Wednesday, November 29, 2006

Teaching: Groups of Three

With my first set of courses done, I'm looking ahead to the next set. One course in particular offers an interesting challenge.

The course is basically a set of game design exercises, meant for small groups (three or four; I think five would be excessive). Each exercise would go something like this:
Tuesday in class: We play a game or set of games, and analyze them from a game design perspective. The choice of games is relevant to what we'll be doing for the project.
Tuesday-Thursday, homework: Students perform some kind of preliminary market analysis that will be relevant to the project, so they won't be starting from square one. For example, if they'll be designing a game to fit a license, this is where they'd take a look at the license.
Thursday in class: Project is assigned. Groups brainstorm ideas together, and submit their best idea at the end of class.
Thursday-Tuesday, homework: Groups take their best idea and flesh it out into a full concept (one or two pages). Something that would be presentable at a business meeting.

Now, the obvious problem here is that it's an awful lot of work outside of class.

Here's what I'd like to do: delegate the homeworks to one person in each group, and rotate it around. With groups of three students each, and nine projects total in the course, that means each student would have a total of six homeworks, none more than a page or two. Much more manageable.

But what do I do if the number of students isn't divisible by 3?

Friday, November 24, 2006

Culture Shock: Class Post-Mortems

In the game industry, it's traditional to do a post-mortem at the end of a project: everyone on the development team sits down and tries to identify what went right and what went wrong, as a way of avoiding past mistakes on future projects. I wanted to bring that same spirit to my classes, particularly the one that I'll be teaching again immediately after winter break.

Now, in both cases, brutal honesty is hard to come by. If you're working on a development team, your desire to be honest for the good of the company is tempered by your desire to not tell your boss why he sucks. Likewise, it's hard for students to tell a professor why his class sucks if he hasn't yet submitted final grades. So that much, I'm used to.

That said, there's a self-selection bias that exists in a classroom but not in industry. That is, at a game company everyone on the team shows up for the post-mortem, because it's part of their job and they're getting paid for it, so the people who got burned on the project will be there as advocates for change. After a class is over, the students who hated the experience are going to leave as soon as possible; they don't want to stick around for a post-mortem. The only students who remain will be the ones who were already loving the class and don't want it to end, so any comments will be skewed very positive. Meanwhile, the students who struggled and had real issues won't be there to let me know, so I'm likely to repeat my mistakes by catering to a certain breed of student while possibly alienating others. (Don't get me wrong. The students who participated made some great suggestions that I'm going to implement next time around; I just want more information from a wider range of perspectives, because that tends to lead to more surprises.)

I suppose that's what the student course evaluations are for, but I don't really trust those either. Many students don't take them seriously and will comment more on the professor's hair style than teaching ability; also, it doesn't give me the ability to ask targeted questions ("what do you think about Homework #4?") so the comments may be too general. Oh, and I had my evaluations done about two-thirds of the way through the course, so anything that happened in the last third won't be represented on the evals.

I was thinking of perhaps using class time to conduct a post-mortem, as a way of forcing the issue, but then I'd have to remove something else -- so I'd be cheating my current students out of valuable class time, in order to benefit the students next quarter. Not really fair.

Any other ideas of how I might get honest feedback from all my students -- both the ones that love the course, AND the ones that hate it?

Tuesday, November 21, 2006

Final is done!

