Space and Games

Space has been a staple of games since the arcade cabinet, and the reason is partly practical rather than romantic. An empty volume is cheap to render, easy to navigate in three axes without the player getting stuck on scenery, and endlessly extendable: a designer who needs another level can add another system without explaining how the player walked there. The earliest space games leaned on this hard, because a black screen with a few bright points was within the memory budget when a forest was not.

The second reason is narrative. Space supplies isolation, scarcity and a countdown without any of them needing to be justified. A character alone in a capsule is alone for reasons the audience already accepts. Oxygen, fuel and distance are resources whose limits nobody argues with, which is why survival mechanics work so naturally there. Designers get a pressure system for free that would take a whole script to establish on a city street.

The third reason is that the real subject keeps producing new material. Every few years a probe returns images of something nobody had drawn before, and those images move quickly into games. The rings of a gas giant seen edge on, the grey dust of a comet, the banded storms of Jupiter photographed close up: each one arrives as reference material that no concept artist would have invented, and each one shows up in a game within a few years of publication.

The distances in space are the single hardest thing to put on a screen. If the Earth is drawn the size of a pea, the Moon sits about seven metres away and the Sun is a beach ball roughly eight hundred metres off. At that scale Neptune is around twenty five kilometres distant. A screen that honestly represents those proportions is a black rectangle, because everything worth looking at is smaller than a pixel and everything nearby is off the edge. This is not an artistic failure, it is arithmetic.

So every game compresses, and the interesting differences are in the method. Some shrink the distances but keep the body sizes, which makes a solar system feel crowded and busy. Some keep the distances and give the player a faster than light jump, which preserves the emptiness while making it crossable. Some use two scales at once, one for close flight and one for the map screen, and hide the seam in a transition animation. Each choice produces a different feeling of size, and players notice the feeling even when they cannot name the trick.

The most honest compromise in the genre is the continuous zoom, where the camera pulls back through several orders of magnitude and the simulation quietly swaps its units as it goes. It is expensive to build and it is the only approach that lets a player feel the jump from a cockpit to a system chart without a cut. When it works, the sensation of scale it produces is the closest a game gets to the real thing.

Orbital motion is the part of real space flight that contradicts a player's instincts most sharply. On the ground, to reach something you point at it and accelerate. In orbit, pointing at a target and burning raises your path and slows your angular progress, so you fall behind the thing you were chasing. To catch something ahead of you, you burn to go lower and therefore faster, then rise to meet it. Almost nobody works this out unaided, and the games that teach it well are teaching real orbital mechanics.

This is why most space games do not simulate it. The majority use what is often called arcade flight: the ship behaves like an aircraft, drag appears from nowhere so releasing the stick slows you down, and the nose points where you travel. It is not physics, it is a deliberate choice that keeps combat readable. A dogfight under true Newtonian motion is mostly two ships drifting apart at closing speeds no human can aim through.

The games that do simulate it tend to become teaching tools by accident. A player who has spent twenty hours planning transfers has internalised the vocabulary of periapsis, apoapsis and inclination change, and has a working intuition for why a launch window exists. That is a genuine astronomy education delivered by a game loop, and it is the clearest case of the two fields meeting.

A hand-built solar system costs artist time per body. A procedural one costs artist time once, for the rules, and then produces as many bodies as anyone wants. The technique is old: a generator takes a starting number, the seed, and derives everything from it through a deterministic sequence, so the same seed always yields the same system. Nothing needs to be stored except the seed and the rules, which is how an early home computer could hold thousands of star systems in a few kilobytes.

The interesting design question is what the rules encode. A generator that picks values at random produces noise, and noise is boring. A good one encodes real astrophysics loosely: hotter stars are rarer and bluer, rocky bodies sit closer in and gas giants further out, moon counts scale with parent mass, and atmospheric colour follows composition. The result is a set of places that feel like they obey a shared physics, because in a simplified way they do.

The limitation is that generated content has no history. A hand-authored place can carry a story in its layout, a ruin where something happened, and a generator cannot invent that without being told how. The current answer in most large games is a hybrid: generate the sky, then place authored locations inside it, so the scale comes from the machine and the meaning comes from a person.

Light is fast but the distances are not small. A signal to the Moon takes about one and a quarter seconds each way, which is why the Apollo transcripts have their distinctive rhythm. A signal to Mars takes between roughly three and twenty two minutes depending on where the two planets are. Real mission control does not hold conversations with distant spacecraft; it sends instructions and waits, and the craft has to be able to look after itself in between.

This is a gift of a mechanic that almost nobody uses, because it fights the basic loop of most games. A player who has to wait eleven minutes for a reply is a player who has put the controller down. The handful of titles that embrace it turn the delay into the subject: you write instructions, you send them, and the drama is in watching a plan you can no longer change play out badly. It is a rare case where accuracy produces better tension than the invention would.

The common compromise is to keep the flavour and drop the arithmetic. Voices arrive with a crackle and a short pause, messages are described as relayed, and nobody is asked to wait more than a beat. Players read the convention correctly, which is a reminder that a setting can feel authentic on cues alone while the physics underneath is quietly set aside.

The lunar landing sequence is the clearest case. A descent with a fuel gauge falling, a landing site that turns out to be strewn with boulders, and a manual redesignation in the last seconds is not an invented set piece; it is a reasonable description of Apollo 11. Games have rebuilt that specific shape dozens of times, usually without knowing they are quoting it, because it is a naturally good encounter: a hard limit, a surprise, and a decision that has to be made now.

Orbital rendezvous is the second. The docking minigame, with its alignment crosshairs and its slow closing speed, descends directly from Gemini and from the procedures that followed it. The reason it appears in so many games in almost the same form is that the real procedure is already well designed: it has a clear goal, a visible error signal, and a penalty for rushing.

