We talk with the man behind endless runner Canabalt about his desire to have a positive impact on people's lives.
Adam Saltsman has run a full marathon. That's 42,195 meters. In Canabalt, the 2009 endless runner that brought Saltsman to independent game design prominence, such a distance is considered an impressive achievement that would eclipse all the current scores on the game's leaderboard. The furthest Saltsman has run in his own game? 12,000 meters.
"I can actually run further in real life than in a game, which seems kind of backwards," he says over lunch during this year's Game Developers Conference in San Francisco.
I meet up with Saltsman a few hours before he is due to give a talk about indie game development. Dressed in flip-flops and a hoodie, the Austin-based developer is laid-back, animated, and pretty chuffed about having completed a marathon.
"This was my first one," he says, proud that he conquered such a distance in five hours and 20 minutes (in Canabalt, a player can, on average, run 1,000 meters in 22.5 seconds; this means that a Canabalt marathon could be run in 15 minutes 45 seconds).
This was the first one...
"This was the first one and I feel like my main thought at mile 23 or something was if you wanted to extract information from an enemy combatant, this would be the activity that you would use to do that," he says. "Make them run 23 miles and then, when they think they have two more to go, you stand there and you go: 'You know, you could stop right now; you can stop any time; you just have to tell us what you know'. I feel like that would be the perfect activity."
Saltsman was clearly in a dark headspace by the end of the race. For five hours and 20 minutes his mind wrestled with his body to just keep going. Saltsman is by no means an unfit person, but by mile 23, the pain of the race was so palpable he could feel his heart beating in his teeth.
"It was a lot of new experiences, like your legs cramping up and you think you can't run but you actually can - it just hurts a lot," he says. "You're already in so much pain that it doesn't actually make that big a difference. A lot of people say marathons are all up in your head, that it's not a physical thing. I used to listen to them and think 'Pfft, you're running more than 20 miles!', but they were right: it's completely in your head."
Everyone hits their limit during a marathon, and most people still manage to finish. Saltsman managed to finish, even if he was crossing the finish line with grandparents.
Saltsman knows a lot about running. He and his team of indie developers at Semi Secret Software built their reputation on a Flash and mobile game that was all about running. A little man dressed like Michael Jackson during his sequinned black shirt era (with fresh white socks and dainty black shoes) crashes out of a window - Prince of Persia 2 style - sending shards of glass flying in all directions. As soon as the man's feet touch the ground, he runs and he runs and he runs. Sometimes, he jumps. Mostly, he runs.
Give Saltsman a distance, and he can translate that distance into pixels. He can map out the speeds of moving objects in Canabalt: the average player travels at 400 pixels per second; a fast player can get as high as (and potentially eclipse) 800 pixels per second. Calculated by the sprite size, 20 in-game pixels equates to six real-world meters. Saltsman says you could calculate how long it would take to break the Game Center leaderboard settings. "It would be like the heat death of the sun," he says, pausing as though to do the calculation in his head. He decides not to.
There's no shortage of endless running games online and in app stores. Canabalt didn't invent the genre, nor was it the first in Apple's app store, but it managed to hit a certain nerve that lured players and kept them coming back for more. The iOS version of the game was an instant success, but it was the Flash version of the game that drew players by the millions.
"The Flash success freaked me out really bad," says Saltsman, who still seems surprised by the number of people who have played his game. "I'd done Flash games before that people liked and usually that meant that after the first few months maybe a 100,000 people had played the game, and that's really exciting.
"But Canabalt was like a few million people in the first month and a few million people the next month and that was just completely unexpected and weird. Maybe there are people who expect an audience that big, but I don't. I still don't. I still don't expect the next game I make will have 10 million or 20 million players. That's not an expectation at all."
There are many reasons why players may have chosenCanabalt over other endless running games: there's the tasteful, minimalist aesthetic, the realistic and visceral soundscape, and the sense of speed as the little man pelts past a flock of doves with a satisfying 'whoosh!'. Or maybe players wanted the same thing as Saltsman: a dash of nostalgia, a hint of fanboyism, and a soothing balm for all the running games that disappointed them.
The nostalgia and fanboyism comes from three games: Super Mario Bros., Out of this World, and Prince of Persia. Saltsman drew inspiration from Super Mario Bros. speed runs, where players complete the entire game in less than five minutes. Two Nintendos and television sets are placed side-by-side as players find the shortest path to complete the game: they never stop moving; they never turn left. It's constant, top-speed, left-to-right motion. Two minutes in, they're blazing, dodging hazards with aplomb; it feels like the fastest game of Super Mario Bros. ever played, because it is. Whoosh!
Saltsman also drew inspiration from disappointments, namely Electronic Arts' 2008 dystopic parkour-esque game, Mirror's Edge.
