Hmm 0 references to procedural generation. I know the article is about design but I think it is still an important part of the discussion. Minecraft wasn’t out at the time this was written but Rogue and many of its successors were. Sometimes the problem with virtual world design is too much design, leading to the player being able to predict the world like a tv show with bad writing. Procedural worlds enable designers to focus on things that matter, and ignore the mundane details like placing specific trees in a forest or deciding what a nobody peasant npc’s hair colour should be.
Because it often is used as an excuse. Unless you’re like, Derek Yu, making a game that uses procgen is hard, unless the systems are that interesting to deal with. And you will almost certainly never get a good plot out of procgen.
And you will almost certainly never get a good plot out of procgen.
I have been wanting to do some experimental work in that direction for years. I can never find the time or resources though. I have a good mental picture of how it could be possible though. Dwarf Fortress lays some interesting groundwork. True most of the plots are either too random and unlikely, or weirdly generic and mundane, but every now and then it hits that sweet spot and you get something really interesting. My plan one day is to try to expand on that principle with the goal of a) reducing the amount of world complexity required (the hard part), and b) optimising to increase the frequency of good plots. It is a really hard problem, but a fascinating one and one that I think will eventually be cracked by someone. Being able to generate plots would be a big deal not just for games but also for AI prose fiction generation, and probably a bunch of other applications.
Some other early procgen games: Elite (1984) and Starflight (1986).
Sorry to keep posting links to the Digital Antiquarian, but I find that blog to be a an undersung yet extremely detailed history of gaming that everyone should at least bookmark or browse around in. The Starflight article is actually preceded (click “back” twice) by an entry-level description of Forth, so the reader can appreciate the significance of writing an entire game in it. :)
I read this book way back when… it certainly deserves its status as a classic. As someone who played MUDs as a kid growing up in the UK in the 80s, and then MOOs etc in the 90s when at university, anyone still interested in multi-user text based worlds would find lots to stimulate their thinking in the pages of this book.
Hmm 0 references to procedural generation. I know the article is about design but I think it is still an important part of the discussion. Minecraft wasn’t out at the time this was written but Rogue and many of its successors were. Sometimes the problem with virtual world design is too much design, leading to the player being able to predict the world like a tv show with bad writing. Procedural worlds enable designers to focus on things that matter, and ignore the mundane details like placing specific trees in a forest or deciding what a nobody peasant npc’s hair colour should be.
Procgen IIRC had a bad rap then. It’s very easy to make games that promise “infinite gameplay” but are hollow because of the lack of design.
That reputation persists to this day in some circles.
Because it often is used as an excuse. Unless you’re like, Derek Yu, making a game that uses procgen is hard, unless the systems are that interesting to deal with. And you will almost certainly never get a good plot out of procgen.
I have been wanting to do some experimental work in that direction for years. I can never find the time or resources though. I have a good mental picture of how it could be possible though. Dwarf Fortress lays some interesting groundwork. True most of the plots are either too random and unlikely, or weirdly generic and mundane, but every now and then it hits that sweet spot and you get something really interesting. My plan one day is to try to expand on that principle with the goal of a) reducing the amount of world complexity required (the hard part), and b) optimising to increase the frequency of good plots. It is a really hard problem, but a fascinating one and one that I think will eventually be cracked by someone. Being able to generate plots would be a big deal not just for games but also for AI prose fiction generation, and probably a bunch of other applications.
Some other early procgen games: Elite (1984) and Starflight (1986).
Sorry to keep posting links to the Digital Antiquarian, but I find that blog to be a an undersung yet extremely detailed history of gaming that everyone should at least bookmark or browse around in. The Starflight article is actually preceded (click “back” twice) by an entry-level description of Forth, so the reader can appreciate the significance of writing an entire game in it. :)
I read this book way back when… it certainly deserves its status as a classic. As someone who played MUDs as a kid growing up in the UK in the 80s, and then MOOs etc in the 90s when at university, anyone still interested in multi-user text based worlds would find lots to stimulate their thinking in the pages of this book.
Here’s the web site for the book, and indirectly, a summary of how well the chapters hold up.