Richard Bartle: There are many, many games I would want to write

The Nexus

The Nexus

▮ The Nexus: In the early days of MUD development, what specific challenges did you face in terms of game design, and how did you overcome these challenges?

Richard Bartle:  When we first introduced the option for characters to come back as new ones after they died, it didn’t really work. Most people just played as themselves instead of getting into role-playing. Realising we needed to encourage a culture of role-playing, I, with a background in role-playing, created a debug character named Polly. I played Polly as a cheerful and lively young woman, to show players that it was not only permissible but actually fun to take on different identities in MUD. This move successfully freed up players to explore different sides of themselves.

Roy and I had also discussed player advancement. We came up with a points system to track players’ growth. Roy proposed it, and was quite amused when I told him that that’s how D&D (Dungeons & Dragons) did it. When it was time to implement the concept, though, I had to do it myself. Drawing ideas from D&D, I introduced the idea of levels. Levels added extra momentum to experience points by setting clear goals. Without levels, going from zero to the maximum points might look daunting, but with levels it was broken into manageable steps. Also, reaching a new level came with rewards such as spell acquisition and stat changes, boosting the sense of achievement when levelling up.

▮ The Nexus: How did MUD capture the attention of players and achieve popularity?

Richard Bartle: It did both by being the first of its kind and being very good. We encouraged others to create their own MUDs (which is why the original MUD was referred to as MUD1 by players, to avoid confusion). We did this not just for the sake of free software, but to give people alternatives to reality. Our goal was to provide more virtual worlds, offering people freedom through diverse gaming experiences. I also put the concept of virtual worlds into the public domain to prevent someone claiming ownership or trying to patent it. Roy and I wanted to keep the idea alive that virtual worlds could give people more freedom and individuality, even if we weren’t there anymore.

▮ The Nexus: What do you think modern game designers could benefit from revisiting the roots of MUDs?

Richard Bartle: They can learn not to have to relearn what we already know. When graphical worlds came out, MUDs were dismissed by some new designers for being irrelevant because they were text; those designers promptly made all the same mistakes early MUD designers made. When VR came out, graphical worlds were dismissed by some new designers for being irrelevant because they were 2D; those designers promptly made all the same mistakes that early graphical-world designers and early MUD designers made. When those of us in the virtual worlds industry saw what basic mistakes were being made by advocates of the Metaverse, it was alarming – predictable, but alarming. They really should have asked some old-timers what to do – and listened to them.

▮ The Nexus: Your 1996 paper, “Hearts, Clubs, Diamonds, Spades: Players Who Suit MUDs,” introduced the influential Player Types Theory, categorising gamers into Achievers, Explorers, Socialisers, and Killers. What does this theory actually say? Have you considered expanding the model to encompass new player motivations and behaviours? 

Richard Bartle: The theory says that people who play virtual worlds for fun find them fun in different ways. Furthermore, if you design your virtual world so that it appeals only to one type of player, it won’t be a success: you need all types of players, because they interact with one another. The model was refined in 2003 to split the types in two, so there are eight types¹ instead of four. This was done to explain why some types (particularly those who like acting on other players) seemed to have very distinct kinds of behaviours and to explain how players progress through types as they play. Ultimately, it explains why people play virtual worlds – but not any other kind of computer game. They may work for other types, but if so I can’t explain why.

There are no other types of players: the theory of player types is exhaustive for people who play virtual worlds for fun. You can subdivide existing types, but there aren’t any extra ones that can be slotted in. If you want to include people who don’t play for fun – gold farmers, journalists, educators, customer service representatives and so on – then player type theory doesn’t apply. 

▮ The Nexus: Your book Designing Virtual Worlds is a seminal text, not just for game creators but for anyone interested in building virtual communities, navigating ethical questions, and understanding the complex bond between design and player experience. In 2021, your decision to offer this book² for free was met with widespread praise and gratitude from the game industry. For new readers, which parts of this book remain relevant and insightful today? Are there specific chapters or ideas that have gained even more importance with time?

