Winter In Reykjavik: EVE Fanfest 2007 Report

The 'sleeper hit' of the MMO business in the last few years has been space trading game EVE Online, which now has more than 200,000 active subscribers, according to recent press reports. Of course, it is well-known that the game makes no apologies about its steep barrier to entry, so it should come as no surprise that the fourth annual EVE Fanfest took place in Reykjavik, Iceland during some very cold and wet November weather. Despite the remote location, EVE developer and publisher CCP Games was able to attract over 1,000 attendees from several different countries.

Particularly for game developers and publishers -- CCP is both -- it would seem that companies tend to hold new features very close to the vest, only announcing them when plans are near completion or implementation is well along. In contrast, CCP Games was remarkably open about a lot of material that is still quite early in development.

The developers and management make no excuses for the fact that they haven't yet figured out how announced systems and technology will work. They also solicited player feedback in a very active and direct manner. The roundtables at the event typified their approach. The developers actively listened to the participants and directly addressed their questions and concerns.

The tone of the interaction was refreshing. There was very little defensiveness on CCP's part, even for issues which they couldn't address to a participant's satisfaction. Each roundtable had a CCP employee present whose sole duty was to keep notes of participants' suggestions. Implicit in the open nature of the face-to-face forums, they put a fair amount of faith in their player community that the participants wouldn't monopolize the experience for each other. And for the most part, their goodwill was justified.

Part of this seemed to be due to the attendee demographics. Although Gamasutra doesn't have an exact breakdown, anecdotally the crowd seemed to skew a little older. Because there was an open bar at the event, attendees had to be 18 or older. Most were considerably so. The average age appeared to be in the late 20s to mid 30s. Many attendees also worked in tech, IT, and game development themselves. Women players seemed to be fairly represented, too.

The three-day event covered a range of presentations, roundtables and activities. For Thursday, November 1st, most attendees were still in transit. The doors didn't open until 2pm, and the sessions were smaller and more specialized. These included an explanation of the development and theory behind EVE's in-game heat generation and dissipation systems, a session on expert PvP strategy, a tutorial on how to create and promote EVE fan sites, a discussion on the upcoming EVE novel by author Tony Gonzales, and an evening group event at the Blue Lagoon hot springs. Roundtables for the day featured opportunities to discuss most-wanted features for future game expansions.

The majority of attendees showed up on Friday. Specialized sessions included a technical overview of EVE's new shader model 3.0/DirectX 10 supported graphics engine Trinity 2, an explanation of EVE's roll-out in China, a presentation of the new quarterly economic newsletter by CCP's resident economist, Dr. Eyjó Guðmundsson, a presentation on EVE's in-game VOIP technology, and an artistic and functional presentation on character customization in EVE's off-ship avatar project, Ambulation.

Presentations in the main auditorium included a series of lectures and discussions on EVE's planned Council of Stellar Management (CSM). The CSM is perhaps CCP's biggest "announcement" for Fanfest. It involves the planned introduction of a system to elect a seven-member executive council with the intention of providing high-level feedback to CCP.

Following an introductory presentation by Dr. Guðmundsson and the author of the CSM white paper, Pétur Óskarsson, industry luminaries Jessica Mulligan and Richard Bartle debated the viability of such a system in light of adjusting the expectations of the player community.

Other main auditorium sessions included a panel discussion on the politics of EVE as explained by high-ranking members of various in-game alliances, a real-time demonstration of digital art creation and commentary by CCP Lead Artist, Björn Börkur Eiríksson, and a formal overview of CCP's Ambulation project. The auditorium was packed for the PvP Finals, followed by "The Great and Marvelous EVE Show," a game show pitting CCP staff members against each other in contests of dubious skill and talent, including "Cow, Lake, Bomb," a rock-paper-scissors variant where "cow drinks lake, lake defuses bomb, and bomb blows up cow."

