Dyad is Shawn McGrath's labor of love. It's a single-player racing/shooting/music game that has more in common with the demoscene or audio visualizers than it does with traditional games. Beneath its psychedelic visual exterior, and beyond the amazing interactive score by David Kanega, hardcore mechanics lurk and Ikaruga-style challenges await.
The game, developed by McGrath's studio ][ (a.k.a. Right Square Bracket Left Square Bracket), is a full package of new sights, sounds, and experiences - something you can't really identify without playing for yourself (luckily there is now a demo of the game on PSN).
In this interview with McGrath, we took aim at the game design side, trying to get at the heart of the PlayStation Network's new experimental beast. Along the way, we learned how he has gone all in with this project. If Dyad fails, McGrath fails.
How long were you working on this? Feels like it's been a long time.
Shawn McGrath: Yeah, three and a half, four years, something like that.
How have you managed to maintain sanity and financial solvency?
Sanity, her. [Points to nearby wife.] Financial solvency, her. I got money from the Ontario government to fund part of it, and I worked for my entire life and spent my life savings on it, and my wife works full-time. And sanity, she keeps me sane, too.
What gave you the confidence to leap off and spend all your savings?
Well if I didn't, my life would be really shitty. I was doing games and consulting and tech work and stuff for game companies ... you know, it was shit. No one wants to do that. So I figured out an exit strategy and this is it. Hopefully it does okay.
Is your future success predicated on the sales of Dyad?
Yeah. Entirely. Which is strange because I didn't -- it's not really a commercial game. As you play it it's like, "Yeah, this is not a thing that should be sellable." But Sony digs it and they're behind it and everybody I've shown it to, except for the bros, are super, super into it. So, it's not like I compromised anything so that it would make money. I just went with what I wanted to make and prayed, and I'm still kind of doing that. But it seems to be working out now.
How did you determine pacing of when to layer in the new mechanics that come with each new area?
Well it took three years just to do the first 12 levels and to figure out a pacing structure that people understood. And that was just through an insane amount of play testing. After that I extrapolated what I thought would work from that and made the rest of the game before showing anybody. It worked. The early part was done through play testing and then the later part was done through guesswork.
I personally feel like the earliest bits feel a little bit empty in terms of what I can do and what's happening.
When we originally designed the game, we had all the mechanics that are in the very last level. That was just one level. And that was all in the game for a year and a half, or so. And figuring out how to split that up was the tricky part. The first thing you have to do is establish pairs of enemies, because then you build off of that. So I said, okay, let's just have pairs of enemies alone in a level. And of course, it sucks when you do that because you've designed a fun interaction around, a million really complex little things. If you just take one thing out and stick it into a level by itself it's really bad.
Once that was separated I spent months, a really long time, trying to figure out how to make just that one little thing fun and the end of the game is, just -- toward the end of the game, the last five, six levels or whatever, they're just crazy. They're insane. Shit's happening all the time and if you have never played the game up to that point, there is no way you can comprehend what's going on. So, I thought that a really nice thing to do would be to make the start relaxing and kind of chill, just so you can play and feel things, and you make music with it -- it's somewhat of a musical instrument.
Once the music and the graphics were made so that it was simple and relaxing and chill for the early levels, then it's an exciting thing. It's fun because it's relaxing and it is something on its own. So, each level -- in the main menu when you open the level, there's a submenu that comes up [with three modes,] "game, trophy, remix." Trophy is the trophy mode, a sort of re-contextualization of the level.
Originally it was "game, challenge, play" -- play was the remix and challenge was the trophy or whatever. When I called it "game" I kind of didn't know why I was calling it that. I thought maybe I'd change it, but when Mathew Kumar (level designer on Sound Shapes) saw it, he was like, "Oh, game, that's the best thing you could have possibly called it because each level feels like its own game." And I was like, "All right. That's why we're 'game,' right, because that was a really good explanation and I really like that."
It really feels to me like each level is a completely different game almost. That's why it's kind of okay and cool that the first levels are really chill and relaxed -- you're making music and then the end of it is just a clusterfuck of awesomeness.
How did you work out the UI here? You have some visual information you need to convey like how many lance attacks you have. But if you're really intensely playing it you're pretty much looking at the center. The main area of the screen you're looking at is maybe two inches wide.
There are a couple things that were done. Some levels play different music loops that are very obvious when you have a full bar, so that's good. Every time you graze an enemy it plays certain sounds. When you graze an enemy that gives you a full lance bar it plays a louder, longer melody. So you can get the information from that. The bars that say "lance ready" flash really, really large into the kick drum. So they flash a lot and they actually obfuscate part of the screen, so you can still see them even if you're focused on that one little tiny bit.
And then for the last half of the game, you can't really see the UI. And it's kind of okay -- and this is demonstrable through experience with other people playing it -- you can actually feel when a lance is ready. If you're playing efficiently, you'll always kind of have a lance ready when you need it. So it becomes less about actually looking at the UI and more about just feeling the game and internalizing what's happening. And the level structure is designed in such a way that it encourages that.