Expanding the Final Frontier – Part One

Gamers are growing disenchanted with the increasingly timid triple A games industry. Risk averse to the extreme, these massive productions are terrified of failure, and thus hew as close to the mob wisdom of focus groups as possible. The scorn heard ‘round the world was deafening when after five years and more than six hundred employees, Ubisoft’s Watch Dogs was released and proved little more than a GTA knockoff where you occasionally hack traffic lights. Watch Dogs isn’t a mess, or a disaster. It’s just the most depressing safe game you could imagine. Afraid to offend, afraid to misstep, and toeing the line so closely that the only feel you get from it is generic.

It’s not fair to pick on Watch Dogs too much, though. The Chicago hack-a-thon is fundamentally a good game. It’s not alone, either. Assassin’s Creed 3, Forza Motorsport 5, Killzone: Shadow Fall, Titanfall, Grand Theft Auto V… none of these are bad games. In fact, a few of them I would even say were very good games. However, none of them are great games. To be great requires something more- it requires the courage and ambition to take risks. You have to reach for the stars, and try something you’re not sure you can pull off. In its current state, it’s nearly impossible for a triple A game to get made that aims for the heavens. Safe, inoffensive, traditional- these are the watchwords of modern triple A. Don’t rock the boat. Give your audience a game they’ve seen before with a little extra sparkle, because if you try to do more than that, you could sink the whole affair.


This wasn’t always the case, though. Once upon a time, we shot for the stars at the highest level… literally. Once, we made space games.

It doesn’t get much more ambitious than space. The final frontier offers a scale and mystery unmatched by any other setting, and that environment defines so much of the game from the word go. The Homeworld series showed us the incredible complexity and beauty of real-time strategy combat taking place on the three axis of space, as fighters danced across x, y, and z planes to dodge deadly plasma fire. Freespace, Wing Commander, and Tie Fighters saw us in the thick of it ourselves, piloting nimble craft on bombing runs and dogfights through nebulae and asteroid fields. Freelancer gave us a small galaxy, and simply asked “What do you want to do?” with no wrong answers. Battlezone 1998 reimagined bog-standard land combat on the moon with lasers and hover tanks. Space games were where imaginations went wild, and for a time these games got the budgets they needed to achieve their great ambitions.

What happened after that golden age of space games is a tired and familiar tale. Development costs grew greater and greater, necessitating fewer and fewer projects per studio, with bigger budgets and bigger teams. With a smaller number of commercial projects in the pipeline, any one failure would be a disaster that could very well shut the studio down, and taking the risks necessary for ambition and greatness lost all appeal. The triple A industry, much like the US government, looked to space and said “It’s really not worth the hassle.”


The industry turned on. Outside of triple A, space games still got made, leaving their respective marks. The excellent FTL showed us starship management as a roguelike (like). EVE Online brought us a baffling but astonishing MMO world that, while niche, was anything but small. Strike Suit Zero resurrected the spirits of Wing Commander and Freespace to give us a fresh space combat sim. The games are infrequent, and don’t tend to rock the industry too hard, but they’re far from dead. It looked like things would continue like this forever.

Then two games, two behemoths showed up to shatter our expectations. Both are yet unreleased, but there can be no question that for scope, ambition, and daring, No Man’s Sky and Star Citizen are two games that dwarf all that came before them. We’ve never seen space games quite like these before.

Made by the ten man team of Hello Games, creators of the motorcycle- platforming Joe Danger series, No Man’s Sky is a game that astonishes with its scale in proportion to its team. A massive two hundred man triple A studio would tremble at the width and breadth of what No Man’s Sky claims to be. Yet this group of ten has proven that the literally infinite sea of stars their game portrays is no smoke and mirrors tech demo- it is fully playable, and completely real. Dropping the player on a freshy built planet, populated with flora, fauna, exotic terrain and diverse resources, the game expands into a vast exploration epic as you fly through the endless reaches of space. Every new world is unique, forged by the game’s astounding fractals (a method of procedural generation) to have its own atmosphere, topography, generated species and structures, and so much more. The size of it all is staggering to comprehend- at first, it seems impossible.


