Arden: a Tragedy of Errors

disclaimer: I want to make absolutely clear that I am not suggesting Multiverse is a bad platform. In fact, the whole goal of my article was to show how Arden *misrepresents* what Multiverse and virtual world technology can do.

Shakespeare MMOG: No-Brainer

So let us say you are a grant officer at the MacArthur Foundation and you are approached by a well-reviewed professor who is renown for being the preeminent "Expert of Virtual Economies." This professor, Dr. Edward Castronova, wants a grant to build a Shakespearian Massive Multiplayer Online Game (MMOG), in which players could experience a Shakespearian world. Through players immersing themselves in the environment, and with a fairly accurate depiction of the time-period, people could get a much more vivid, clear picture of the world in which Shakespeare wrote.

It's not surprising that MacArthur awarded him $240,000. Sounds like a no-brainer, right? Seems like a great idea with project leader with plenty of credentials; at worst it'd be short and cute, at best it might inspire a larger project and lots of attention. That's what I thought when I first heard about the Arden project over a year ago.

For his environment, Castronova initially selected Multiverse, a MMOG-kit designed for small developer studios. Multiverse, for those not familiar with the platform, states its purpose as being "designed to change the economics of virtual world development by providing independent game developers with the resources they need to enter and compete in the $2 billion online game market." Multiverse was - and still is - in a Beta development phase, less than three years out of the box.

However, after months of toiling with Multiverse, the Arden project was scrapped on Multiverse. Arden instead was Beta-released as a patch to Bioware's aging Neverwinter Nights MMOG, after recreating most of the artwork, whose format in Multiverse was incompatible with Neverwinter Nights, according to the Arden development team. Soon after, the project was terminated entirely.

"It's not fun," Castronova explained. "We failed to design a gripping experience."

There's no doubt that the end-result was a failure. What puzzles me, as I learned details about how the game was produced, is why anyone would have expected it to succeed in the first place.

The Larger Picture: Education and Virtual World Development

Allow me to add some perspective. How would this failure look to a potential investor in virtual worlds, or an educator looking to enter the MMO space? How would it look to a development company such as my own, Involve, Inc, regarding the initial platform, Multiverse, or other potential philanthropic organizations like MacArthur?

One might easily infer that the Multiverse platform is not as capable as it was promoted to be, although a developer (like me) looking at the situation would quickly learn that Arden failed because of a mismatch of expectations, not because the Multiverse platform was inherently flawed.

But for a potential investor or educator with little virtual world experience, there might be a whole different implication. If some "genius" professor (Wired Magazine's word), a purported "expert in virtual economies" can't make an educational experience for a quarter of a million dollars, why bother with virtual worlds at all?

Given the high stakes at play here, I will say it: This is not about the failure of one singular project in virtual worlds. Instead, as a well-funded, well-publicized project, Arden's failure can cast doubts over the entire virtual world industry.

There are critically important lessons to be learned by examining what happened with Arden; when we do, we learn that the mistakes made were human, not technological. The nature of the failure revolved around the project management, not the essential viability of virtual spaces, or Multiverse as a platform. Most important, we see that the mistakes made were predictable and avoidable. Any professional development team could have seen this train wreck coming miles down the track.

Let’s explore why.

The Tragedy of Errors

An interview with Arden's lead artist, covered at the MMO blog Channel Massive reveals a glut of odd decisions made about the project. Now, let me be clear - this developer was a full time student and I would not expect him to devote anywhere near the time nor experience that a professional lead project developer for a game development studio could provide. From what I read, he did a good job with the situation he was in, but it is a crying shame that he was put in such an awful position to begin with.

The nature of these odd decisions is further illustrated by an interview with Castronova in the current edition of Wired Magazine. In it, Castronova shares a series of five tips that are absolutely indisputable, but an absolute "DUH!" for anyone who has ever developed a real game.

Since I am a technologist by inclination and profession, let’s get to the gritty stuff.

The Fool's Errors

(All quotes from the aforementioned, wonderfully frank, Massive interview)

1. Their lead artist, who would become their lead developer, was a full-time college student. Gee, perhaps a major project like designing and constructing a MMORPG deserves someone devoting his or her entire attention to it?

2. Not only was the lead a full-time student, but the plan was that the staff would rotate every year. What kind of game development plan includes rotating the staff yearly? How could anyone guarantee even a modicum of continuity?

