Adam Saltsman, a.k.a. Adam Atomic, is the indie game developer behind the bare bones game Canabalt, among others. Canabalt stripped down gameplay to it's most basic element, 1 single button press to jump. Yet, in doing so, proved that by concentrating on making the most simple of gameplay fun, the rest of will fall into place.
Adam was nice enough to exchange a few emails with me. We talked about the minimalist approach to game design and whether the big developers could learn from it as well, the social network integration trend in gaming today, and the assumed story in Canabalt.
Alex: Did you go into development on Canabalt with the intention of making it a minimalist sort of game design?
Adam: Totally, it was built for the Experimental Gameplay Project, a monthly game design showcase about small games with specific themes. The August theme was "minimalism"!
Alex: Do you think the big game companies could learn a thing or two about getting their gameplay down to the core, most fun mechanics possible, instead of just adding more and more stuff?
Adam: This is going to sound elitist and grumpy, and I swear it isn't; I'm actually talking about this at GDC with some friends too. AAA companies can't design. I don't mean that they choose not to, or that they are not capable of it, but they just can't. Designing something new and then filling it with their standards for content and length would take...well, look how long it takes Valve and Blizzard to make a game (and they're not even innovating that much!) It just makes zero financial sense for them to invent new things. Do I think big companies should have a not-for-profit sector of their company, subsidized by their sequels, that helps to pioneer amazing new stuff (I AM LOOKING AT YOU NINTENDO)? If my little bitty company can do that, surely they can.
Alex: Social network integration was a big part of what made Canabalt popularity spread, but other games, like Uncharted 2, have tried and failed to include similar features without annoying people. Do you see social network integration growing and evolving with games, or fading away?
Adam: Growing, but awkwardly, haha. These things are not a fad, and for small developers especially they are really important. Figuring out how to use them without bothering your friends is gonna matter more and more, I think.
Alex: Your integration, bragging rights, was a lot less annoying then, "So-and-so played blah-blah last night." But where I'd like to see it go, is a Majestic sort of mystery game, that drops clues on Twitter or Facebook once someone figures out a specific part of the game. Make it happen!
Adam: I like it! We're looking at doing simple ASCII art for our new game. There are a lot of ways to use this stuff!
Alex: Maybe I'm crazy, but I played enough Canabalt that I became convinced there's more at work than just randomely generating building. Am I insane, or is there more at work here? Seems like when I'm at full speed, long gaps are commonplace.
Adam: You're not crazy! About halfway through development I changed it to generate buildings based on your velocity, not your distance into (or out of?) the city. This was pretty critical, and I have Farbs (Captain Forever) and Steve Swink (Shadow Physics) to thank for their input there!
Alex: You're dropping hints! He's not escaping the city, he's going into it to save someone, isn't he?! It's safe to say there's no end, right? I have a friend who got to 17,000 meters, so I'm going to assume there isn't.
Adam: It never ends. Well it might but it's at like 1.78-e36 or something! Lousy 32-bit numbers.
Alex: There seems to be a loose story to the game that is completely left to the player to infer. Are the robots in there to make the game more aesthetically pleasing, or were you intending the player's imagination to paint a story?
Adam: Very much both! Pretty much everything you put in your game should be trying to do both. Anything you put in there should be able to help communicate an idea that will stick around after they've finished playing.
Alex: Since you've managed to boil down your game to a single button, is it safe to say you're moving on from Canabalt, are you going to continue experimenting with a single button control formula, or maybe even start to add more to it?
Adam: Definitely onward, not necessarily upward? I'm working on buckets of stuff right now, none of which is even in the same genre much less similar control scheme or look or feel. There will be a little bit of Canabalt in everything I make from now on though, I learned a lot while building it!