With around 80 million registered users, 6 million unique visitors, and 400 million page views on its websites per month, the web-based teen centric online hangout/play space Habbo Hotel is one of the most popular online worlds on the planet - but is relatively little talked about in the game industry.

Sulka Haro, the lead designer of Habbo Hotel at Finland's Sulake Corporation, delivered one of the most entertaining and informative keynote speeches at last month's Austin GDC. Here, in conversation with Gamasutra, he takes on the very nature of play and what defines a game. He also talks more about Sulake's successes with implementing Scrum, and ponders what the attempts to define a "game grammar" mean to design.

Some people have been saying that while we don't look at products like yours essentially as games straight out, we probably should be expanding what our definition of "game" is. That's because Habbo Hotel is the sort of thing that people are playing. What do you think about that?

Sulka Haro: I guess I really don't look so much at the definition of "game" as much as I look at the definition of "play." If you look at Habbo, nobody can say that people aren't playing in there. People really do play in all of these environments, so I would use that as the unifying metaphor for discussing the different environments and products you can use to play. It's more clear.

Obviously there's products that are more "game," and they define gameplay, and the algorithms, as Raph [Koster] put it, where the meshing actually has a way to compute the thing that's going to happen next. As opposed to the purely social play, like Habbo. But people are still spending time doing something that could be really said to be "play."

Do you think there needs to be a new term for that? Is it important to label that, or does just the definition of "game" need to expand?

SH: I don't know. I guess I'm also looking at this from the perspective of not being an English speaker. In English, the word "play" is fantastically good. It's defined so broadly. In Finnish, there's several different words for different kinds of play, which means there is no one unified word we can actually use in Finnish to quantify the types of play that people do with games, because you probably do need to use one of the more specific words for different aspects. Also, the word for "game" is pretty broad, so that could be applied to different kinds of games, but that doesn't include the kinds of games like Habbo, which is more like the "gameless game" kind of thing.

Though there are games within it.

SH: Yeah. So if we really want to find a global way of speaking about it, it should obviously coming up with something new would help, because we could apply that anywhere. But then again, that's like coming up with new terminology. It usually doesn't work.

Global words are insanely difficult anyway. If you took the word "play" and said it phonetically in Japanese, for instance, you could get "play" or you could get "pray." It's not easy.

SH: There are people in the area of computer games saying that "game" means something that you can play with. I guess then that it would make it easier, and there wouldn't be any discussion as to whether products like Habbo are games or not, because they wouldn't like the word "game" to actually cover it.

Right. A game that plays with you is an interactive game, right? It would be hard to call it a video game if it's a board game and it's just you, because you could play all the sides I suppose.

SH: Yeah, but then again if you look at Habbo, if you go to a room and you're alone, it would be just furniture. There's no play. You're kind of playing, but it's the same as if you had a doll's house, just moving the furniture, which you can also kind of play, but it's definitely not a game.

Right. If you have the knowledge that other users are there and you're playing by yourself, and at some point you can invite them -- even if it were only on your own screen on your own computer -- you could invite someone over to your house and play.

SH: Well, in a case like the doll's house example, if you were hanging out in a group of people who all had doll's houses at home and you knew that they were playing with their doll's houses as well, and occasionally gathered everybody's houses together and played together, would that change it? It's essentially the same as being like you're in this big network [with] your own room, and you're playing and you know other people are there. Of course, with [Habbo] being virtual, you can actually just jump to other peoples' rooms immediately and start to play with them. There's none of the limitation of having to physically be there. But in terms to play, I don't think that it actually does change it that much.

It seems like you've added more of what we traditionally consider video game-like game elements to Habbo over time. What was the reasoning behind it?

SH: I guess the initial couple of games we did were very small. There's this funny diving kind of game, which is not a game at all. You just dive down, and you can do moves while you're diving. It's completely silly. We actually had a prototype of that even before the hotel itself was created, but that was not launched publicly ever, so the idea kind of lived, and we decided to do that with the hotel.

Then there's -- the name's just horrible -- Wobble Squabble, which is a... what do you call it, when people stand on a log, and you try to push the other guy off? With the thing that you pull? It's kind of popular as well. There's some mad players who have played like 50,000 games total. They've really lived in Habbo over a long period of time just playing the game.