3. "In our first year of development we had planned to create a working prototype of the game with around 80 hours of content," said the lead developer. Considering that most video games, with multi-million dollar budgets, have something like 10 hours of content, Arden’s base expectation of how much work was needed was off by a factor of at least 100 times.

Let me say it again: The estimate of work that they hoped to do was more than 100 times what the budget would have allowed.

4. "An element that I particularly enjoyed was how we had planned to record every line of NPC dialogue. I'm sure that any experienced MMO designer would laugh at this idea and tell me how unreasonable it is." Yes.

5. "The big turning point ... The whole team was already stressed because we had switched engines and this had a significant impact on the every facet of the project. We pulled an all-nighter, which for some ended up being two all-nighters back to back without any sleep." Most game developer professionals could tell you that, on occasion, these happen. If it couldn’t have been avoided, a professional project manager should have seen to the care and feeding of the staffers working in crisis mode. Without that level of management: "Our nerves were absolutely fried and anytime someone left to take a break, the rest of us would question their devotion to life in general let alone the project. ‘Is so-and-so serious? They must not really care about Arden at all.’ The 40-hour hell march killed our spirits."

6. "Network stability should have been the last thing we had to worry about – we were still trying to complete the game! The server would crash every 80 minutes." Ensuring network stability should have been a top priority. The interview mentions no further detail into why the game was so unstable, so I can only conclude that they did not ever find the root of the problem. Rather than complaining and giving up, a professional team would have been able to trace down the issue and provide a specific explanation.

7. "Departure from Arden was somewhat staggered. Several team members – mainly the volunteers – more than likely left in frustration." ‘Nuff said.

8. "A fundamental approach to the MMO genre would have really helped us. It was obvious that we weren't going to be making the next big thing; however, we were really trying to create an engaging and immersive MMORPG. This was unrealistic and ultimately part of our downfall. If I could go back to that time, I would suggest scaling down the project and really focusing on a few, key gameplay mechanics. For example, a primitive crafting system coupled with a small-scale PvP arena." Yup.

9. "If the technology is not working and the team is inexperienced, better technology is needed. Immediately." To me, this reads as a slight to Multiverse. Whether Multiverse failed was the chief question in my mind since hearing about Arden's failure last year. As a virtual world developer working across multiple platforms, I wanted to know if Multiverse was inherently flawed.

Last week, I had the chance to find out, when Castronova spoke at the Life 2.0 conference.

10. At Life 2.0, I asked Castronova directly why Multiverse didn't work for Arden. Castronova said that he wasn't a game developer, that he spent five months working with Arden in Multiverse before he realized the platform didn't have sufficient built-in content, such as medieval style houses, etc.

He expected content to be there, and it wasn’t. And to add insult to injury, it took five months figuring this out.

Why Didn't Castronova Know?

In all my close research of Multiverse, I have never seen any promises that independent game developers would be able to build a game without having to provide their own content. Did somebody at Multiverse mislead Castronova? Did Castronova consult any professional game developers at all?

Or, what seems more likely to be the case, did the professor just fail to do his own homework?

"I hadn't realized Multiverse was so new", Castronova said at Life 2.0.

Seriously, I'm not kidding. This is a direct quote.

I am not sure how he could have missed this crucial point. How does one choose a platform to create an MMOG and secure a quarter-million-dollar grant to do so without researching the progress of the technology and what it is capabilities it offers? Did Castronova miss the barrage of press releases and buzz that was printed from a variety of sources describing Multiverse's Beta opening at the end of October 2006?

Conclusion: More Harm Than Good

Arden's blunders have tarnished Multiverse's reputation and also raise questions about the viability of education in virtual worlds. In fact, I would go so far as to say that Arden's failure has done more to hurt than help education in virtual worlds. This borders on disaster for the industry, and Castronova's "Mea Culpa"just isn't sufficient.

Of course, Wired Magazine (still stinging from its own poorly designed, executed and reviewed foray into virtual space) dumps its usual pooh-pooh on virtual worlds through the article's headline. The headline on the print article is "Playing the Fool: Trying to Design a Truly Entertaining Game Can Defeat Even a Certified Genius." This read to me as, "Gee, if a genius can't make a good educational MMOG, no one can!" That is the exact mindset plaguing pundits of the virtual world industry that we as developers need to eschew. The errors made in this project would have been obvious and avoidable for any professional game or virtual world designer.

