; I had a great time, the students seemed to be enjoying it, and the students seemed to be learning something.

So I was ready to declare the methods a success, and indeed I think the methods I used were good ones; but subsequent iterations of the class didn’t have the same feel. Part of that is doubtless chance (e.g. the specific students involved), and part of that is that I was less actively investing mental effort in the later iterations. But I bet that the fact that I was clearly experimenting had an impact on how the students saw the course, and did so in a way that’s similar to a magic circle effect, treating it as an explicit alternate space that muted the impact of certification on their learning.

Interesting stuff, I wish I understood the interplay of forces here better. I hope we’ll talk about this more in future VGHVI Symposia (of which there will be one this Thursday); follow the VGHVI blog if you want to participate!