So these would be like the super minigames, which are really popping out more and more nowadays, but then we've been expanding into doing more complicated stuff, where the user is actually playing what you would identify as a game, like the snowboard game, and you have proper games, like throwing snowballs.

I guess, partially, it's good business. There are people who actually want to play, and they pay money for it. But also, at least in my view, if you're looking at, like, a 13-year old guy, who is used to playing games, it's easy to communicate that, "Hey, there's this game-game here as well, and if you start off on playing that, maybe you'll get used to talking to the other users and get excited to meet people and eventually do the other activities as well." It broadens the scope a bit as well. The action in Habbo is really in the rooms themselves, so...

I don't enjoy MMOs or virtual worlds or things like that, but in the past I have used MySpace or other social networking sites. Habbo Hotel isn't that far removed from that, honestly. Well, I guess it does have one step of removal, because you have an avatar that you create, but I suppose that you could create an avatar on MySpace too.

SH: Theoretically you could play Habbo like you would play a social networking site, and certainly people have, so they can grab hold of contacts. But then again, there is so much more that you can do with the contacts. Look at LinkedIn, for example -- it practically died at some point, and I think it's because everybody who was interested in gathering contacts went there and got them, and...

There was nothing to do.

SH: There was nothing to do. Whereas what we have is that you can come in and get contacts, and then actually start doing stuff with other people. There's a fundamental difference as to what you can do and what you can't do, and social networking sites... I think there's plenty who are really struggling to add activities. With Facebook, for example, applications are just brilliant, and add so much you can actually do.

Without applications, it probably would pretty much die away, because you wouldn't have anything to do with the other people except pass messages, and there are other tools which are just so much more efficient for just passing messages between people. Being able to scribble onto somebody's walls, or play some games, and invite people to be vampires -- which Raph is so sick of, and I'm getting sick of that as well.

Luckily I'm not on Facebook, so I don't know.

SH: But anyway, there's a lot more. Even sites that were perceived to be purely social networking are expanding to be more like games, to have stuff for people to do.

It's funny that people really talk about the film and game convergence and stuff, but the game and social networking convergence is much faster. It's like a year or two years, that it's really ramped up to where you can't extract one from the other in many ways, like with MMOs or Xbox Live Arcade or things like that.

SH: There's some really interesting discussion going on currently, it's fairly new, about "social graphs." You can probably Google it up. People are envisioning that we should actually have systems whereby you can move your social graph of people from one service to another, and that there probably should be an independent third-party similar to, for example, the domain name system, which would upload your contacts, so that you could sync and have the same contacts available everywhere. It would be really cool if we could have that, because then if you went to a new service and were playing a game and your friend came online, you'd be able to automatically link friends if you want. Just moving the contacts between the different sites is a messy thing, though.

It's especially difficult, obviously, because the services really don't agree with each other, and function differently, even in the contacts side. With Xbox Live, you can add friends pretty easily, but with something like the Wii or the DS, you have to have friend codes, and you have to share those in an "outside of the console" context in order to make that connection.

SH: A problem that I'm hitting is that I'd want there to be more than one identity for myself within even one application. So for Flickr -- I'd want my streamed photos, like high-quality stuff. Then I'd like there to be another stream for stuff I take on the phone, which is completely different views. There's the proper photography, and then there's the [random] stuff happening. But I still want to be able to aggregate that in the same profile -- I don't want the hassle of having multiple accounts.

You're talking about different layers of information.

SH: Yeah. The same goes for my blog, and everything. Really, having an easy way to manage multiple identities and actually have some level of control as to how people see those would be really good. Also there's like... so if I go to a new service, maybe that service has implemented a way for you to import your contacts from another service? It might be that I actually don't want my friends to be able to import me from another service, because I'm doing something that I don't want people to know. But then again, just having to have multiple e-mail accounts just to be able to do that is a hassle as well. There really should be more refined ways to manage contacts and identities between services.

I'm sure someone's working on it and figuring out how they can make money out of it right now.

SH: If you look at the social graph discussion, that's basically pretty much it. But they're also saying that if somebody did start to make money out of it, it really doesn't work. So it probably should be independent.

And people always want to make money out of everything. It's always the way.

SH: Maybe we'll get some visionary willing to fund it non-profit, someday.

