Saturday, September 21, 2013

The Stitching Process

Wouldn't it be nice if all games formed around some strong, clear, inspired thematic core -- a Vision, with a capital V! -- and the great ideas kept sprouting from that seed. I think there's a misconception among some game players that games are made this way. I think they rarely are, not even the good ones.

Game development feels more to me like an archeological dig where all you've got to go on is a hunch. You start digging and maybe you find something, but you probably won't, not at first. You start to second-guess what it is you're even looking for. Maybe you find something exciting along the way, but it's only exciting to you, and you're part of a team looking for something else. Maybe you're hell bent on finding this one specific thing to the detriment of other valuable discoveries along the way. In time, what you find down there probably isn't as perfect as the image you had in your head, but it's something.

Games, I think, are often put together from a bunch of dusty old pieces, not all of which fit. You probably find too many pieces in your search. Apart from all the skill and craft involved in unearthing the pieces without ruining them, there is art in identifying which pieces are the ones that can form a whole, and in connecting them with care.

Part of the reason an archaeological analogy works for me is I don't believe in new ideas. Everything has been done, done well, better than I (at least) could ever do it. Thankfully, though, memories are short and tastes change with time. Synthesize some good old ideas at the right time and people call you original. The paradox is that in uncovering old things, you discover new things.

When I work on a game I always think I know what I'm looking for. Accepting that what I'm looking for may not be the right thing, may not be what we find, may not even be what we end up looking for, is a nonstop process.

In time, if you're lucky, you get to the point where you've unearthed most of the pieces and what remains is putting them together. Finally, you're getting somewhere. The work is cut out and just needs to get done. But it's here, relatively late in development, when some of the most important creative decisions get made, through synthesis, through omission, and through connection.

To convolute my archeological analogy: You're not just unearthing something, you're not just piecing it together. You're bringing it to life. It's a Frankenstein's Monster type of process. So then, once the pieces are splayed out, there's still a need for some good stitching to tie it all together into something recognizable as a distinct entity, preferably one that isn't an abomination but rather something that seems like it ought to belong in this world.

I find that writing provides one of the lowest-cost, highest-efficiency forms of stitching together disparate game elements.

In a simple mathematical model, you can form a line between any two points in space. In real-world situations, writing I think is the key to linking any two otherwise-unrelated things. I practiced this in college. Taking English lit classes, I faced the same challenge as every other student: What could I possibly say about this book that my professor (rather, the teacher's assistant) hadn't heard before a thousand times? A solution I eventually discovered consists of three steps:

  1. Open assigned reading to one page at random, and note the first passage that catches your eye.
  2. Open assigned reading to another page at random, and note the next such passage.
  3. Write essay arguing there exists a meaningful connection between those two passages.

I tried this several times as an experiment when I was concerned that I had no original ideas about what to write. I felt cynical and like a total hack when I did it, all the more so when the work inevitably yielded better grades than essays I put a lot more thought and effort into. But then I thought about it more.

Meaning does not inherently exist. Meaning is manufactured, produced, contrived, any number of ugly words. There are nicer words. Meaning is felt, experienced, inferred. But deep down at the heart of it that shit is made up. Fiction. Narrative. In the case of media, the meaning of a work may initially be created by the author, but in the end it is shaped by the audience. As audiences, we have a strong capacity to sense when a work is cohesive and when a meaning can be found in it, and we like to experience that discovery through passive participation in the work if not through active exploration. Our brains do their pattern-matching thing and we take pleasure from finding hidden connections, from realizing the greater scheme of the work, the grander more intricate design. Meaningful work is impressive work.

So to tie this back to game development: To me it is often not sufficient for a game mechanic to be fun or interesting in itself. It needs to be meaningful somehow -- I want to be able to find meaning in it, to connect it to some aspect of my own life experience (or someone else's), to see why it exists as an abstraction of some deeper truth, and to really understand why it's part of the game I'm playing. And so, when dealing with a bunch of disparate mechanics or ideas, all of which are no doubt cool yet not all of which are cohesively connected or inherently meaningful, I believe there's still a chance to integrate them meaningfully -- to stitch them together -- using words, if only as a starting point.

To give you an example from Bastion, that game has a system called the Shrine, in which you can modify the game difficulty to suit your preference. This system was not the Shrine until rather late in development (for a long time it was the Bestiary) but the underlying idea was always that we wanted a difficulty system that didn't put a blind choice before the player at the start of the game, and instead let you tune the game's challenge in a granular way, beyond the opaque and judgmental easy/normal/hard choice given by many games. So, after much wracking of minds, from a writing perspective we found that a promising angle on it was to structure it around religion. After all, the difficulty effects were global, somewhat magical, and somewhat strange. To engage with our difficulty system was to test one's faith in one's abilities, so a religious metaphor just made sense. What's more, it served to deepen an aspect of the world's backstory (involving a cultural dispute between neighboring nations), which in turn deepened the plot.

I think everyone on the team was happy with how this system came together, and it's one of the best specific examples we have to show for when our collaborative process goes well. More personally, I like what the Shrine system did for the game, and not just mechanically. I like how we were able to make it fit. And I appreciate that, if not for the Shrine system, Bastion would not have broached the subject of religion at all, which I know was a meaningful part of the game for some people (at least some of whom have written to me or talked to me about it). The desire to include a difficulty system created an opportunity to instill meaning into the system, by stitching it onto the rest of the game.

* * *

You're probably familiar with the term "to retcon". It's a pejorative term, frowned on by audiences, referring to the act of retroactively creating continuity where none existed. We see retconning as a sort of public embarrassment, a collapse of the fourth fall, as with the infamous midichlorians example from the Star Wars prequel. However, much like how I came to accept that my college English tactics were effective, I've come to accept retconning as an important part of game development. As much as I'd love for all meaning in a work to flourish from some pure and artful place, from some never-changing point of origin, I'll take it where I can get it. That means always searching for opportunities to create continuity where none exists, or strengthen continuity where a hint of it is detected if only by coincidence.

Ideally all this is done behind a curtain, so that when the finished work is presented, it feels like it has Vision. There is no hint of retconning from the audience's point of view, and indeed, no retconning could have possibly occurred with Vision such as this! So, I'm all about retconning, because I want everything to fit and that's only possible once all the pieces are there, late in development. Retconning is a way of executing on ideas, and early on in a project when all you have are ideas, they can be more frustrating than anything else. All throughout development I like looking for those vestigial little pieces of game, whether it's a conspicuous art asset or a design contrivance or a limitation of the engine or a story trope or anything really, and I'll try to connect them to the game's center of gravity, so that in the end, the finished work gives the impression that it knew what it wanted to be all along.