Mastodon Icon RSS Icon GitHub Icon LinkedIn Icon RSS Icon

The Challenge of Infinite Space

A.K.A. why Elite: Dangerous, Star Citizen and No Man’s Sky could be doomed to fail

Since the beginning of the videogame era, developers aim to reach a perfect simulation of the world. “So realistic!” is probably one of the most abused comment to a videogame ever. The videogame industry tried to reach realistic graphic, realistic sounds, realistic landscapes and realistic Non-Player Characters. However, reaching the perfect realism is a never-ending quest. As soon as we reach something, desire drives us to seek more.

Now that we have extraordinary graphic capabilities and stunning landscapes, some developers are going to raise the bar to a much bigger task: simulating the universe (or at least, the galaxy). We are not in the first (game) generation who is trying to achieve that (the original Elite is a very old game indeed), but we are for sure in the first generation who can reach this goal; at least to some extent.

We have the technical capabilities for doing amazing things: we can generate and simulate hundreds of billions of star systems, we can allow the player to travel to every one of them, explore and land on every square unit of the planet surface, every asteroid, every satellite. Software/games such Space Engine can clearly show us what a lonely developer can do in its spare time. Imagine what a game developers team can do on their full time job! It is not a surprise that several studios spotted the niche in the market and tried to use this power for a real game. After all, it is not “simulation of experience” one of the meaning of “game”?

The amazing view of a generated landscape in Space Engine 0.9
Figure 1. The amazing view of a generated landscape in Space Engine 0.9

The players, as far as they are concerned, can understand this and are overly excited every time a game promises so much contents. A year ago Elite: Dangerous was released. It was the first, early release, of a group of massive space simulation games born at about the same time. The other two are Star Citizen, who recently reached the astounding fund-rising amount of 100 million dollars, and No Man’s Sky, the hyper hyped game of the new era. These games are quite different from each other, but they share a common trait: they aim to simulate a vibrant and huge world. They explicitly state that their core element is the exploration of a procedurally generate, almost infinite, world. A galaxy or the universe, in particular.

The idea seems amazing. A procedurally generated, practically infinite universes to explore, to live in! Many players share the same feelings of wonder and excitement. Then “Elite: Dangerous” comes out and, after the first enthusiastic experience, the game met mixed reviews. I played with it a bit, so I can confirm the main criticism players described: lack of gameplay. Frontier, the developers of Elite, failed to provide the right amount of gameplay in their game. This was a latent opinion for a long time. Recently, when they released a paid expansion to the game, the player base has become upset and the frustration for the developers started to raise again. I do not know how this chapter of the Frontier studio story will end, but this made me think a lot about this kind of games.

Elite: Dangerous boasts an excellent flying simulation.
Figure 2. Elite: Dangerous boasts an excellent flying simulation.

Did the game “fail” because of Frontier or because of the concept of simulating an infinite world? In other words, is it possible for a finite amount of gameplay to fill an infinite space? Is it possible to provide enough gameplay for a game that aims to emulate the infinity of our space without highlighting the emptiness of the game world (and thus making the game boring)?

After all, any amount of gameplay you dilute into the infinity leads to “zero gameplay density”. No matter how many players you get in the world you are simulating, they will spread like gas in the emptiness of space, becoming millions of lonely players. No matter how much contents you can put in your games from your artists, those contents will repeat infinitely in the infinity of the generated space. At the end of the day, your world becomes just a library composed of copies of the same book but with scrambled words. Some of them could be interesting, original and amazing to discover, but the majority of them will be just uninteresting “random white noise”. Elite: Dangerous, maybe because of its early release, made this clear as the Sun. However, I think many of these issues are not game-specific, but, instead, a general set of issues: the “infinite space problem”.

If this problem is true in general, if it is impossible to fill infinite space with enough gameplay, contents or whatever to make a game fun for everyone, then we can already say than Star Citizen and No Man’s Sky will be doomed to fail in the same way, even if with different times. These two games will be and are actually forced to challenge this problem and bring us a solution. In the meantime, we can explore some ideas.

Minecraft… ehm no… No Man’s Sky… ehm… Let’s say “something in between”
Figure 3. Minecraft… ehm no… No Man’s Sky… ehm… Let’s say “something in between”

The first way to avoid the “infinite space” problem is to make the game interesting in the local scale and then eventually expand it to a greater level. Your game must be fun at the smaller level and eventually go to the infinite. Minecraft is a good example of this. Even if the game world is technically infinite, nobody really cares. The Minecraft gameplay involves crafting, surviving, mining, do low-scale exploration in order to get resources for crafting, surviving and mining.

The second way is more complex. The idea is: if you cannot fill an infinite space with a finite amount of gameplay, then fill it with infinite gameplay. We could talk about procedurally generated gameplay, a new active and interesting research field, but it is not needed. Building an infinite amount of random empty dead planets in a huge galaxy is not enough. You have to use PCG to provide variety in your galaxy, even at the macro - level. However, this requires a lot of contents and ways to use PCG to generate stories, empires, assets in an automated way that may be out of reach for many developers.

The third way instead is the easy one. Do not use the “infinite space”. In fact, infinite space is not needed in order to create a vibrant, living and interesting universe simulation. EVE Online is an example of this. The EVE “galaxy” is just an infinitesimal fraction of the galaxy of Elite, but it is way more full of life and stories! Smaller size allows a more controlled gameplay, force players to interact each other, allow a more complex-but-controllable economy. Really, you don’t need the “infinite space” for every space game.

The EVE Online map is incredibly small compared to the Elite: Dangerous one, but is full of history and lore.
Figure 4. The EVE Online map is incredibly small compared to the Elite: Dangerous one, but is full of history and lore.

These are not the only solutions. No Man’s Sky and Star Citizen will be the first games who need to demonstrate that they have been able to solve the “infinite space” problem. They will bring us new interesting solution to this problem. In the meantime, I will be there waiting, hoping to not be disappointed again. Excited as always to navigate in these infinite spaces.