If I were rich, I would do things like that all the time. But I'm not rich, and I probably never will be! At your keynote, you briefly discussed Scrum. I want to know what you thought were the ups and downs of Scrum for developers. I know a lot of people have tried it.

SH: I guess yeah, if you tried it and abandoned it, it's probably because you just couldn't tolerate the initial pain of switching how you work. Changing processes really... if you're used to doing things a certain way and actually start to do things differently, it's a horribly difficult thing to do. I know from experience, and switching to Scrum takes two or three months at least before everything goes over. The first planning sessions are bound to be horrible. It's a fact of life. Because people are just struggling to start doing things differently.

If you're a producer who's really used to having exact control over what's happening, it's probably pretty hard to let go, because actually you do need to give room to people. But I guess the benefit really is that once you do do that, the people actually will be more motivated, and they will be doing things that they know are more productive.

For example, looking at scheduling, for example, if the schedule's wrong, with Scrum you at least know that real quick, whereas if you're using Microsoft Project, it could be that somebody works on something for two months, and only after that comes and tells you that the project schedule is not going to work. If you're having fast iterations, you have to at least... if you're going to be screwed up on schedule or design or whatever, you should know that pretty quickly, which allows you to actually react pretty quickly. You end up saving a lot of effort for projects that would fail.

I don't know if this affects you, but it seems to reduce crunch time for some people. You don't necessarily have specific dates that you need to hit with this kind of product, right?

SH: Yeah, but hitting the dates is easier as well. If you're really focusing on coming up with designs that are implementable within one sprint... once you get the system up and running and you start to hit the sprints, it will mean that you will know exactly what you'll be getting in the next thirty days.

If you would have a project that would fail the whole schedule, you will learn that sooner. So instead of working on the project for a couple of years and only then figuring out that you need six months to a year extra, you'll probably start to know that pretty early on. Because of that, it's going to be easier to go back and talk to whomever is financing the project and say, "Okay, we either need to change the scope or change the schedule, or come up with new ways of actually implementing it."

If you actually let the developers tell you, it could be that you will figure out new ways of accomplishing the same goals as you had, but something that's actually accomplishable within the schedule, which might be pretty hard to do if you had a big spec that's supposed to be implemented.

I also meant like within Habbo, since you don't have to target holiday releases or things like that. You don't have to deal with the typical...

SH: What makes you think that we don't have releases for the holidays?

I guess this is my question -- do you have fewer or do you have more periods where you're like, "Wow, we have to get this thing done right now!"

SH: Yes and no. Obviously for things like holidays, we need to have stuff out.

I suppose you need to have different content?

SH: Yeah. Then again, what we're doing nowadays is waiting for people to get ideas as to what we should be doing. We generally have an idea of whether it should be something that's doable within one sprint or two sprints, and how many teams will be working on it. We just cut down the scope so that it's something that is doable. That also means that you don't need to do that much crunch, because you already knew how big of a project it's going to be, and you can make the calls on how big [the project is] you're going to be doing.

How large is your team size?

SH: I can't remember exactly. Something like thirty people.

That's quite small. By comparison to your userbase, it's incredibly small.

SH: That's just the developers, though. Obviously there's a lot of people who work on different aspects.

Network administrators and things?

SH: Yeah, exactly.

Do you anticipate continuing Habbo for a long time, or do you have other types of projects that you hope to pursue as well?

SH: I can see us working on Habbo for a long time. There's millions of users who are pretty happy about it, so...

Yeah, and it doesn't seem like it's going to necessarily going to slow down anytime soon.

SH: Yeah. I want to continue working on it.

To what extent do you think it would be beneficial to expand onto console as well? Is that of interest in any way?

SH: Well yeah. It's a shame that you weren't there for the keynote. Really, the fundamental thing is really like text-based roleplaying, and with consoles, people don't have keyboards.

Though they could.

SH: They could, but they don't. It's a fact.

You could just plug in a USB keyboard.

SH: Still, people don't. We're trying to not fight the force of nature. If people don't have keyboards, I don't think we would like to be the first people to go onto a console and try to get people to use keyboards.

The first people to fail at that?

SH: Yeah. Obviously, we're really trying to push an update every month globally. That basically also means that if we did go to the consoles, I'm not sure if the mechanisms to actually do such a rapid pace of change in products is out there. Obviously you can download stuff to the hard drive, but still...