Saturday, EVE Fanfest wrapped up with a small session on environment creation in Ambulation. Large presentations in the main auditorium included a historical overview of CCP Games by company co-founder, Reynir Harðarson, a joint lecture on the advanced features of Trinity 2 by CCP Chief Technology Officer Halldór Fannar and NVIDIA representative Phil Scott, a tour of EVE's new expansion content, and a "World Domination" keynote by CEO Hilmar Pétursson followed by an open Q&A session with the audience. CCP Games capped off the fanfest with a party featuring live music courtesy of Röxör, a rock band comprised of EVE developers. At the end of the set, lead guitarist Runar Thorarinsson thrashed his axe Pete Townshend-style to the delight of the Tuborg-fueled crowd.

There were a few other unique offerings on hand for the fanfest. In cooperation with new CCP partner White Wolf Games, the fanfest hosted a tournament for the EVE collectible card game. Also, White Wolf designed and executed a Live-Action Role Playing event based on the fiction of the EVE universe. All fanfest registrants were welcome to participate. The game involved a series of missions spread among four tiers of difficulty. Tasks included individual, as well as group challenges, a mission that required players to log into EVE Online to acquire information in-game, and a round of the EVE collectible card game. Rewards were doled out in ISK, EVE's in-game currency. The ISK rewards were redeemable at the CCP company store or within the game itself.

Also present were the publishers of EON, the official print magazine of EVE Online. Published in the UK, EON magazine is funded entirely by subscription. It is a high quality, full color magazine that garners much of its content from the activities of the player community. Interestingly, the advertising is placed by the players, and is paid for using in-game currency.

There were some other unique items and concessions that CCP Games had built into the event. One was their Sisters of EVE program. From the EVE Online website: "This year's event will celebrate the unsung heroes of EVE Online: Significant Others. The Sisters of EVE Program will give your partner a chance to visit Iceland along with you: while you discuss the nuances of the Rokh's PVP setup with fellow corpmates, they can take a day trip to a geothermal spa, shop in the downtown district, or tour the Icelandic countryside."

Another was the promotion of Fanfest on the EVE-Online.com website. Ad banners showing in-game avatars featuring recognizable player characters, saying "Will Be There!" And indeed, more than one player had signs proclaiming they were looking for other players (or griefers) with the intention of giving them a piece of their mind. Interestingly, there appeared to be a bit of recruiting for in-game corporations. Players would discuss their skill ratings and daily ISK earning potential. There was even a bulletin board in the bar area that players were using as an impromptu job-posting system.

CCP's attention to detail extended to the venue, which was intended to loosely approximate an EVE space station. In fact, CCP Games CEO, Hilmar Pétursson, commented that the architecture of the location was quite Caldari (one of EVE Online's empires) in style. CCP adorned the fanfest with full-color signage, and a sense of humor that extended to the bathroom stalls. On the inside of each door, they had posted a small sign that read, "Are you sure you want to jettison those items?"

Despite a few allegations of developer misconduct, CCP has managed to build up a high degree of community goodwill. Their willingness to openly interface with their player community, even for features that are very early in development was much in evidence during the event. According to company CEO, Hilmar Pétursson, initiatives such as the Council of Stellar Management are the next logical and necessary step in that process.

Developer responsiveness and community management has been a fair contributor to EVE Online's average 80% annual growth. If EVE Fanfest 2007 is any indication, CCP Games certainly looks to maintain that focus as the game moves into the future.

Gamasutra also got a chance to sit down with Pétursson and discuss the fanfest, why it's OK for gamers to hate EVE, the formation of the Council of Stellar Management, developing (and redeveloping) a scalable MMO, and more.

There are a lot of people here who've come to Iceland in November -- over a thousand -- from many different countries to participate in this fan convention.

