Saturday, February 1, 2014

"Get Better" dev diary 1, idea and notes

I just got news about an arts grant that I was part of -- and it turns out we now have some funding! Hurray! I'll talk more about those details when the exhibition organizers announce it, but for now, I want to start documenting my process in making this game -- which I am tentatively calling "Get Better" as a direct challenge to the rhetoric of the mainstream gay-industrial complex.

(Well, originally it was called "Ludonarrative Disco-dance." But that makes it sound too much like a game about games, and this game isn't primarily about games.)

In terms of actual prototyping and production, I'll probably be building on top of my existing first person framework, but I haven't actually done anything yet. Mostly, I've been sending e-mails to possible collaborators and contractors. (The small chunk of arts grant money is making the asset contracting possible. Yay for having a budget and paying people for their work!)

Instead, I'm trying to sketch out the structure of the game first. So here are my actual game notes, along with some remarks on my notes...

- porting slave of god... 
- too much abstraction to represent my experience?
- why does everyone think Slave of God is about drugs?

Slave of God is a highly recommended Stephen Lavelle game, set in a dance club. I think it's a strong work because it's fairly specific about the premise and the perspective (you are at an intense dance club; you love someone on the dance floor) but somewhat vague about certain... variables:

What are you supposed to do? Whatever you feel like.
When does the game end? Whenever you feel like it, you can leave "early" if you want.
Who is your romantic interest? You decide. They're abstracted stick figures.
Is it intense in a good way or a bad way? You decide, you can leave "early" if you want.

At the same time, the problem with letting players "tell their own story" is that their stories often suck. It is difficult for anyone to improvise stories on the spot! This has led to some descriptions of Slave of God that strike me as lazy... like, to me, Slave of God isn't about drugs, because there aren't any interactions or parts about using drugs, and so any imagined link to drugs is obviously a metaphor for some other feeling. And yet, to so many people, that's what they take away -- "it's an acid trip simulator" -- which is an incredibly basic reading at best, and represents a profound lack of imagination at worst. (To a certain extent, I suspect that many who think Slave of God is about taking drugs and dancing -- they've probably never actually taken drugs and danced.)

My work is usually pretty representational stuff on the surface, and that's the space I like to work with. I want to be able to define more variables for players. Still, Slave of God suggests that I should try giving up control in certain aspects, and maybe I will.

- look buttons: compare screenpoint coordinate with guiRects
- John Malkovich-style piloting / cockpit

The first bullet has to do with implementation and coding. When I'm conceptualizing projects, I usually think of "how to make it" at the same time too. It helps me narrow down possible ideas while suggesting new avenues. Here, I'm thinking about how to implement and present player interactions.

The traditional way of detecting what a player is looking at, in 3D space -- is a raycast, which you can think of as an invisible instantaneous laser. If we shoot a laser from the center of the camera, in the direction the camera is looking, then we can see if the laser hits something. If it hits a wall, then the player was likely looking at the wall? This is great for world-based interactions, but it feels hacky for GUI / HUD-style interactions. Instead, I think I'm just going to use old fashioned math to compare pixel coordinates? We'll see if that ends up working.

The second bullet has to do with the user interface. This is going to be a VR game where you walk places. However, I think a lot of VR walking games fail because of their movement interfaces. (This is why there are a lot of vehicle simulation games now; they work best within the real-life context of sitting in a chair, facing one way.) What if we just framed the interaction in a different way? I'm thinking something like the vignetting in Being John Malkovich, which situated the actor inner monologues as these sort of voyeur-pilots inside someone's head. We'll see if that ends up working.

verbs (always):
- turn, LOOK LEFT / RIGHT... turn left / right
- walk, LOOK MIDDLE / UP > STEERING MODE, walk forward, look in middle
- dance, LOOK MIDDLE / DOWN > HEAD BOB... head-bob in bar, or head-bob on the floor
- phone, LOOK DOWN... 

verbs (contextual):
- (bar) order a drink
- (bathroom) wait in line, pee, wash hands?

