Book Review: Designing Virtual Worlds
It has been over a year since my last review of a vintage virtual reality book. I’ve recently come across a good one that I’d like to share.
In 1978, Richard Bartle co-authored MUD, the very first virtual world. In 2003, he shared his twenty-five years of virtual world and MMORPG experience in the book Designing Virtual Worlds. Here are some excerpts from the preface:
Too much virtual world design is derivative. Designers take one or more existing systems as foundations on which to build, sparing little thought as to why these earlier worlds were constructed the way they were.
Are designers even aware that there are decisions they can unmake? Although a good deal of design is evolutionary, that does not mean designers can’t be revolutionary, too.
The key is in recognizing the face that what seems eminently logical to you from your usual perspective might turn out to be disastrous when viewed from another angle — and then realizing that the worlds you’re drawing inspiration from almost certainly contain elements designed by people who didn’t recognize that fact until it was too late.
Obviously, the preface resonated with me on the topic of metaverse design.
The book is an incredible seven hundred and fourty-one pages, filled with decades of experiences and observations in virtual worlds. According to Wikipedia, it has been called “the bible of MMORPG design”.
The book is not a step-by-step guide to implementing multiplayer virtual worlds. It was published in 2003, so it doesn’t have significant comments to make directly on the topic of virtual reality or more recent MMORPG design elements. But that isn’t why you want the book.
You want the book to open your eyes to new possibilities as much as you want it to help you to question you own design choices. He covers a very wide range of topics from server architecture, player motivations, character attributes, ethics, and so much more. He introduces a topic, explains what has and hasn’t worked, and makes a few comments and suggestions of his own.
The only real controversy in the book seems to be in his observations into gamer psychology. This is based on his 1996 paper Hearts, Clubs, Diamonds, Spades: Players Who Suit MUDs. He proposes four major categories of players: Achievers, Explorers, Socializers, and Killers.
I made much more sense of the Killer category as those looking for dominance over others. A combination of competitive players (the kind that would enjoy DOTA 2), griefers, and in-world politicians. I think that the label of “Killer” is what leads readers down the wrong path, not necessarily the thoughts behind it.
So you can see how his four categories fit on the horizontal axis of Players vs World, and the vertical axis of Acting vs Interacting. In understanding what Bartle is trying to say, I think it makes more sense to put a different label on the vertical axis. Instead of Acting versus Interacting, it should be Dominating versus Exploring.
“Killers” wish to dominate people. “Achievers” wish to dominate the world. “Socializers” wish to explore people. “Explorers” wish to explore the world.
When I shared this observation with Richard, he agreed that the “Killers” quadrant is badly understood and badly named (inherited from a late 1980s perspective). He originally called the vertical axis active/passive, but that didn’t really capture what he wanted. It also had problems during peer review (active is not on the same dimension as passive, just as dominating is not on the same dimension as exploring).
While dominating/exploring may help us better understand what is going on in a two-dimensional player graph, it turns out that it doesn’t fit into the larger picture. Why?
Bartle had developed a third axis to distinguish between explicit and implicit actions. A player with a social interest could be intentionally networking with other people, or just casually making friends. (Our own friends who are developing social VR experiences should take notice.)
His view is that “I wouldn’t say all socialisers were exploring other people anyway; on the 8-types graph, the networker-socialisers would be but the friend-socialisers wouldn’t be.” That said, he thinks the labels aren’t quite capturing what is going on here. Perhaps after reading his book, you may be able to find a better set of labels.
If you are interested in seeing where you fit in to the 4-way classification, you can discover your own primary and secondary classification in the player interest graph via an online test. I’m primarily an achiever/explorer type.
I found it amazing that one of the most relevant titles in my vintage collection of virtual reality books ends up being the one that makes the least number of references to virtual reality itself. But as the author explains, we’re just inheriting and iterating on virtual world design decisions that have been made by others over the past thirty-seven years.
One last thing. Richard points us to his own early metaverse experiment.
Metaverse-wise, players could “seamlessly” walk from MUD to the second world I wrote, Valley (which used the MUD engine). I didn’t let them take anything with them, though, because it was too open for exploits in a game-oriented environment.
Maybe we’re just looking for a new way to implement a long-lost feature.
JUNE 8, 2015: Richard Bartle’s response has been integrated into the article.