Overall, it went very well, and I'll definitely consider this style of game-demo-exam in future classes. Things I learned, in no particular order:
  • I'm not exactly a whiz with video cameras. Even though the final was taped, I have no idea if it recorded sound, or even if it recorded at all, so it's a good thing I wrote everything down as I went.
  • Preparation is key, and I didn't do enough of it. In particular, I forgot to make sure there was a PS2 in the room, so I had to take a short recess in order to hunt one down and connect it. Time was very short, and delays like this were deadly.
  • Students were absolutely terrified going into the exam; this comes from the prospect of being asked only two or three direct questions, and what if I draw a total blank? As soon as we started, everyone was much more at ease -- it felt a lot like the discussions we've had in class all along -- and I think many of them forgot they were even taking an exam :)
  • Students had a tendency to repeat each others' answers, weighing in with their opinion, apparently forgetting that I was only asking for NEW information. These took up a lot of time, particularly on open-ended questions in which everyone had an opinion they wanted to share. I need a new rule to prevent getting bogged down on a single question; I'm considering limiting the "buzz-in" responses to something like three per question instead of unlimited.
  • I realized when I had asked a single round of questions that we were about an hour and a half into a two-hour exam, and decided on the fly to go immediately to the "lightning round", and then ask some bonus questions (individually written responses on index cards) with whatever time we had left. This ended up being a reasonable way to deal with time pressure. Unfortunately, it did mean that there was more of a focus on older games (and more class lecture, less reading and terminology) than I'd have liked.
  • This really only works with a very small class (<20),>
  • Ask the most interesting and thought-provoking questions first, so students can get into the spirit of the exam and really start paying attention. Save the most fun, entertaining questions for last, so students will leave the exam (and the course) on a positive note.
  • It helped to talk like a gameshow host, referring to students as "players" or "contestants", asking for a round of applause for our sponsors, and having parting gifts. First because it sounds so ridiculous that it really removes tension, second because it makes the video camera in the room feel more natural.

Monday, November 20, 2006

Marble Sadness

There are times in your life when you realize that it's nearly impossible to share some experiences of your youth with the next generation.

Those kids who saw the Star Wars movies starting with Episode I (instead of IV like the rest of us) will never be able to experience that shock of finally realizing, after all of the conflict, that Vader is actually Luke's father. You just can't un-see the new trilogy to experience the old trilogy with a clean slate.

Similarly, a couple of weeks ago I found (to my surprise) that most of my students had heard of Marble Madness, a wonderful arcade game first released around the time that today's college students were just being born. Why? Because the game has been ported to just about every console machine Atari could get its hands on. My students haven't seen an arcade model (many didn't even know it was an arcade game), but they've seen the NES version.

This is a shame, because there has never been (to my knowledge) a decent port of this game to a home system. The great thing about this game is its play control: you roll this heavy trackball and your marble on the screen moves in the same way. Converting it to a d-pad control simply doesn't work, because it removes that visceral rolling that forms the core of the game; it's like playing a dance/rhythm game with the sound turned off.

So, all of these kids today think of Marble Madness as this mediocre race game for console with lousy play control, not as a perfectly wonderful arcade game with the curse of never surviving a port intact. And that makes me sad.

Saturday, November 18, 2006

Teaching Backwards

Due to low enrollment in one of my Winter courses, the focus of the course has changed to something completely different. This happened all of a sudden at the very end of classes this quarter, so it's a bit unsettling.

I'm excited about the content though. This will be a pure design course, with no art or programming to get in the way. It's project-based: students will work (probably in groups) on a set of exercises. I'll cover some physical card and board games, and some paper designs (and project proposals) for digital games. I plan to keep the students on their toes by offering them interesting sets of constraints; I did this before on a semi-regular basis at Hi-Score, so expanding that into a full class shouldn't be too hard.

Here's the strange thing, though. This Fall, I taught a course about rapid prototyping and iterative design, from a technical game design perspective; this was a course where students were expected to actually implement their ideas, sometimes doing some light programming or scripting in the process. It's a pretty advanced course for undergraduates.
Meanwhile, this Winter I'll be teaching a practical game design class. And this Spring I'll teach a game design class that covers the theoretical foundations of the field.

Through a set of circumstances outside of curricular matters, I'll teach these courses in the reverse order that students should be taking them!

Lesson learned: next time I'm introducing a new set of courses to an existing curriculum, offer the lower-level ones with fewer prerequisites first.

Tuesday, November 14, 2006

Culture Shock: Going Independent

Occasionally, a jaded game developer gets fed up with harsh industry practices and decides to strike out on their own, starting their own game company to make their own games. This usually involves moving to a cheaper apartment and eating ramen for awhile.

And occasionally, I'd imagine a jaded university researcher gets fed up with stuffy academic bureaucracy and decides to strike out on their own, doing research as a commercial venture. This also involves ramen, unless they manage to secure a grant first.

