Procedural Generation, Game Length, and Dead Horses

One of my biggest irritations with TV shows is when their writers seem to write without any regard for a potential end of the show. They write on and on and eventually the show drags on for so long that it might as well just feature the main characters beating a bloodied horse corpse. Video games suffer a similar fate when developers want to offer the player an endless stream of activities to do, ever vigilant to bring about the prophecy of the “ultimate game.” But the problem is these endless games most often leave the player hacking away at the soulless husk of the game they once loved.


Especially in the “games as a service” era of the game industry, developers want to offer big games to their audiences. This means big maps, extended storylines, or replayability. The thing is, I really like a game I can sink hours and hours into and still enjoy, but developing that kind of game isn’t easy. First and foremost, developers need to know exactly what they want their game to be. Sometimes developers don’t know so they over-extend the game and the end result suffers. It takes a lot of skill, insight, creativity, and discipline to make a game that people spend large chunks of time playing. We can see many developers try to compensate for the lacking of some of these with repetitive tasks, bland worlds, or the over-application of procedural systems.

I like procedural systems, they can lead to some pretty cool games, but they can also make a game seem lifeless and bland. What might have been better suited to a smaller world designed by a specific, human creator falls prey to the ambition of being too big. Take the example of the recently released Subnautica. A survival, base-building oriented game like it could easily have utilized a procedurally designed world in the hopes players might be more interested in exploring a world that is huge and unique to each of them. But Subnautica knew what it wanted to be and it worked diligently to create it. What resulted was a tight survival game that has a light, but interesting narrative that drives the player forward. Had it gone with a procedural world it might have been more ponderous and stretched its gameplay too thin.

But procedural systems can be used for more than world building. Games can also utilize them to build new missions for the player to complete. Done well, these missions can help bring you in to a reactive and immersive world. Done poorly, they can feel like lifeless systems meant to appease players when the main questline is done, as with Skyrim‘s much-reviled generated missions. The key is to make the procedural systems give players quests that make the world reactive rather than simply having them give the player more to do. A great example of this is how the Nemesis System in the Middle Earth series makes orc captains respond to what you do with them and has them act accordingly. Shame an orc captain too much? He might chase you across the game world and try to kill you. Leave one of your followers to die? He might cheat death and come back to kill you. Perhaps their objective doesn’t change too much, but the context given to the player is great. Though, maybe don’t try to drag out the story too much as with Shadow of War.

I love the things procedural systems can do; they have led to many of my favorite games. Though it might have not lived up to its hype, No Man’s Sky was an extremely impressive example of the power of these systems. Likewise, Elite Dangerous has done some cool things with their jumbo-sized universe. But both of these games have given up a bit of their potential character in the interest of making their games as big as possible. For some games huge worlds work. Minecraft seems to be a game that can be played forever because it’s less about the world itself and more about what players create in the world. Likewise, the Civilization series has featured procedurally generated maps that allow for a more organic expansion of empires as opposed to static worlds that might encourage simply racing to strategic locations. Meaningfully applied systems are far more important than those tacked on to artificially extend the experience.

This over-extension is also the effect of an emphasis on game length over quality. Oddly enough, one of the accusations constantly leveled at Call of Duty games is that their campaigns are too short, often clocking in under ten hours. I always wonder at these comments, because Call of Duty is primarily a multiplayer game. Also, as one of the few people who usually buys games from that series primarily for the excellent campaigns, I can say their brevity is a positive. As a serial non-completionist in video games, I find a nice, short story mode to be refreshing. Shorter, cheaper Hellblade: Senua’s Sacrifice was an interesting and encouraging game to emerge out of 2017. I hope to see more AAA developers follow Ninja Theory’s example and set out to develop smaller, tighter experiences, especially if they strive to break free of the $60 price point.

I love long games. I enjoyed every single hour of The Witcher 3 and I want to play through the campaign again someday, but it was an incredible game built over years of development with a team of highly skilled individuals. But I also love small games centered around big ideas. Many of these small teams we see developing games that utilize procedural systems do so for all the wrong reasons. As with any other essential gameplay mechanic, procedural generation is a powerful tool that needs to be limited to interesting and central mechanics of the game as a whole. Whether it’s Borderlands‘ bazillions of guns, Middle Earth‘s orc captains filled with personality, Crusader Kings II‘s dynamic family trees based around character traits, or even simply Star Citizen‘s procedural walking, I want to see more great uses for procedural generation and I want to see more short, great games.

Leave a Reply

Your email address will not be published. Required fields are marked *

Scroll to top