"I was really disappointed by Mirror's Edge when it came out," he says. "I had really high expectations for it and there were a lot of aspects to the design that I felt were undercooked; it felt like it was really rushed. You're on rooftops. You're running and doing acrobatic things. But it's clumsy and unnecessarily hard. It's like the polar opposite of Bennett Foddy's games (QWOP, GIRP, Pole Riders).
Saltsman says that in Bennett Foddy's games, there's nothing between the player and that feeling of rocketing at extremely uncomfortable speeds right away. The games are hard, but only if players expect running 100 meters like an Olympic athlete (QWOP) and rock climbing off a cliff face while being attacked by seagulls (GIRP) to be easy. Players can get a sense of the weight of their flopsy bodies and the potential speeds they can achieve right away - the games are as tricky as they need to be. In Mirror's Edge, Saltsman feels that many of the constraints are unnecessary and dull the experience.
"It's really hard to go fast in Mirror's Edge," he says. "You're in this beautiful city - why would they give you this big, beautiful amazing place to be in if the only way I can be in it is really clumsy most of the time, until I've played it over and over and over and I've memorized one specific path through and then I can get a little taste of that?"
Driven by dissatisfaction, Saltsman wanted his game to feel exhilarating. He wanted running on rooftops to have a sense of flow. The idea of flying across rooftops shouldn't be work; it should be a dream-like feeling of speed and weightlessness, the opposite of a gruelling marathon. Hence: Canabalt.
Creating speed that isn't there
Movement and speed in video games are tricks shown to our eyeballs. The arrangement and rearrangement of pixels on a screen tells us where we are, how fast we're going, and where we'll end up. So when a developer has only a tiny screen at their disposal, how are they to make us feel like we're flying when we're really wedged between sweating commuters on a city train while staring at little iPhone screens?
"There were a couple of things that were fairly intentional: one was the form factor, which was very specific," Saltsman says. "It's 3:1, which is wider than a cinema form factor, and it's to solve a problem that games about speed encounter.
"With Canabalt, I wanted it to feel like you're running really, really fast, which sounds completely simple, right? Velocity on the X axis equals all of it, and off he goes - he goes really fast! But speed places certain demands on the player and the controls. One of the main things is how far you can see in front of you, because the faster you go, the less time you have to react to objects. You want to be able to see well in front of you, but there's a presentation problem. There's a visual aesthetic problem."
"I WANTED IT TO FEEL LIKE YOU'RE RUNNING REALLY, REALLY FAST, WHICH SOUNDS COMPLETELY SIMPLE, RIGHT? ... BUT SPEED PLACES CERTAIN DEMANDS ON THE PLAYER AND THE CONTROLS."
Saltsman gives an example of travelling in a plane. When a person is up in the air they can see for miles in all directions. They may be travelling at 400 miles an hour, but when they look down to the earth they appear to be completely still because their frame of reference is too large. If the same person on the plane was armed with a telescope and asked to zoom in on the ground to narrow their frame of reference, the ground would rip by and be invisibly blurry. There would be some serious whooshing.
"It has nothing to do with the actual speed they're moving at," he says. "It has everything to do with the frame of reference for what you can see around you while you're moving, and the general rule is you have a tiny frame of reference for speed and you need a huge frame of reference for reaction time.
"So what do you do? What a lot of games do is put the camera behind the player. Canabalt is almost a driving game from the side. In F-Zero they have the camera behind the player and that way you can see really far ahead, but it seems like you're going really fast because the side of the track has all these little objects on it that rip by in the foreground and you get the sense of speed and you can see far enough ahead. You can't do it in 2D without doing some kind of actually considered compression of space."
Hence the 3:1 cinematic ratio. This particular ratio keeps the frame of reference really small on one axis so that players can get a sense of speed, but the screen is wide enough that players have anywhere between a quarter to three quarters of a second to react.
"I really wanted to have a rich sound texture and this really blazing sense of speed. For me it was about accomplishing that and also having it be mechanically sound and fun to play. I felt really good about accomplishing those things."
The whole experience is just torture. Saltsman has run half marathons before, and he would go so far as to call them enjoyable. But this - this darn marathon - is something else entirely. As soon as he passes the 14 mile mark, he begins teaching his brain a weird sort of math. He convinces himself that for every mile he runs, he has actually run two miles: one because he now has one less mile to run, and one because he has just run a mile. This, to the mind of a marathon man, equates to two miles. It makes perfect sense.
"I was just fighting myself for hours," Saltsman says. "The last 11 miles might take two or three hours to run and it was exhausting, more than physically exhausting. I've never fought myself for two hours straight and it was a weird experience."
