[game_edu] Brenda Braithwaite's game_edu rant at GDC

[Anthony Hart-Jones]

On 08/03/11 19:51, Bill Crosbie wrote:

> Anthony, you make a good point, but may I also point out that the 

> three years at university made you ready to excel in your internship. 

   I never did an internship.  I was one of those people who jumped 
straight into the deep end, using my degree experience as a starting 
point.  The trouble was that it was only really a starting point; the 
one group project we did to make a game ended with a complete failure 
and so my only development experience was my final year project.

   I am certainly not saying that academia is without use or even that 
apprenticeships are the right answer, but I do think many people learn 
better by doing instead of starting with a grounding in 'theory' and 
working through theoretical examples.
   Industrial placements (sandwich years) are great, but my university 
didn't have the links with games studios and so I had to search (and 
fail) for myself.  On the other hand, my first UK job was at a company 
who have a strong belief in employing placement students.  They had a 
university student on his placement year when I started and currently 
have a design placement for students on the Train2Game course.


> You weren't doing the specific tasks of industry in school, but you 

> were developing a mindset to be effective. So the question we are 

> grappling with here is how to extend that kind of experience to 

> everyone with the hunger to "go for it" and to do so from the start of 

> their design careers.

   Of 360 credits, I think maybe 60 were in subjects useful to the 
industry.  I mean, I learned assembler for a semester and 'Engineering 
Maths' was the same stuff I was learning at 16.  I can even build a 
working trebuchet thanks to my degree!

   The problem is that the degrees need a fixed number of points and 
there are required subjects to make it into the right kind of degree.  
The subjects that are useful are timeless, C++, engineering practices, 
even UML was still useful when last I checked, but (for example) we were 
taught Waterfall as the be-all and end-all of software development.  
Agile was unknown, or those who understood had not yet re-written the 
curriculum to accommodate it.

   If you want to teach kids how to make games, you need to get them 
making games and analysing the games they play.  Teach them C++ if you 
want, but heck; teach them D20 and GURPS too.  Teach your students to 
play Dungeons and Dragons, to analyse how it works, and even programmers 
will benefit.  At it's heart, D&D is just a game of Neverwinter Nights 
with the mechanics and systems exposed and visible.


> I love Jim's idea of going visual (gamemaker, processing) and 

> providing real projects. It reminds me of how I started coding (when 

> you had to enter source from the back of a magazine to play the 

> games... yes...I'm old.)

   Really?  That's how I started and I am only 30.  I would worry about 
Game Maker, given that it provides too much abstraction from the 
underlying system.  You can certainly go into the source code, but most 
students will not bother unless you push them.


> What strikes me about Jim's approach is it starts with a real goal and 

> relies on the students to work to make the end result possible. It 

> seems that in traditional CS curriculum we start with small building 

> blocks and then if we have time, ask the students to try to assemble 

> these pieces in to a coherent whole at the end of the semester. We 

> lose people along the way to boredom or confusion. There is no 

> framework to structure the individual components and so each concept 

> is seen in isolation from how it is used.

>

> Or something like that.

   I heartily agree.

   One of the best modules of my degree was making a simple 3D engine.  
The lecturer cobbled it together from MDSN help files, but it didn't 
change the fact that we were building a DirectX game engine from scratch 
and had working code (a spinning cube) by the end of the first lesson.  
Yes, he did start a couple of lessons with the words "I just found out I 
was doing something wrong, so we'll change it", but he kept our focus 
because we all knew what we were working towards.  In one semester, we 
had scratch built a simple toy that let you fly the pre-packaged biplane 
(from the SDK) around a scene.