Do you have to localize content for regions quite heavily?

SH: The UI is always local. Especially with teenagers -- it's important that they actually use their native language to play the game, because they're obviously talking in their native language as well. We have local people, globally all over. There's like 19 offices, total, I think. The people in those offices obviously know the teenagers of that part of the market, so when they're hanging out to do the community management with the people in that particular hotel, it's going to be easier for them to work with the teenagers and drive the community. So yeah, we're definitely localizing, but not in the sense of actually coming out with functionality that's specifically aimed at individual markets.

In many ways, it's kind of a visionary project, in terms of the broad scope of the targeting. How did you envision that something like that was possible? How is it that you realized, "Hey, we could actually do this thing that a lot of people would actually like, that is this social networking-slash-game-like thing."

SH: The first Habbo started out about seven years ago, and it was a pretty small project back then. The company only had like eight people working. It's not like it was developed for years -- it just kind of picked up.

So did you not have the idea that it was going to take off like that?

SH: Well, there was a vision of letting users play around and do the content and all the activities, but no, people didn't know exactly if it was going to live three months or a year. I do remember this one day at the office when we realized that we had like 100,000 registered users, and being happy and realizing, "Oh wow, we're going to be big!" and not really imagining that there'd be like 80 million people -- this year's total right now.

What percentage of the population of the world is that right now?

SH: I don't know. The market penetration in some of the markets is incredible... I don't know exactly, but almost every single teen in the whole country who is in that age group has actually been there. It's kind of funny -- if you go and look at like eighteen-year-olds, or people who are already past the teenage age, they still have this thing in common, that they actually have been to this service and have played out. It's kind of funny, sometimes, to talk to people who are way beyond it already, but still remember the funky stuff that they did.

Do people eventually move out of the target age group for it?

SH: Yeah, definitely. Obviously, the fact that we have the teenagers in there is a big turnoff for the older people. In the States, I would guess that by the time you get your driver's license, you're getting interested in really meeting people live. It depends on the market. The age groups that have already... depending on which market we're talking about. But yeah, when you get to the age that you really want to meet people in the physical world, it kind of changes.

To return to a previous question, I was wondering how it was that you came up with the idea to let users play around with stuff. It hadn't really been done too much on a scale where it was easily accessible like that.

SH: As I said in the keynote, the people who founded Sulake -- the first core group of people -- they all had multimedia-slash-web backgrounds, and [were] not the games people. So we didn't even have this notion of stuff not being done before. It's kind of like really looking at all the websites that were already back then doing a lot of content -- obviously not to the extent where it is now, but really just looking at the past experiences and knowing that people want to do it.

Do you get a lot of people coming to you asking for advice about this kind of space now? Because Habbo has really taken off.

SH: Yeah, I guess. Obviously, we're here talking at a conference full of people who are really interested.

It kind of puts you into a visionary role. Are you prepared for that?

SH: Yeah. The weirdest thing I saw in a while was when Nicktropolis launched -- I went to the blog of the main developer, who confessed that he was an ex-Habbo fan who started off by copying Habbo, and then came up with technology and actually got licensed to develop Nicktropolis. I'm not sure if that developer's going to get in trouble if you put that on Gamasutra...

Well, if it was on his blog...

SH: So basically there's a lot of people who used to play in Habbo who copied it and then came up with worlds. There's plenty of products out there where you can see the solutions being copied. Some of them are good, in that people really got what it was all about, and actually knew what they were doing. And then there's obviously lots of products where you just see the people copy the UI, and not really understand what the game was all about. It could mean that they just screwed up somehow, and prevented people from having fun as a result.

There's also certainly the intent behind it, which is, "Are you trying to create a thing which is going to be really interesting to people, or are you trying to create this thing because you know people enjoy playing it?" Do you think that may have to do with some of that?

SH: Yeah, I'm aiming at the "fun" part myself, and money will come as a result of people having fun. It's much harder to do it the other way. If I'm out to make money, people don't have fun.

I wanted to talk about the concept of "game grammar" as espoused by Raph Koster.