Richard Bartle: I don’t think any chapters have gained more importance; most were important then and are important now. It’s quite dated in places, though (remember, it came out before World of Warcraft!) and many of the discussions concern what were issues then that are no longer issues today. Chapter 6, which is to do with how virtual worlds are looked at from different academic disciplines, is very out-of-date now. Thousands of papers have been written on the subject since then.

▮ The Nexus: You previously stated: “The key point of designing a game is that it’s not craft that’s important, it’s art.” Could you elaborate on it?

Richard Bartle: When you design a game, you are saying something to your players. You can say it in a number of ways, but the central one is through gameplay. Gameplay is what games have that nothing else has. If you only wanted to tell a story, you’d write a book or a screenplay. Games tell the players things procedurally, through their systems. Now, if you regard them as purely craft, you get something serviceable but with no message. It’s purely an engineering problem: how to arrange gameplay elements in such a fashion that they work together to meet the project requirements. The resulting game will be soulless, though.

You do need craft, but you also need art: the game designer has a vision that they’re trying to express, and they’re doing it through games because gameplay is the medium they have to use to express their vision. If they could express it some other way – as a painting, or a symphony, or a dance, or a sculpture, or a poem, whatever – then they would. Because they’re trying to say something through their game, they have a consistent, coherent sense of what must and must not be present in its design (which actually makes designing easier, believe it or not). They don’t necessarily know what they’re saying, only that they’re saying it and that it means something. This gives it its soul.

▮ The Nexus: As a game designer, what are some common pitfalls or red flags you see in games, even those that may be commercially successful?

Richard Bartle: People don’t know why they’re designing what they’re designing. They think it’s sufficient to make one that they, personally, will find fun. Players aren’t great at designing games (neither are designers, they’re just orders of magnitude better than players). As a result, we get games that look the same, play the same and have nothing to say. There is room for these games, but there are just so many that they bury the treasure.

▮ The Nexus: Professor Bartle, looking back on your years in game design, how would you say the role and capabilities of AI in games have changed over time? Do you think the rise of large language models like ChatGPT has the potential to further revolutionise AI in games? If so, in what ways might this unfold?

Richard Bartle: I did my PhD in AI back in the 1980s because I wanted intelligent NPCs in virtual worlds, and I’m glad we’re getting closer to seeing that. In the past, AI was always hamstrung by the fact it was slow, but Monte Carlo Tree Search helped speed that up and with AI-as-a-service we’re getting to the stage where sophisticated AI responses of certain kinds are possible in real time (as far as players are concerned). GPT does have potential here, but you have to make sure it’s trained on relevant data. You wouldn’t want to be chatting to a guard in a medieval Fantasy setting and suddenly find the conversation has turned to the merits of various real-world soccer teams.

AI isn’t all good, though. If you think discoverability of your games is hard now, wait until companies are churning out a hundred games a day and putting them on app stores. It would cost them practically nothing to create them, so any income they happen to generate by accident will be pure profit.

▮ The Nexus: What are your thoughts on the current revenue model for games? What do you think can be done to promote fairer and more ethical monetisation practices?

Richard Bartle: It depends on the games and the revenue models. Some are reasonable for some games but not for others. I don’t like pay-to-win in any of its forms. It’s like paying to have heavy weights attached to your competitors in a sprint. As for how to promote fairer and more ethical monetisation practices, well the first thing to do would be to have a fairer revenue distribution. Apple takes 30% of the revenue that any app generates for its platforms, despite contributing nothing to their creation. If developers were allowed to keep more of the money that their own work has made, they wouldn’t have to try to squeeze every last drop of money from their players at the cost of their own integrity.

▮ The Nexus: As a professor who teaches game design, how do you approach teaching students, and what are your primary goals in the educational process?  

Richard Bartle: I assume that my students want to be game designers. Most don’t – they either want to be developers (but not designers) or they’re only taking my module to avoid another one that they perceive is harder (until they find out too late it isn’t!). My primary goals are to look for the sparks in the students’ heads and throw petroleum gasoline on them until they burst into flames. A lot of the time, too much fuel for thought can stifle a student’s creativity, but those students weren’t destined to become game designers anyway. For the remainder, they take off and go in whatever direction their creativity takes them – whether game design or something else.