In "Poetics", Aristotle says something to the effect of, "character is action." It's a very screenwriter-y sentiment; that a good movie involves characters doing things to demonstrate their character. It helps to think about interaction very formally and very literally sometimes -- like, what is this thing that the player must so do with their hand, and how does that input map to this action -- and what are the roleplaying consequences, how might the player feel when doing these things? Also, the range of possible actions suggests something about the player's character too.

I'm very interested in a "head bob" or "nod" as an input mechanism in VR headset games. I've done a few tests as to how to detect it... there are optimal ways for people to nod comfortably / rhythmically. The harder part here will be to figure out how to prompt the player to headbang, I think.

headbob to begin

Since headbanging will be the least intuitive action here, it's effective as a "start button" to confirm intent to start the game, while prompting the player to think about it as a specific input to train and remember.

I'm also thinking heavily about the context for presenting this game: in a public exhibition space. Whenever I've gone to these spaces, I've usually preferred the works that "give me something" quickly. Difficult work deserves exhibition too, but I've always felt like I've been "hogging" the kiosk if I play for more than a few minutes. I want a game that's fairly short but is fun to watch someone else interact with...

black screen. shower running. 
fade in... in the shower, can look down at your body if you want.
start hearing fan sounds / no shower, cut to bathroom mirror
at bathroom mirror, taking a naked selfie >> your face will always be blocked / blurry
start zoomed out, gradually zoom in, phone lerps to player look average
SNAP! this is now your app icon
start hearing music... then cut to the club

boom boom boom boom

This will be the beginning of the game. I write my narrative first person games with a bit of a screenplay sensibility, as to "where the camera is moving" and how the camera cuts. I think a lot about fades and cuts and timing those right. Coding and implementation notes merge with more conceptual notes.

I want this first scene to do a few things: (a) feature nudity, (b) emphasize bodies, (c) give it a gay cis-male context, (d) hearken to contemporary culture with the selfie, and dramatize it a bit as an expressive device, which was one of the few things I liked about Grand Theft Auto 5.

Bathrooms are for character building; they're usually these private, intimate things. In bathrooms, we make ourselves vulnerable. I think that's why so many selfies happen in bathrooms -- mirrors are useful, sure -- but it's also this private-public moment where you can be proud of your appearance a little.

relatively empty bar, only a few people around, you can roam around
after a little while, more people will start coming in

people dance in randomized groups / clusters
people's faces are their app icon?
if you stare too long, you will creep them out for a while, they'll turn away and/or leave
if you approach people, they will either turn away from you or turn toward you (randomized at start)
people who'd turn to you = look at you frequently, but look away if you look at them, etc. >>> game about looking? looking as flirting? looking is expressive?

if you dance with someone for a bit, they will start following you?

you can leave at any time, whether with someone or not

My first impulse was to make this a strategy tactics game. Then I started designing which actions would hook into which actions, and realized I was boring myself. A "mechanic" is only one way to think about action in games, and assumes all interaction should be viewed through this narrow lens of optimal play and tactics -- when optimal play is rarely the point. If you play golf against your boss, or you're playing basketball or Monopoly or poker with someone you have a crush on, you should NOT play too well nor too poorly, because it's not about who wins. Sportsmanship is our blanket term for all this rich expressive context which gets lost if you distill a game into a spreadsheet.

Designed systems don't actually have to be all that complicated in order for people to read complexity into them.

Still, much of the joy of interacting with a system is seeing it respond to your inputs. So I think for this game, a simple and responsive system is best. I'm perfectly okay with people figuring out the "black box" and consuming this game in 5 minutes... all I want is 5 minutes of your time. Can you imagine the profound hell that would be every game demanding dozens / hundreds of hours of your life?

cut to shower, can look down at your body if you want.
if you left with someone following, the shower tile is a different color?
cut to black. go to attract mode.

To me, this is the only ending that would make sense.

... More updates to come later, hopefully! We'll see how this goes.