An EXAG Science IV

Screen Shot 2013-08-30 at 15.15.20

(This is a series of short ‘previews’ of papers to be presented at the upcoming Experimental AI for Games workshop at AIIDE 2014. Tune in live on Twitch on October 8th to catch the presentations of these papers, or find the PDFs online at

Procedural content generation (PCG) is a thriving area for games. Everyone from indies to AAA developers is using PCG. Spelunky, Minecraft, Diablo, Dwarf Fortress, and many others use PCG at the core of the game. But are the games we have now using PCG in all the ways they can? Where has PCG been and where can it go next? Gillian Smith, in her paper “The Future of Procedural Content Generation in Games“, covers five major lenses on PCG and what unexplored areas the future might hold. Read on for a preview. 

Current PCG research focuses on personalizing games to players or improving how PCG tools interact with designers. Personalization tailors content to player preferences or skills. Improved PCG tools let designers give features to guide generated content. For example, adjusting how difficult a level should be to get different possible levels. But these are only a small slice of what PCG can be.

Also, PCG has only looked at a narrow band of player experiences. PCG mostly creates opportunities for challenge (Spelunky) or open-ended discovery (Minecraft). Are those the only kinds of experiences PCG can support?

With this context, the paper traces five trends in PCG, how they’ve shifted over time, and where they might lead next. One of the key trends is the aesthetics developers have used PCG to support. The paper considers experiences other than challenges or discovery that PCG could support. Recently PCG has become popular among player communities, creating a sense of fellowship around uncovering how the PCG works and ways to game the system. But what about giving players an experience of self-discovery? Or, are there ways PCG support the experience of a drama?

Another trend is the shift from single-player content (Spelunky) to shared multiplayer content (Minecraft). Could PCG go further? What about generating content with multiple instances where players influence each others’ worlds? Rather than working in a shared universe, a game could support players creating content that influenced the content generated for other players. What new kinds of games could arise from this kind of PCG and what new genres might emerge?

These are only two of the five trends in the paper. Check out the paper to learn more about data vs process intensiveness, the level of content interactivity, or shifting control over the generator. You can also find more examples of related games and changes PCG can embrace.

To see the paper’s full version, you can visit The paper will be presented at EXAG, and streamed live on Twitch on October 4th at

Leave a Reply