SH: Establishing a grammar which would be accepted by everybody is obviously impossible to do. Given that we're talking about a global scope -- you'll definitely have to agree on a language to be used, when a lot of people are used to local dialects. That's going to be pretty hard to do. But I do subscribe to the idea of actually having a many dialects for communicating between people. That's kind of defined in the sense that they can actually communicate the idea. Especially if you have a problem in the design -- you can point to the exact point in the design that you think is flawed and causes the game to be not fun.

As I was saying earlier, there's this long talk on four different projects where people say that a game is not fun, but they actually can't say what's causing the game to be not fun. It could be the tiniest little detail, for example the user not getting adequate feedback for his actions. That might be immediately obvious and a designer could actually communicate that.

But especially if you're talking with someone who is not a designer -- it's real hard for these people to analyze these designs and be able to communicate exactly what their meaning would be, in the issues that they're trying to bring up. At least, if the grammar can be made simple enough so that even the non-designers could actually use it, it would be really beneficial for the industry, and I guess even for people outside the industry.

When I was talking to Denis Dyack, I was saying that with film, people understand the basic ideas like long shots, cuts, and things like that. People understand the terms that make up films. The only terms that we've established within the game industry, for the most part, aside from genre, are very technical terms that people don't understand. It wouldn't make sense to artificially create those terms, but if we could naturally evolve those terms so that people could understand those things easier, I think that would be really beneficial.

SH: Yeah, but if you look at the movie industry, those terms are pretty technical as well. A cut is really concretely referring to you cutting the film and changing to a different plate.

But it is conceptually easy to understand, because there's a break there. People know there's a break. And also for a long shot -- people know it's far away.

SH: Yeah, but I guess those are visual terms, so in the context of games, if people wanted to use similar things, you would be referring to the art of the game, and not actually the gameplay. You actually could be using the majority of the same vocabulary to talk about visuals within a game. It would be completely applicable, but Raph was talking about the more abstract part of the game -- the various activities that people are doing.

I guess there are some concrete examples of the abstract stuff that people actually know the terms for, like grinding, which is an activity that people do in the game that can be used to describe pretty accurately what it's all about. There's grinding, and Raph's actually talking about a grind as well, and saying there's a lot of games where you do grind, but the activity itself is obviously very different. But you can actually quantify different types of activities as grinding. I recall that he had some specific examples as to why that could be avoided.

I guess the vocabulary actually is establishing itself, but that's also... if you do look at grinding as a specific example, it's also partially because the games are actually doing similar things. People are used to doing grinding on so many different games. It's a phenomenon that's recognized by a population of people.

Given that it's actually hard to look at the abstract game design and separate that from the visuals, which is pretty much what Raph was talking about, it's harder to communicate about the game behind the visuals, because you need to go a couple of levels of thought deeper into what it really was that you were doing. And there's no need to actually do that.

Certainly he's really talking about the more binary level of games -- the games that take place within the games each second that you're playing.

SH: Yeah. I've been looking at the vocabulary defined by Eric Zimmerman in his books, and there's some really concrete, simple things that he's been quantifying, and telling that these things need to be in place for a game to be fun. Some are really concrete things, like the example I used earlier -- that players, when you do an action, need an immediate response, so that they know the action was committed. There are games where that's lacking, and that really does make the game not fun.

If you don't get a response, you don't know what you did wrong, and as a result, you know there's something missing. If you're talking about a game design and that's missing, it might be that the users can't really quantify and tell that that's the part that they felt was missing from the design. But if you actually do know that there is this thing and that vocabulary phrase telling that thing, it's going to be easier to communicate it.

It seems like there's almost a higher tolerance for that in the casual game market, because people expect computers to take some time to do something, especially with more casual users. That's just my perception. I don't know if you have thought about that at all. It seems to me like the delay of cause and effect is not as much of an issue there as it is in the more hardcore area.

SH: I don't think that's true at all. If you look at any of the Web 2.0 sites that are doing really well, like Flickr, it's packed full of these little visual cues that something's happening. Practically every time you actually press a button, the button changes, or you get a little progress indicator that something's happening. It might take a long time for the next page to load, but you at least know that something is happening.

That's especially important with AJAX stuff. The browser can't tell you that something is being loaded, so most of the really popular sites have implemented cues where the user actually does think that the UI is being responsive. It actually does work really well. If a page takes like five seconds to load, and that's the consistent behavior of your site, people will get bored pretty easily. You do have to do something about it, and tricking the UI to look a bit more responsive is a good tool.