“This will be terrible,” a bumbling and shy Sean Murray from Hello Games says to an assembled crowd of seated journalists at a Sony-mandated session for No Man’s Sky. He’s not wrong. He launches, umming and erring as he goes in to a history of the company that spawned Joe Danger, and soon, the frightfully ambitious No Man’s Sky.
“Just to be clear,” he reiterates, “this will be a disappointment.”
He explains how the company that’s making one of the most intriguing and ambitious games of this new generation is staffed by just ten people – and how only seven of them are busy with No Man’s Sky. He explains that for the first year of No man’s Sky’s existence, only 4 people were making the game. They all came from a bigger studios, leaving Criterion after years of making the same thing. “I was burned out on Burnout,” Murray says. It’s not the sort of game he and his colleagues wanted to make, so they left to be more creative. This was before developers leaving large studios to become indie developers was even a remotely viable thing.
They wanted to do something different. That something, asserts Murray, is No Man’s Sky. “It’s the game we’ve always wanted to make,” he says, realising that it’s the sort of thing just about every developer says. The guys who make Rainbow Six would say the same thing, and then just make another one the next year. “We mean it this time,” he says.
He speaks like an odd, erratic genius, lost in his own thoughts. He’s pleasant enough to listen to, but nothing he’s said brings me, or anyone else in the audience closer to understanding quite what the hell No Man’s Sky is all about.
What it’s about, Murray says when he stops feeding us history, is emotion. The focus in the media up until now has been all about the game’s technology and tools, but it’s really about the emotion a player would feel when discovering something new for the very first time. It’s about landing on a planet that’s alien to everybody, and finding something new. The game itself, the playable bits of it, are really difficult to summarise, but Murray attempts it.
There’s a galaxy, that everyone shares. Players all start at the very edge of the galaxy. If you were the very first player, you’d be able to pull up a galactic map and see it littered with planets – but each would be listed as unexplored, as nobody’s ever been there. the point of the game, largely is exploration. To do that though, to break free from the starting galaxy, you’ll need interstellar travel, as the basic ships you start out with are incapable of hopping between galaxies.
To get a better ship, you’ll need money, and there are a number of ways to generate coin. Players will likely take on one of a number of roles. You’ll get the weird folk who’ll opt to be resource traders, gathering minerals or resources from planets and selling them along trade routes. They’ll likely stick to known space trading along established routes. Another way to make money would be to become a space pirate, attacking other ships and stealing their loot. All of this eventually allows you to afford upgrades to your ship and equipment, getting you closer to the middle of the universe, where the rewards become greater.
Each of those planets, and everything in them is created through mathematics, procedurally generated by a seed. It’s all maths or some sort of arcane sorcery, fed through a fancy algorithm. What’s rather magical about it is that none of the planets,, or the terrain, or anything about the game is built or designed in the traditional way. It’s all generated on the fly, based on set inputs run through a formula. Nothing generated is stored locally, or on some sort of cloud. A planet you’re on now, for example, will essentially cease to exist once you leave it, but its numerical seed and the game’s algorithm ensure that when somebody else visits it, every tree, rock, and blade of grass is in the same place it was when you were there.
It all sounds like it’s a game of infinite possibilities, but Murray was quick too point out that it’s not quite infinite. As “some guy in the comments” is quick to perpetually point out, Murray says, nothing in computers is ever infinite – but No Man’s Sky comes close. The game, at one point, used a 32 bit integer as its universe seed, which would allow the game to generate up to 2,147,483,647 planets – which would take a single player about 5000 years to see any sort of repetition, if they discovered a new planet every second of every single day. As a bit of an “up yours” to that guy in the comments, they’re now using a 64 bit one, which would take the same player 4 to 5 billion years to see any form of repetition.
It’s maths, and it’s impressive as all hell – but I still have no real idea just what the hell No Man’s Sky is. I do know that it’s beautiful, and insane, and ambitious – but not very much more than that.
Last Updated: August 14, 2014