The Role of Procedural Content in Games
I find it so interesting how the role of Procedural Content in Video Games has changed and Evolved since the first hit games back in the 1980’s.
When video games were first emerging, procedural and procedural-like generation was used out of necessity. Consoles and cartridges of the time had next to no memory, and it was impractical in some cases, and impossible in most, to store all the content required to make a satisfactory game. This meant developers had to get creative. One of the ways developers of the time got around their memory limits was to rely on procedurally generated levels, rather than pre-defining and storing levels in the code. Developers would come up with a world generation algorithm which relied on a pseudo-random number generator, and then just plug and play seeds until they found levels which were adequate for their needs. They could then store just the algorithm and the seeds, for a consistent experience which took up a lot less space. This technique was employed for “River Raid”, an Atari 2600 game [1]. One of the more famous games using these techniques, Elite, employed an algorithm so successful that they were asked by Acornsoft to reduce the number of levels so as not to “rub the player’s nose in the artificiality of what they were enjoying” [2].
As resources grew, this type of generation largely fell by the wayside, in favour of pre-designed levels. And as resources grew even more, pre-defined stories, characters, and dialogue beyond “The Princess is in another castle” became expected. In recent years, procedural generation has come back into vogue. Probably the biggest driver of this trend is Minecraft, whose procedurally generated worlds have captivated both players and developers. Open-world, sandbox style gaming (with or without PCG environments), have also become extremely popular. Players like being free of restrictive level design, and enjoy being able to do what they want, as opposed to what the game wants them to do.
Today, our group is looking at using procedural algorithms to control both the environment and the story of our upcoming project. We are looking to create something with an extra dimension to games like Minecraft. We want to create a world that not only builds its environment, but that also builds its own narrative. The player’s choices and actions will contribute to this narrative, and hopefully involve them on a much more personal level than existing games today. We already have some great ideas as to how we can communicate this to not only developers watching us, but the actual players of the game. With limited time and resources, we don’t expect that our game will be complete, or industry changing by next April. But we do hope to push the boundaries of what is currently possible in gaming, and have something to show the world next year.
Sources
[1] http://www.riverraid.org/riverraid_history/index.php
[2] http://www.theguardian.com/books/2003/oct/18/features.weekend
