[MUD-Dev2] [DESIGN] Music Industry teaching the Game Industry (Was: A rant for Vanguard)

[Sean Howard]

"Nick Koranda" <nkk at eml.cc> wrote:

> The game design process seeks to find what players want in a game and
> design the game to fit those wants.  The music industry is different
> since bands dont ask what music they should create, they create music
> that they enjoy and hope others enjoy it as well.  The successful bands
> are the ones that have created music that has struck a cord (no pun
> intended) with a large populous.

That seems like a grossly simplified and idealistic version of the music
industry. There is quite a bit of manufactured music designed by study
group from a band of pre-teen males selected to fit a specific type of
hunkiness. You don't think Britney Spears writes her own music do you? She
has people who do that for her, just like she has agents, back up singers,
roadies, bouncers, and a whole host of people supporting her. Britney
Spears is a pretty face on what is ultimately the same sort of corporate
crap people accuse EA of putting out. The successful bands are the ones
designed to be successful and marketed to be successful.

But there is a stronger garage sentiment in the music industry than in the
game industry. It is far, far, far more likely for a small band of
teenagers to create a band in their garage and get signed - amateur game
developers don't get that because we are too busy judging the game
industry by the Britney Spears and not the Ramones.

> I would imagine that the game industry in part would do well
> to find ways to place the creation process in the hands of individuals
> or small groups and let lots of "bands" (dev groups) create their
> "albums" (small MOGs) and let the playing community decide which are big
> hits.

I agree, but frankly, there's not a whole lot preventing anyone from doing
it. Most of the tools you need have free alternatives. There is plenty of
book and online tutorials and communities for support and knowledge. If
you wanted to create a MOG tomorrow, there's literally nothing THEY have
that YOU don't... except that creating a MOG is about a million times more
complicated than writing music, requires about a hundred times more
effort, and instead of writing lyrics, you're producing five hundred
different animations for a single NPC.

In theory, there's nothing keeping a garage developer from doing what the
game industry does - but the game industry does it all day every day
(literally in many cases) while the garage developer gets, if he's lucky,
weekends and and an hour a night. Personally, I've programmed a total of
three hours over the course of the last three weeks - I'm a really fast
programmer and I use the down time to plan my next steps, so it's probably
worth a full day's work in the industry (probably more, judging from
experience). It will take me about seventeen years to make a single game
at the same level, despite having all the intelligence, tools, knowledge,
and skills needed.

> What if game tools allowed the same?  Instead of having
> an artist draw a tree (one chord), have a function that has the ability
> to draw a million different trees (procedural synthesis = multiple
> chords.)  This example extends to rocks, streams/rivers, mountains,
> etc.  Procedural synthesis is not just bound to natural objects either,
> but could extend into living creatures/beings as well.

Procedural content generation is, I believe, the ONLY way this stuff will
happen, but it is insanely complex on a deeply intellectual level. It'll
happen when there's actually a field of study for PCG and a lot of that
complexity is ironed out into formulas and instructions for others to
follow. I don't think one guy will be able to make any significant
advancement in the PCG arena unless he is really brilliant and perhaps
maybe insane or autistic.

> Expand those procedural functions one level higher to make entire
> forests, mountain ranges, or a clan of orcs, each different looking but
> still all orcs.  I am glad to see the gaming industry exploring more
> into this as I see it as the future.

The one insight into this that I've had that I think makes a difference is
to use a model-view-controller system, where different procedural
decisions are made at different steps toward different purposes. For
instance, the layout of a level would be something the controller creates
specific to a game engine, but the general flow of the puzzle tree and
narrative happen at a more abstract and independant level. So, basically,
the model is like the outline, from which the controller fills in the
blanks to a specific implementation - from a single outline, an infinite
games game be made, from an infinite number of outlines, well, an infinite
number of genres can be made.

-- 
Sean Howard