"I feel like I have more perspective now," he says. "I occasionally have what I would call a transcendental experience that reminds me that there are a lot of things that video games are not good at doing. Going to the Grand Canyon a few years ago was an experience like that - it was a really overwhelming place to be. I sort of think about all these things that games are really bad at, not in a cynical 'people need to get outdoors and stop playing video games' way, but more like, there are things that I feel have inspired art for thousands of years and the sublime majesty of nature is one of those things, and human trials are another one of those things, and this marathon was very much a human trial experience and it gave me the perspective.
"I realized the way we talk about hard games, challenging games, or masochistic games is exaggerated. I've never played a game for two hours and mentally fought myself to exhaustion only to just barely finish it and feel like my whole life ahead of me is different now. That's not an experience you get in a video game, and the question for me isn't whether or not you should have that experience in a video game but whether or not you could. To me it's a really interesting question."
People often speak of having transcendental music experiences or transcendental architectural experiences: as the ultimate mash-up of so many existing media, Saltsman wonders if there could come a day when video games will be able to provide such experiences.
Passage, a game about living a life by walking from left to right
For Saltsman, it's not that games will matter once they become transcendental - they already matter, no question about it. But he's curious about whether games can do for him what a five and a half hour marathon did.
"If you made a game that was as painful to play as a marathon, I think you would be put in jail for trying to sell that, but that's what we did," he says. "We paid $50 to enter this marathon and we had this five hour experience and it was horrifying and transcendent at the same time, and a lot of video games have a lot of weird similarities to that but they don't go as far. The effects don't last as long."
He cites Jason Rohrer's Passage as an example of a game that people have purportedly had transcendental experiences playing: perhaps they've had a shift in their outlook on things like morality, which is cheesy, but not uncommon.
"I don't know if you could design a transcendent experience without having a bunch of them," he says. "I'm kind of curious about that. I mean you can design a fighting game even if you don't get into street fights all the time, I guess, but being able to recognize what that is and the forces that are going on is really important. You could almost argue that fun is a temporary transcendence and there are all these other experiences that have a permanent transcendental effect, but it really has to do with fun."
Ten thousand kicks
Saltsman still thinks about his marathon. After the roaring success of Canabalt, he went on to create the official Hunger Games mobile game in collaboration with other independent developers and is now working on Semi Secret's elusively titled game,Hundreds. He still wonders, though, if the transcendental experience of running 42,195 meters while octogenarians in the same race cheered him on can somehow be captured - even in the tiniest of capacities - in a video game. The experience wouldn't have to be as brutal as running a marathon, but the challenge and the sense of mastery and accomplishment should somehow better the lives of the people who play it.
"Bruce Lee said that the trick to doing the perfect kick is to do 10,000 kicks and there's this idea that there's something about that number 10,000. If you do something for 10,000 hours, you'll have most of it figured out. I think Bennett Foddy's games have a bit of that," Saltsman says.
"IF YOU DO SOMETHING FOR 10,000 HOURS, YOU'LL HAVE MOST OF IT FIGURED OUT."
"A game like GIRP says 'No, you don't just go right up to the top; you have to do the challenge and the problem solving part'. I actually rock climb as a hobby, and while GIRP is not exactly the mental and physical process of rock climbing, if you've ever tried Mega GIRP - which is GIRP but instead of playing with a keyboard you have four dance mats taped together and the keys are on each of the dance squares - the climbing becomes very physical and requires a fair amount of stamina.
"You get kind of Twistered out so you have two legs over here and you have one hand back here and you're trying to tap this other thing with your other hand and there's no in-game timer; there's just the timer of your body. How long can you maintain this absurd position physically as a human in order to progress? I think his games kind of tiptoe to some of that difficulty. There's something in the way he uses humor really well. He uses happy accidents really well. There's a bit of a gambling quality involved - how many complex psychological processes that he wields with apparent ease in order to get people to enjoy it."
Saltsman doesn't know if or when he will be able to design a game that gives players a transcendental experience, or even a fraction of the challenge, satisfaction, and mastery that Bennett Foddy captures in his games. He says it's not a matter of sitting down on a Monday and deciding that he's going to make a game that changes people's lives. It is not an active goal, although it does occupy a snuggly spot in his mind.
"For me, my philosophy of art and craft is just to make things that make people's lives better, and not necessarily in a 'we didn't have a well in our village before and now we have water' way, but in a 'they found something in there that was valuable to them, even if it's temporary' way," he says.
"So a transcendental experience in the form of a video game or board game, something somehow, would totally fall into that for a lot of people. I think they would have this weird experience that somehow they didn't expect or change their outlook on things. If I can do that, that completely satisfies my desire to have a positive impact on a lot of people's lives."
Until then, he'll always have a marathon.
Adam Saltsman, Terri and Tom Saltsman, Semi Secret Software, Nintendo, Bennett Foddy, Electronic Arts