Hilmar Pétursson: We've been quite amazed ever since we held the first fanfest back in 2004 -- the willingness of people to come to Iceland in November -- which almost is the worst time to come here. But I think nobody has ever come without feeling the trip more than worth it. We see people coming back year after year, and as the player has grown -- EVE has been growing about 80% per year since we launched it -- the fanfest has been growing in a similar way. I think a lot of that has to do with repeat visitors: people who have come for every year.

Initially they might have come here to sort of listen to what we were saying about the future development of the game. But I'm seeing people coming here more and more to meet each other and to participate in various... even player organized events during the fanfest while they're over here which we definitely celebrate and want to continue to support.

Let's go back a bit and talk about the history of your game and your company. You said you've had pretty active year upon year growth in your community. It's obviously not as mass as something like World of Warcraft, nor do you try to be a mass-market property. But somehow you've managed to maintain your success being in some ways what could be called a niche property.

HP: Initially, we wanted to create a space fighting/trading game. We wanted to found the company in Iceland and do it from here. The company was founded in 1997, and we started developing the game in 2000. Up until then, we were focusing on professional service work and building up the company -- recruiting the team and all that.

Even though we said that we're doing a game, we wanted to do it in a way that the game would really be there for the players to build. Not in the terms that they would be modeling the content themselves in Maya, but that everything interesting in the game would be emergent from systems that we would have created, that the players would really have built upon.

That theory was maybe very similar to what you see now in social networking websites and things like that. There are a lot of fancy terms for these things now, but we sort of intuitively had the idea of building a product that was more like a service and more for all the players.

When you think of it, we always thought we were planting a seed that would grow to a tree. Exactly how tall the tree would become wasn't necessarily something we aimed for. So when we see the game grow from 50 to 100 to 200 thousand players, it isn't necessarily a surprise. It's more sort of realizing the vision that we have created fertile enough ground for the tree to grow this tall. We will continue to focus on fertile ground for the tree to grow rather than trying to force it to happen. And our vision for that has been to create a product that people love... but not like. We want people to love or hate it. People shouldn't play it because they like it, they should play it because they love it.

What that implicitly requires is that your community puts a lot of work into the game because they're as much stewards of the game's direction as you, who create it. What were some of the mechanisms that you thought of when you envisioned the game, and have those mechanisms followed in good stead since you've enacted them?

HP: I would say what ties it all together is the economy. The economy of the game is very much controlled by the players. All prices are decided on the market, CCP doesn't set a price on it. It's freely traded on an open market and people price it based on how they value their time or the scarcity of what they have to sell. And then we have systems built around the market: we have a manufacturing system, we have a resource-gathering system.

And then the game very much focuses on [the fact] that you're always at risk in terms of all the players attacking you or taking something away from you. So that creates very interesting interactions between war and the economy. And war and economy is something that has created a lot of events in human history. That is essentially what we maybe have put in place to drive the storyline. But then, the players have used those systems to create something much more spectacular than we could ever have envisioned in the beginning. So I would say, the economy is the tool to create this.

For soliciting community feedback, we have used various methods throughout the four years. And we're trying to evolve those as our world has evolved. You use different methods for a community of 50,000 players than you do for a community of 200,000 players. Especially when all those players live in the same world. It's different when you have sharded worlds down to smaller shards, and you just have more shards. And you have to tackle the community of each shard. Then you can use the same method, but scale it up.

But when the community fundamentally grows as it has in our case, then you have to adapt and evolve your method of soliciting community feedback. And we're now, at this fanfest, introducing a new idea which we call the Council of Stellar Management which involves allowing the community to elect representatives for a council. And we'll do this through voting. So this council will then be a venue for exchange between the community and CCP so that it is a more meaningful discussion than us talking in a non-structured way with 200,000 people which... um, achieves very little in its current form.

I just sat in on a panel discussion with two game theorists and two representatives from CCP. On one table, we had industry luminaries, Jessica Mulligan and Richard Bartle and they were maybe playing devil's advocate with the idea, because this is an introduction of an idea. It's something you're thinking of, but haven't yet implemented.