The third is the failure scenario. A cascading systems failure far from help, with power, air and heat traded against each other, comes out of Apollo 13 and out of the long literature of station incidents that followed. It has become the standard structure for survival games in space, and it survives because the trade-offs are real ones that a designer would struggle to invent as cleanly.

Astronomers have been drawing the sky as a map for two thousand years, and the conventions that survived are the ones that turned out to be usable. Bright objects are drawn larger rather than brighter, because ink has no brightness. Lines connect stars that have no physical relationship, because a pattern is easier to hold than a list. Coordinates are angular rather than spatial, because from a fixed viewpoint direction is all you have.

Every one of those conventions is in the navigation screen of a modern space game. Dot size stands in for magnitude or importance, lines connect systems that have no physical connection but a travel one, and the map is usually angular around the player. Designers arrived there by usability testing rather than by reading history, which is a good sign: the same problem produced the same answer twice.

Where they part company is in the third dimension. A real sky atlas is a projection of a sphere and does not try to show distance, whereas a game map usually must, because travel time depends on it. The most common solution is a flattened slab with height lines dropping to a reference plane, a device borrowed from architectural drawing rather than from astronomy.

The governing relationship in rocketry is unforgiving: the speed change a craft can make depends on how efficiently its engine throws mass and on the ratio between its full and empty weight. Because that second term sits inside a logarithm, buying more speed costs exponentially more propellant. This is why a launch vehicle is almost entirely tank, why stages are dropped as they empty, and why every kilogram of payload is argued over for months.

Games almost always drop this, and for a good reason. A ship whose handling changes as its tanks empty is realistic and unpleasant to fly, and a player who cannot get home because of a planning error made an hour earlier will usually stop playing rather than admire the accuracy. So fuel becomes a simple bar, thrust is constant, and mass is fixed.

The titles that keep it find that it reorganises the whole game around planning. Routes are chosen days ahead, a heavy cargo run becomes a genuinely different flight from an empty one, and the most satisfying moment is not a combat kill but an arrival with the margin you calculated. It is a narrower audience and a deeper one, and it is the clearest example of a real constraint producing a real game mechanic.

Sound needs a medium, and space does not have one in any useful density, so an explosion outside a hull is silent to anyone outside with it. Nearly every space game ignores this, and the reason is not laziness: audio is the main channel through which a player learns what is happening off screen. Remove it and combat becomes confusing in a way that has nothing to do with realism.

The films settled this argument long before games existed, and games inherited the settlement. What is interesting is the rationalisation the medium invented: many titles explain their sound as a deliberate audio cue generated by the ship, a synthesised representation of sensor contacts played into the cockpit. That is both a decent excuse and, as it happens, close to something real aircraft do.

The handful of games that commit to silence get something specific in return. Vibration through the hull replaces sound, the player's own breathing becomes the loudest thing in the mix, and an explosion seen but not heard is genuinely unsettling. It is a narrower design that trades readability for atmosphere, and it works best in games about isolation rather than about combat.

Deep sky images are not snapshots. They are long exposures through narrow filters, each isolating the light of one element, then assigned to red, green and blue channels by the person preparing the image. The result is real data honestly presented, but the palette is a choice rather than a view. The best known images use one particular assignment of elements to channels, and its gold and teal look became instantly recognisable.

Games absorbed that palette wholesale. A generation of skyboxes uses the same gold and teal clouds, not because anyone examined the underlying physics, but because those images were what space looked like to everyone who grew up with them. It is a case of a scientific processing convention becoming an artistic convention through sheer familiarity.

The irony is that a human eye at the same location would see almost none of it. Nebulae are faint and largely colourless to unaided vision, and the naked-eye sky from deep space is mostly black with very sharp stars. A game that rendered that honestly would be accused of looking cheap, which is a neat illustration of how thoroughly the processed image has replaced the actual view in the public imagination.

Games position everything using floating point numbers, which hold a fixed number of significant digits regardless of magnitude. Near the origin that gives precision far finer than a millimetre. A few hundred thousand units out, the gaps between representable positions grow to centimetres, then metres, and geometry starts to shake visibly. Anyone who has seen a distant object jitter has watched this happen.

The standard fix is to stop moving the player. The camera stays at the origin and the universe moves around it, so the numbers that matter stay small. Large worlds add a second layer, dividing space into cells with their own local origins and rebasing as the player crosses a boundary. Both techniques are invisible when they work, and both are the reason a seamless planetary approach is a genuine engineering achievement rather than a content one.

Time has the same problem. Simulating a planet across millions of years cannot be done by stepping a physics loop, so long-term motion is usually solved analytically from orbital elements, and only nearby bodies get a stepped simulation. That split, analytic far and simulated near, is the same split professional astronomers use for the same reason.

The most common observation is that the sky is too busy. Generated skyboxes tend to fill the frame with nebulae and dense star fields, whereas the real sky is mostly empty, with structure that is faint and sparse. A second is that asteroid belts are drawn as rubble fields dense enough to dodge through, when the real ones are so sparse that passing through one without seeing anything is the normal outcome.

A third is lighting. In space there is one dominant source and no atmosphere to scatter it, so shadows are hard and unlit sides are genuinely dark, lit only by reflected light from a nearby body. Games often soften this with ambient fill because a half-black ship is difficult to read, which is a defensible choice but the most visible departure from what a camera would record.

What people who study the sky tend to say they enjoy is the opposite of a complaint: games are now one of the main ways a non-specialist ever encounters an orbit, a transfer window, a light-delay problem or a scale comparison. A title that teaches a player why a launch has a date has done something a textbook often fails to do, and that is the part of the overlap worth taking seriously.