Dan Roy: Thanks for participating in this conversation, Raph. Let's talk MMOs. Throughout your career with MMOs, did you think at all about mobile (e.g. checking auctions via WAP or SMS)? What possibilities with mobile excite you, if any?
Raph Koster: Every project I've ever been on except for the very earliest ones, there was the question of what, if anything, we could expose via mobile. Ideas like letting people check friends online, check status of auctions, do crafting, and so on, were always common proposals.
Personally, I suspect that mobile is going to rapidly grow into being more than just a minor adjunct in the sense that those proposals imply. I think there will be full-blown mobile clients to virtual worlds.
Dan Roy: I agree completely that mobile will become more than just a minor adjunct and we will see full-blown mobile clients to virtual worlds. How do you think players on mobile clients will interact with players on more robust clients like desktops, laptops, and consoles? Will cross-platform interactions emphasize asymmetric roles and reduced time pressure to even the playing field across different interfaces?
Raph Koster: There's really as many answers to that as there are possible platforms… obviously, mobile platforms tend to be much more limited not only in their rendering capabilities but also in their interfaces and their network capabilities. Until there's greater parity, I think it is safe to say that the roles played online by users on different clients will have to be different.
Dan Roy: How much of a priority have features been that allow players to access data from or aspects of the game from a web browser? Do you see the browser as an opportunity to let players interact with MMOs from any computer? What opportunities does that create?
Raph Koster: Oddly, web integration with server backends is something that has come slowly and grudgingly. It took a long time for stuff like UO's guild metrics, DAoC's player metrics, and so on to show up. Now of course, the SOE "Players" offerings and the WoW metrics site have opened up quite a lot. I suspect that some of the reluctance has come from the tradition of embedding everything into the server. In the old mud servers, discussions forums were in-game, mail was in-game, and so on. What metrics existed were also exposed in-game, rather than via the web.
Similarly, I think that the pattern of requiring dedicated clients has precluded a lot of the thinking around web clients. I've written before about how by nature, virtual worlds want to be streaming – that's effectively what a text-based protocol IS, and that's how these worlds got their start. A true streaming protocol that was client-agnostic would open up the door for web-based clients, as well as much more – alternate clients, vertical applications for clients, and so on.
Dan Roy: What will it take to really open the door for all of these different ways of accessing a streaming world? Are we more likely to see many different access methods being developed by the studios that first create the virtual worlds, by other studios or individuals if the stream is opened up?
Raph Koster: Well, the core thing needed is really some standards on protocols. Right now, so much of the thinking around virtual worlds by the big developers is around walled gardens that typically, they don't even share network protocols within one publisher. Having common standards will be a pretty dramatic shift, but it is one I am sure is coming.
Dan Roy: How does the player experience change to be able to access the stream in many convenient ways? This would let players stay immersed in the virtual world more of the time and further blur the boundaries between game and life, right?
Raph Koster: Yes, it would. But more importantly, it would open the virtual world to lots of other kinds of interactions. Picture virtual worlds that can spit out game data as web services – or that you can access functionality from a browser. A network stream that could be directed to multiple clients at once that have very different display methods, like a tactical view versus a first-person view – the military is always asking for this capability.
Ironically, virtual world developers have done this a lot – they all have tool clients, and network testing clients, and the like, that monitor the health of the world using radically different displays.
Dan Roy: What do you think motivates players most to invest in improving a character in MMOs, whether it be leveling, getting better gear, joining a better guild, or any other way. Do you think seeing oneself become more masterful is a major part of that? How much of a priority has it been on your various projects to make sure players could clearly see their own increasing mastery (both in their avatar's abilities and their own skills)? How much of a priority was it to make this increasing mastery more socially visible - to group members, guildmates, the whole server, and the whole game population across servers?
Raph Koster: A sense of progress is critical to keeping players engaged in any game. Designers pay a lot of attention to providing that feedback. The sense of mastery can actually be conveyed in many different ways – levels and gear are far from the only ones. In more social settings, mastery can be measured by size of the friends list, for example, or by number of greetings when logging in, or by social influence.
It's always been important to present something of this sort to players in order to set social context. It's incredibly important that other players be able to see your level of achievement and respond appropriately. Doing it across servers is less important than doing it within the frame of reference of a given player, though – it's more important to get the social feedback from the people you interact with, than it is to have a game-wide leaderboard. Very few people reach the top on a large leaderboard, and it can have the opposite effect from what is intended, forcing the player to realize that their mastery is in fact nothing special.
Dan Roy: Can you give some examples of design choices you made, in spite of other drawbacks, that favor enabling increasing mastery or making it more visible to players and social networks? Can you give examples of the reverse, where you sacrificed increasing mastery or visibility for something else?
Raph Koster: Hmm. Certainly, in the general sense showing relative mastery is a good thing, but not always. So leaderboards is the classic example. Sometimes you create inadvertent leaderboards – for example, in UO we had bounties for playerkillers. The bounties were posted in town so that you could see who to hunt. But we sorted based on the amount, which made it into a high-score table of sorts – who could get the largest bounty? The result was that playerkillers went and worked to climb that achievement ladder. We ended up hiding it because it promoted the opposite of the behavior we wanted.
Dan Roy: Thanks for the interview, Raph.
In addition to designing MMOs, Raph writes a great blog.
