By Derrick Jean-Baptiste

No Man’s Sky is a relatively recent game that has taken the entertainment industry by storm. The game was pitched as a technological advanced action-adventure survival game, which relied on procedurally generated assets in order to develop the universe that surrounds the player.

The Player is able to perform with the entirely of a procedurally generated open universe. One of the common claims of the developers is the fact that players would have the option to survive and explore on over eighteen quintillion planets each with their own flora and fauna.

With these strong aspirations, it was to be expected that the developers could not hit all of the marks. Yet, no one expected for No Man’s Sky to fail on such a grand level.

There has been a lot written about this game. Ranging from the game being sixty dollars even though the game played more like a twenty dollar independent game  to being sued for false advertisements.

I’d like to approach the game from the perspective of looking at the specific technical choices made in the developmental process of the game and focusing on procedural generation.

In all fairness, I think there’s a subset of people that will enjoy it for exactly what it is, which to their credit, is more or less what the devs promised. I also believe that there’s also a subset of people that will find that this game is rather  soothing. It’s a game that you could grind in or simply play while you’re binging a tv series.

On those grounds, it is completely awesome – you could play it twenty-four-seven three hundred and sixty-five days a year and never run out of worlds to explore, species to catalog, atmospheres to descend into and star systems to warp to. And, not to be discounted, the developers promise much more to come.

On a more general level, though, I think much of the lack of thought for the game comes from the exact thing that people laud about it the most. Procedural generation can if used well, greatly lessen the load on artists and engineers in creating spaces for players to inhabit. But the issue is that those spaces are, uniformly, not particularly interesting.

There are always going to be the super fascinating edge cases, and I suspect we’ll see those screenshotted and shared in the weeks to come, but think about it this way: for every Niagara Falls and Grand Canyon, you have almost four million square miles in North America  that are less distinct and less interesting, and in many cases are indistinguishable, acre by acre.

In other words, nature can create unique and notable beauty, but that’s the exception and not the rule.

On the other hand, every major city in the world has a skyline that often includes one unique, notable building that you can identify by silhouette alone – man creates things with intent, and intent drives meaning.

This meaning is often what you search for, and why many people are pointing out that they would rather pay thirty or even sixty dollars for an hour long game over a game like this. There’s meaning in that experience, and that meaning is what drives emotional connection.

Maybe there’s a point in a game that doesn’t treat the player as the center of the universe, a game that doesn’t allow you to complete it, a game where you are merely just a plankton in an endless ocean.

Maybe – but the issue here is less that the narrative doesn’t treat the user as the ultimate savior, but that there’s no narrative at all. You may explore a hundred solar systems, but the only reason, the only drive that you have in doing so is yourself.

But don’t get me wrong there is Nothing wrong with procedural generation, but it only works if it paves the way to emergent gameplay. If Procedural Generation does not serve for emergent gameplay, it gets bland quite fast.

We as a species are really good at recognizing patterns. We quickly get a feel for the process of the Procedural Generation algorithm behind the scenes. When you see a few universes an algorithm generates, it’s like you’ve seen them all. Emergent gameplay is what breaks the patterns.

Developers struck gold with the Survival and Procedural generation genre. Survival by crafting in varying landscapes and conditions gives way to many emergent scenarios.

Minecraft obviously comes to mind. Don’t Starve is another. In these games, the challenges the procedural generation algorithm throws at you are unique challenges. There is a carry over in learned skills from one game session in a particular universe to the next, but it takes a few playthroughs to be confident that you can survive in almost all environments the game throws at you.

Another method for emergent gameplay is, of course, multiplayer. Humans are incredibly complex so put some humans in a universe and give them abilities; they’ll create the emergent experience for all.

My point is that if a procedural generation based game is bland, it rarely is the fault of the decision to make the game procedural generated. The actual game mechanics should give way to emergent scenarios that break the pattern that is our understanding of the game.


Please enter your comment!
Please enter your name here