And on the other side, you had Dr.Eyjó Guðmundsson and you had Pétur Óskarsson who wrote the white paper for the Council. I thought Jessica and Richard brought up some really good points. Their main issue was managing player expectation in terms of how you implement this. It isn't necessarily something which exists in the game fiction. It very much is something which happens in the so-called real world as a high-quality feedback interface between your fans and CCP...

HP: We have this idea of creating a council which fairly represents the world. Because it's easy to get the vocal minority to be represented -- they're vocal and you hear them. But there's a large silent majority in the world. We're hoping that this method of having them democratically elected will at least be the process that will evolve to the point where we actually get representation from the silent majority.

We have sort of developed a system for that and we have presented that here at this fanfest and we definitely wanted to get devil's advocate-like feedback from industry luminaries. We're very honored to have Richard Bartle and Jessica Mulligan to give us some very pointed questions and guard us from the obvious holes we could trip into. And they did an awesome job today and it was a very... I would say an exchange of ideas and a discussion which I felt brought us closer to the correct solution, and we will continue this throughout the day with roundtables and player feedback and will definitely make sure that everybody agrees that is a path on which we want to go on.

We very vigorously think this is the right thing to do for the game at this point, and we build that on our experience from managing the community from 50,000 to 200,000, seeing the sort of trends that evolve, trying to predict how it will be when it's up to 300-to-400,000 people. You would definitely have to install something like that to make sure that you have a fair representation from everybody within the world.

As with our method of soliciting community feedback and that has evolved throughout the years, our code base has definitely evolved in a similar way to manage the scale. And even though we're not mass-market in the terms that we don't have millions of people playing our game, we're big in the way that everybody's playing the same game on the same shard. So the scalability challenges that we're faced with are maybe larger than [those] of sharded games where you can basically manage scale by adding shards. We don't have the luxury of doing that so have constantly been evolving and innovating both in terms of hardware and software, to be able to manage the scale.

Last year, we completely changed our architecture and moved it over to 64-bit computing from 32-bit computing. Before that, we installed solid-state hard drives into our SQL server to be able to reduce I/O waiting queues which were creating latency within the game. Our next step is something that we are introducing at this fanfest as well: it's the move over to supercomputing where we will take traditional supercomputing solutions from, basically research and academia which have been the fields traditionally using high performance computing solutions. But it's now becoming commercially viable, and HPC as a whole, to the point where you can do simulation-based activity that isn't only batch oriented as it is currently sort of built for.

And we're now in a massive research project with Microsoft and IBM where we are going to construct a substantially large supercomputer, which we estimate will be in the top 500 supercomputers in the world, which we will employ to break some of the limitations our designers and programmers have had to work around. And I'm hoping to talk a little bit about that at my closing keynote this fanfest and explain our motivation for doing it, and also the opportunity we see from doing it.

One thing EVE has always been known for its beautiful visual presentation. It has very strong art direction. In fact, reading through your book, The Art of EVE, one of the quotes from company founder Reynir Harðarson was, "If it looks good, it is good." This is very different for a game development studio because typically, they're run by engineers. And then artists are subservient to the designers. In EVE's case, it seems like that model's flipped on its head. The artists and their visual design seem to drive a lot of the functionality. In fact, in that same art book, there's a section on user interface entitled, "Style over function," that explains that the interface was subservient to the design.

HP: CCP was founded by Reynir Harðarson who is a graphical genius. I have yet to meet another man who has such a strong ability to create his vision in Photoshop. Essentially, he can create any graphical effect that we use in EVE in Photoshop. When I joined initially as the head of the programming department -- I was hired as the CTO in 2000 -- I very much understood how much the art can drive an engineering team. Because all engineers aspire to enable something beautiful. We had something beautiful and we had to enable it!

So I sat down with Reynir and Torfi [Olafsson] who was our technical artist and saw what they were doing in Photoshop, what they were doing in Maya. And when I ultimately had to end up creating our first 3D engine, I tried to make a tool for them to do similar things.