The secret to how Hello Games did the impossible is revealed in the words of Sean Murray during an interview with Polygon [1]. “The new consoles are really good. But most of the games we have seen feel like experiences you could have had with the last generation. They have next-gen graphics but we are more interested in creating next-gen gameplay.” To that end, they have harnessed the processing power of the new generation of systems not to create more complex shaders and bump mapping (though the game looks quite nice thanks to its art direction), but instead harnessing that computing might to create content. Instead of needing hundreds of people to build all the assets for the game, they simply spent a year designing an engine that will build the content for them from a collection of smaller parts.  ”It’s hard to create this system, but then once it’s created, we’re kind of freed from content. We don’t have to worry about levels, or building light boxes, or whatever,” Murray told The Sixth Axis. [2]

By building this system from the ground up, Hello Games has found an entirely new way to build games. Procedural generation will never replace authored content when it comes to telling a specific story, and sending a specific message- the incredible emotions and concepts that a single artist can convey in a handcrafted game can’t be replaced by random generation. But when the goal is sheer breadth of content, the potential for a well-coded algorithm to churn out surprisingly, vivid worlds full of mystery is an incredible tool. No Man’s Sky isn’t the first game to try this- titles like Spore, Minecraft, and Dwarf Fortress have taken this technique to incredible places already. With No Man’s Sky, these English developers aim to show the world just how far this tool can be taken, and how it can be the basis for an entire game.


It remains to be seen just how the final product turns out- will the game be any fun? Will these worlds that our electronic overlords have built for us be worth exploring? What isn’t up for debate, however, is if this technique works in the first place. In a rare display of openness, Hello Games has been eager to show gaming press how well their technology works. They’ve offered multiple live demonstrations of how the game can create a thousand unique ships, planets, creatures, or even trees with the simple press of a button. It may have taken a year to build the base algorithm that powers these amazing feats, but clearly it was worth it. The potential of this system is incredible.

That is the part that really gets me excited- of course, I’m hyped for No Man’s Sky because it looks fantastic, but the most promising part of it is what it means for the future. Games like Spelunky and Terraria have already shown us a surface layer of what procedural generation can do- by taking premade objects and terrain, it can combine them to create an infinite number of levels. The constituent parts are still the same, however- the generator may place a Spelunky arrow trap in a thousand different positions, but it is still the same arrow trap no matter its position. It does the same damage to the player, it triggers in the same way, it is aimed the same way.

Hello Games is taking it one level deeper- using procedural generation to take premade parts, combining them to generate objects, and then generating levels with those objects. It’s an exponential leap in the variety that can be created- it doesn’t have to choose from a list of ten enemy ships to spawn, it can combine a fuselage with wings, fins, engines, and guns to create any of thousands of possible ships, all distinct yet visually recognizable as being part of the same art style. That this technology has its limits should go without saying- the more you automate a system, the less there is of that intangible yet distinct human touch that adds character and charm. Still, the ability to go a level further exponentially multiplies the possibilites, and allows small teams to create vast, diverse worlds.

Assassin’s Creed IV: Black Flag was made with a team of nine hundred people, all told. Like any project, not all of those people were working on its for its full development cycle, but certainly it was a massive undertaking, with a huge budget. As a triple A open world game, ACIV had a lot of small side activities- treasure chests to find, islands to map, cities to explore. Imagine how much money and work could have been saved if those optional open world features- a miniature Caribbean that takes half an hour of fast sailing to cross- were created by an algorithm instead of human hands. The game still contains a lot of content that would need to be hand authored- story missions, important side activities, and more- but so much of the ambient detail could have been handled by the software.


This is the potential of this development style- the freedom to outsource sections of your game to the machine, freeing you to make bigger games with more variety using fewer people. For many people who thought of a concept for a game, but realized that it was far too big for them to ever pull off, this flips that notion on its head. There will still be games that are beyond your means to make, too big and too hard with too much authored content for your team to create. With this incredible new tool, however, maybe it’s time to take another look at that idea you shelved. Maybe it’s not as impossible as you thought.

While No Man’s Sky is able to do so much with so little thanks to brilliant use of the processing power of new consoles, Star Citizen seems to be a more traditional development style… at first blush. Further examination shows that it’s a game so vast, complex, and bewilderingly ambitious that I couldn’t possibly cover it all here. Next week, we’ll dig into Star Citizen- and what makes it so incredibly fascinating.