If a teacher decides they've had enough of the system, though, there's not much they can really do. Tutoring, maybe, but it's not really the same. I suppose it's because academic programs are accredited; you could start your own university, but it takes millions of dollars, and there's no easy way to do it "on the cheap" that I can tell.

Not that I'm looking to do any of this, mind you. It's just one of those differences I noticed the other day.

Saturday, November 11, 2006

Final exam rules are finalized

Just finished writing up the rules for the interactive final. I'll hand them out in class Monday, but you get to see them here first!
  • Students get assigned a unique random number from 1 to 17 (there are 17 students in the class). Randomizing prevents anyone from arguing that I intentionally gave them harder questions, or that they got screwed by sitting next to the wrong people.
  • Starting with #1 and incrementing, each student is asked a question. (Most questions will be about a particular game that I'm demoing at the time). The student gives their best answer. Answers are scored as a normal test question.
  • After the student in the "hotseat" is done, anyone else may raise their hand to elaborate or disagree. If several students want to chime in at once, start at #17 and decrement. (In this way, a particularly bright student who keeps completing everyone else's answers can't keep control of the spotlight -- once they answer out of turn, everyone else gets a shot before they can try again.)
  • For those parts of the answer that the student elaborates on correctly, they split the points with the "hotseat" person, 50/50. In this way, the final is collaborative; even if you don't answer your own question fully, you can reclaim some of those points with help from other students. At the same time, students who answer other people's questions can get above 100% on the final. I expect this to have interesting ramifications on students' study strategies...
  • For those parts of the answer that the student elaborates on incorrectly, they lose a quarter of the full points (i.e. half of what they would have been entitled to). This is to discourage random guessing. Educated guesses are encouraged (I hope), since it's a 2-to-1 ratio of gains to losses.
  • If a student doesn't show up for the final at all, the questions that would have been theirs instead become bonus questions for the group. Everyone gets to answer on an index card, with points given (or taken away) as if they had answered out of turn. Students can choose to only answer partially, or not answer at all.
Time permitting, I'd like to add a "lightning round" at the end, quiz-show style: rapid-fire questions, all with very simple, clear answers, directed at each student in turn, with no answering out of turn. This will be the ideal place for those little tidbits of information that I want everyone to know about the industry, but that can't be demonstrated easily by playing a game.

The final is 2 hours long, which is frighteningly short if I'm asking even just two questions per student (it comes out to about 4 minutes per question on average, including time for me to ask the question, time for the student to answer and time for other students to chime in out of turn). Not sure what I'll do if I run out of time; I'll definitely be practicing everything ahead of time to make sure that I can breeze through my own parts, at least.

So, that's what I've got right now. I'd love to hear your comments.

Monday, November 06, 2006

Teaching: Oh yeah, study sheets!

One of my classes has a final exam. I was planning on actually designing the thing next week, since there's a solid week between the last lecture and the date of the final and I'll have nothing else to do in that time.

Until then, I was planning on working hard to get the last of the grading done, so students will know their grade going into the final. I always liked that as a student, being able to compute exactly how well I need to do on the final to get what grade. I'd like to do that for my students. But I don't really have time to do that AND come up with all of the questions for the final exam.

But then I realized I should really give out study sheets for the final exam (I'm ashamed that I had to be reminded of this by a student). But in order for those sheets to be accurate, I have to know the content of the final...

So, I either have to keep students in the dark about their grades, or keep them in the dark about the final exam, or take a guess of what content I'm going to test on the exam and then try to stick to that. I'll probably do the latter.

Now I know why in some classes, the study sheet would bear no resemblance to the actual content of the final... it's because the final wasn't written until after class was over!

Wednesday, November 01, 2006

Elementary Schools

Sometimes, just when I feel like I'm too old to be in touch with my students, they surprise me.

Today they surprised me by recognizing Math Blaster, Number Munchers, Carmen Sandiego and Oregon Trail from their own early childhoods. (This is, like, ten years after MY early childhood when I experienced these games.)

So, apparently the classics never die. Or maybe elementary schools are just hopelessly behind on the technology curve. Or maybe we just haven't had any decent educational games in the past ten years. Either way, it means something.