In any event, after the termination of Arden, Castronova is talking about the insight gained from this “research project.” But as research, it failed to meet even basic criteria. The knowledge gained was already commonly available. The MacArthur Foundation, I assume, did not intend to not spend $240,000 to train a professor and a few students on what not to do in interactive game design. I mean seriously, for that price, you could have paid for the entire undergraduate education of several students (and there are any number of schools doing really superior work in game design and technology).

And I wonder: Does Castronova's lack of prior knowledge about these issues indicate a larger issue with educators being out-of-touch with the industry? I’m hoping it’s an exception, as I have seen plenty of educators within virtual worlds doing much better jobs with much less money.

***

Takeaway message: They should have known.

Game designers and software engineers have understood and thoroughly exploited these design and programming technologies for more than a decade. This is why it is important to use professional developers, people who have been doing this sort of thing for awhile. With major demonstration projects like this, there is so much potentially at stake for the industry; it's a shame to see a great idea that had an opportunity to shine instead falter and collapse due to lack of sensible decision-making, and then have the technology blamed for the failure.

Rosedale Quits CEO, Kapor's Vision, The Problem with Second Life, and The Future

Kapor and Rosedale: Visionaries

If you live under a rock: Philip Rosedale was the founder of Linden Lab, creator of Second Life, the first popular online 3-D virtual metaworld. Before Second Life, Rosedale made a bunch of money coding some streaming stuff for Real. Mitch Kapor founded Lotus back in the 80s, whose software was considered the first "killer app" and quite literally aided the world into using computers for widespread commercial business use. Kapor and Rosedale both changed roles at Linden Lab recently; Philip stepped down as CEO to become Chairman, and Kapor stepped aside as Chairman to remain a board member.

To understand the ramifications of these moves and what these two fine innovators are going to be doing for virtual worlds, let me juxtapose a third individual, a virtual world power-user named Eric Rice.

Spin, The Gadget Hound

Last night on Twitter I had a fun back and forth with blogger, writer, and gadget-hound Eric Rice (SL: Spin Martin) regarding his frustration with Second Life. Namely, Eric has multiple sims, has experienced bad performance, and recently can't even log onto Second Life Mac client with the latest patch. He's so frustrated because he invests hundreds of dollars into Second Life every month and basic things like logging in, walking around, or being billed on time are flubbed.

Each represent a different failing of Linden Lab. The first being a lack of proper resources in QA to test Macs, the third being Linden Lab lacking the proper infrastructure and/or staff in customer service, and the third being the technology. Spin is somewhat of a developer - he builds, he creates a great deal of I.P. content, though he's not a programmer. He's used to spending money on things and having them work.

Now, I empathize with Eric's frustrations, even though I called him out on Twitter last night.

Some of the things he complains about is inexcusable. I count my lucky stars that Linden Lab hasn't messed up billing for my company, because I've heard too many horror stories of messed up billing locking up accounts. I came close when Linden Lab failed to bill my AMEX card because they had outsourced billing to Ireland and AMEX had flagged that as suspicious. But that's a pittance of a problem when compared with, say, Anshe Chung, the poster child of entrepreneurial success in Second Life, being locked out because the timing of when Linden Lab billed simulators was out of order and locked her account.

Other issues raised by Eric Rice are less blatant. Why doesn't the current release work on his Mac? Why does he have problems doing "basic" tasks like "walking" on his PC, which he describes as a "God Machine" he purchased recently? Mac releases have always been somewhat behind the PC version, though, on the flip side, sometimes a release will come out and break the PC one, but Linux, for example, may still work. For the record, I told Eric to call concierge support.

No, Seriously, It's Not a Game, Right?

And then there are issues that Eric just is not fully appreciating. From Twitter last night, "This 'not a game' shit, has to stop. STOP. Because that's probably the whole architectural problem. I have just HAD it." and then soon after, " BTW, I can run fucking CRYSIS flawlessly. CRYSIS. Explain that one to me."

And I did explain it, of course. Crysis is a game with static content made entirely by professional developers and QA'd to death for any system. Microsoft's recent "Games for Windows" effort has included a huge push to ensure Windows games *just plain work*. The game, like any other major video game release, has a Hollywood-sized tens of millions of dollar budget. On top of that, you don't have pesky things like:
- User generated content driven by an in-world building engine
- Scenery that changes every day
- Tens of thousands of users occupying the same space

And so on and so forth, but at the same time, Eric represents the enormous percentage of users that *aren't* sticking with Second Life - people that either don't make it past orientation, don't see the value in Second Life and don't settle down, or just get frustrated and leave. So why aren't they staying?

