Tuesday, February 12, 2008

A Storm of Dissection...

There are a lot of theories out there as to how to talk about games. Most people seem to want a formal vocabulary for discussing the aspects of game design.

I disagree strongly. A formal language? Formalizing what? We don't know anything, so how can we formalize it?

We're inventing all these languages that divide up a game into such tasty-sounding pieces, but based on what? To what end?

For example, I like MDA - mechanics, dynamics, and aesthetics as posed Marc LeBlanc. It's not the first thing he's posed, it probably won't be the last, and it's a fun brain exercise.

But people are taking it to heart, treating it like it's a formal, scientific approach.

It's just a general way of thinking about games.

This happens a lot: someone posts something that has categories and subdivisions and people say, "oh! It's internally consistent! It must be right!"

They are arguably formal, but don't mistake "formal" for "right". Astrology is also formal.

Thought exercises are grand things. They're just about all I do. But don't mistake them for solutions. We don't know the first thing about games: how could we accurately represent them?

Someday, thought experiments like these may lead us to stronger knowledge. Maybe, someday, we'll be able to have a formal grammar that isn't just voodoo dressed in a tux.

But not today.


Matthew Rundle said...

I think that this kind of model is useful because it advances the cause of knowing what we're talking about.

If people are advancing new ways to think about games, then even (or especially) the broken or incomplete models are useful, if their strengths and flaws can be identified and further models are drawn off of those conclusions.

Basically, I think that there are enough people thinking about games regularly that popularizing and canonising theories isn't a bad thing - even if only because it incites argument about the specifics that could spur further theory.

Ryan said...


Couldn't have put it better myself. People tend to lose sight of the fact that these 'formal' frameworks are just humans ham fisted attempts to impose some kind of order on things because generally, as a race, we hate mucking about with all that nasty uncertainty.

They have their uses, facilitating communication between people on the topic, or simplifying complex ideas down to a simple label. Sometimes they work the other way and needlessly complicate a simple concept. But those who think they are the one 'right' way are the same as those who would have stoned Copernicus for saying the earth orbits the sun and not the other way around.

Craig Perko said...

Sure, I obviously think they have value or I wouldn't use them myself. But they aren't accurate, and that needs to be remembered. Consider them theory fan-fic rather than theory canon, yes?

Chris DeLeon said...

The same is true for all models we use in physics and engineering, or any non-abstract field of study (math, symbolic logic), right? As in, they are at best useful approximations of relationships that we cannot fully describe.

Except, even those fields are grounded in something (theoretically) objective and concrete; to this extent, I believe that these ways of framing the game design problem space have more in common with literary styles, or art/design movements, rather than something ultimately answerable. As I see it, different purposes are better served by different semantics, from both development and critical perspectives, but beyond that I do not see that speaking of accuracy or correctness is meaningful beyond someone's subjective values.

What do you suppose will have to happen between now and someday to render these languages and frameworks more useful? Do you foresee a definable end goal for such theories, or an objective way that one proposal may be judged as more 'correct' than another? (I am genuinely curious for a response; these are not intended as rhetorical questions!)

Craig Perko said...

There is currently no realistic way to test if a given theory is right or wrong. I'm pretty sure they're all wrong, since we're basing them on wild guesses.

I think that in order to test these theories, you would need to form some kind of... group. An engine that could be tweaked in tiny amounts, tweaked into lots of different formations by people seeking to test theories. Then, of course, the games would have to be tested, and information recorded (not simply "oh, it looked like he had fun...").

I'm not sure I care to reach that future. I'm pretty happy just struggling in the mud. But I get less happy when people mistake the mud for solid ground and build castles on it.

Chris DeLeon said...

Thanks for the prompt reply!

Please pardon the verbosity; this is partly me 'thinking aloud', and partly an attempt to hone in on a very specific point from a few different directions at once. It isn't intended as a rant or disagreement, a disclaimer I'm inclined to provide since my lengthier writing has a way of coming out more preachy than I intend.

I'm still questioning the deeper assumption. "There is currently no realistic way to test if a given theory is right or wrong" - what do you mean by "right" or "wrong" in these uses?

Fun to whom? Of value to what end? Fun the first time, fun again and again, fun for hours, fun only for children? What does fun mean in this context, why is fun the implicit goal... questions that I think hint that the concepts of "right" and "wrong" are a bit slippery in their embedded assumptions of objective valuation.

As I understand it, MDA doesn't really try to address these types of things; it and other FADT's (a now 8 year old concept, courtesy of Doug Church) mostly seek to assign words to patterns, giving people handles to refer to formulas or patterns to be improved upon.

If MDA proposed, "spend twice as much time on Aesthetics as Mechanics," I'd be appalled. But by instead providing a few new vocabulary words to refer to bits and pieces that otherwise are difficult to separate in conversation, it seems a suitable tool for speaking about certain aspects of a project.

Saying, "I don't like this sequence" isn't as helpful of a building block in refining videogame literacy as, "I don't like this sequence, and I think that's because it lacks perceivable consequence." The latter can be argued, refined, redefined, or replaced; the former doesn't even provide mud to build upon.

Often, I think that these devices are more useful in criticism and classification than synthesis. Trying to treat them as creative formulas is likely to lead to systems that are pleasant to talk about but otherwise worthless to end users - much like trying to make music based on theories and formulas, which seems to happen for big name commercial groups, but indie musicians tend to shy away from.

It may inevitably find a place in synthesis on larger development teams, however, where communication has to be externalized that can be kept internalized for one-to-few member projects. Being able to cite a FADT gives slightly better grounding than saying, "because I think it'll be fun," and when the argument outcome translates to weeks of expensive art/engineering effort. trusting established patterns is, for better or for worse, typical practice. I've even seen this form of argument play a role in some of the large indie teams I've been on, and it worked.

Designers that can articulate why are more likely to be listened to than designers that can only articulate what, and these attempts at generating a common lexicon or tool set give us a way to do that besides, "It's like the boss fight from level 6 of that game from 1997..."

To that extent, the more terms and perspectives that we have on the subject as handles for what exactly we're trying to say, I think the better off we'll be from having a richer language. Just because someone invented the screwdriver doesn't mean we all have to use it on nails.

But that is why I am confused by the idea that any of these handles can be intrinsically right or wrong, correct or incorrect, accurate or inaccurate; they seem to me as different viewpoints on the same target, a target that can necessarily only be fully described in and as its completed game form.

Craig Perko said...

I'm still not saying they aren't useful! I think they are useful.

What I'm upset with is that they are being treated as a kind of scientific thing, as if talking about things in MDA (or whatever is popular today) resolves something.

These kinds of grammars give you a new perspective and the ability to share that perspective. However, that perspective is not connected to reality. If it happens to align, then yay! If not, you can't be fooled into thinking you got it right because you talked about it right.

Also, why the categorization? Categories are just for ease of communication, and there's something to be said for that. But they certainly don't reflect an underlying truth. There really is no clear line where you can say "this is mechanics, this is dynamics". Some things are clearly on one side or the other, but some things are blurry or messy and are harder to classify.

Classification is just something we do to make talking easier. Reality is not so neat.

Classify, but understand that you're not unveiling hidden truths of game design. You're just talking, just thinking. Thought experiments, not solutions.

Chris DeLeon said...

Amen. :-D