So, rather than create an engine per se, I more created a tool to replicate what I saw them doing in Photoshop and Maya. And that was really the foundation that created the Trinity graphics engine that we're still running on. Their ability to realize their artistic vision through that tool is really what creates the graphics. The graphics of EVE are not a technical achievement. They are definitely an artistic achievement on behalf of them. I just created something so that they could realize it. And then we have really driven all of our outward facing marketing initiative on the look of the game.

Everybody has respected EVE for being "too good to be true." We had some very... [laughs] strange discussions with publishers initially where nobody believed these were screenshots from the game we were sending over to them. And now that we see the rest of the world catching up, we're now taking the next step where we have basically re-written our 3D engine; we have redone all the content to be a "next-generation" qualifier. And we have also taken advantage [of the fact] that the hardware on the GPU side has evolved quite a bit. So we're doing a fully shader model 3.0 enabled graphic engine which in some cases will run faster than our old fixed function graphic engine that we still run on.

So not only will the graphics... It just looks amazing. It's too good to be true again. I think we'll have to argue with people that these are actual screenshots from the game again. Which we celebrate to have the opportunity to do. But also due to the use of shader model 3.0 and what they have there, I think we have the ability to even run faster than our fixed function pipeline on top-of-the-line computers.

Underneath all of that architecture -- the display, some of the faster computing that's going on server-side behind the scenes -- in rebuilding a lot of that code, have you thought more about scalability this time around?

HP: Definitely. We foresaw massive fleet battles in the beginning. But they have risen to a scale that we didn't really anticipate... to have 500 people in an all-out shootout, all graphically rendered on a PC is something that we couldn't necessarily have built a better strategy for. But our new engine definitely enables that on hardware that is out today. Our old 3D engine is not really able to leverage the GPUs of today and tomorrow to enable those experiences to be completely smooth. But our new engine has the tools in place to allow people to get rid of the client-side latency part of massive fleet combat.

Have you thought about future generations of hardware as well?

HP: 3D engines as you create them today are structured around shaders. And as the graphics systems evolve, you have the ability to do more and more sophisticated shaders. Our engine is structured in a way that it provides facilities to shaders to take advantage of the hardware. So as you get more sophisticated hardware, you can easily upgrade your shaders and get better effects from that without having to change the overall structure of the 3D engine that much.

Most of the more recent 3D engines that you see created today use that approach, and we have very much taken advantage of that and we have people, much more capable than I ever was back in 2002, to really understand the way today to build a future-proof engine so we won't have to go through this massive effort of re-creating again. We can continue to evolve the look of the game over time as the capacity of the hardware allows for it.

You're upgrading portions of the game -- hardware and software -- that work behind the scenes. So basically you have functionality which will just become more efficient. At the same time, you're working on new features -- new technology -- that you're implementing in the engine and networking. And on top of that, you're adding new gameplay features such as the Ambulation project. So how much attention do you put toward something new versus fixing things from the past or becoming aware of things that you might have to deal with in the future?

HP: As the company has grown, it has very much been the challenge of juggling multiple goals. We have been continuously evolving our project management infrastructure. We do a lot of scrum-based sort of agile development initiatives now, and those have helped a lot. Especially around cross-disciplinary initiatives such as Ambulation which has to involve artistic elements, game design elements, and technology elements and also, just our overall vision for the game.

Ultimately the game is about people, for people, and not having people there has always been a challenge to explain, that the game is about people when all that you have to show is space ships. So Ambulation for us is very much... a tool to better explain what EVE is all about. It's about people, and all of our emergent behavior comes from people interacting with each other. So we're putting more of a human face on the game. And also, it just looks amazing... what the guys have been doing representing a person in the future, working around a space station. In my opinion, this is maybe way better than people what people have strived for before. We wanted to do for the character aspect what we have done for the space aspect in terms of feeling real and almost movie-like.