Linden Lab is to Blame for the Game Perception

Gee, maybe it's because Linden Lab's whole marketing push from Day 1 has been to make it look like a game?

The evidence:
1. The name "Second Life". The screams out MMORPG. I've always thought the name stinks and have not been afraid to tell Linden Lab that to their face. While the metaphor is apt, Rosedale referring to Second Life as the ultimate Lego kit during speeches doesn't exactly scream "web platform!", either.
2. The website. Second Life has always had a game-like website. It's geared toward users having fun, exploring, socializing, etc. There's never been a business-driven site, though I and other have suggested this over and over.
3. The interface. It's a video game interface. You use WASD or arrow keys to move around, there's a 2-step pie menu for doing actions, buttons are arranged at the bottom like a video game HUD.
4. The lack of integration with the web. This is perhaps the saddest to me, since decade-plus old ActiveWorlds had this as a basic feature.
5. It's been marketed at gaming conventions, Linden Labbers have spoken at Serious Games Summit, etc.
6. There is a lack of ability to "be yourself" rather than a fictional avatar. Rosedale is definitely an Immersionist.

Mitch Kapor Explains The Universe
(at least, the metaverse)

The overview of his speech from my notes, for my readers to follow along:
1. SL in the DOS era
2. Kapor's epiphany
3. Business use in a world dominated by global warming
4. 3-D camera tech and new interface
5. Why virtual worlds have taken so long
6. Comments on Rosedale stepping down

1. SL in the DOS era

Kapor, being a veteran of software, the Internet, and personally overseeing Second Life as its til-a-few-days-ago Chairman of Linden Lab, sees the technology driving Second Life (and metaworlds) as analogous to Personal Computing and the DOS era. DOS was not a friendly interface, and even when Windows came along, it was still a far cry from the way we use computers today. At the same time, DOS represented an interface that had come a long way in and of itself.

2. Kapor's Epiphany

Kapor was watching Suzanne Vega perform in Second Life (a show produced by my company, incidentally) and realized that Second Life was no longer an island to itself; he had seen two worlds overlapping and working together. Of course, Second Life faced (and faces still) a glut of challenges to make the worlds working together more seamlessly. Its these challenges that Kapor focused his talk.

3. Global Warming and Business Consequences

Kapor is a very far-looking individual. He related that global warming will have a serious effect on the way business operates. He specifically noted that there are no near-future alternative air-travel propulsion alternatives, so businesses will have more and more reason to meet in virtual worlds for meaningful discussion. "Virtual worlds will work themselves into the fabric of daily life" Kapor predicts.

But to get there, we look at the crazy, steep-learning-curved virtual world interfaces. For example, Linden Lab's own research states that the learning curve of Second Life is approximately 20 hours; this is basically the time shown of hours logged on where users generally stay after completing that mark. Before 20 hours, they are much more likely to not continue to use Second Life. 20 hours is a huge learning curve for an Internet interface. So why is Second Life so difficult to learn? Well, my analysis:

- Lots of features
- Game-like interface unfamiliar to non-gamers
- Orientation area is crowded, confusing, and outdated
- Users are inundated with so many things to learn all at once, rather than using a more proven method of introducing them over time or in discreet, isolated steps

Kapor continues with his speech, focusing on interface innovations.

4. 3-D Cameras

I've been talking about using 2-D cameras to control avatar facial expressions for years. However, Kapor's recent research is in 3-D cameras.



Pictured above Mitch Kapor (MitchK Linden, bottom right corner) shows a slide about his 3-D camera research at Kapor Enterprises. On the left on the slide is an avatar flying; on the right is a human raises his arms to activate flying mode. Kapor indicated that they would have a demo for YouTube within days / weeks.

He noted that 2-D cameras did some neat things, but weren't sophisticated or nuanced enough to be an intuitive enough control as can be done with the added dimension. He described his prototype set of controls as owing to the Segway human transport, leaning forward to move forward, etc.

How teleporting between places will work, or clicking? Maybe an air mouse like Gyration's?

5. Why Virtual Worlds Took So Long

Virtual Worlds have been around on computers since the mid-80s. However, Kapor reasons that technology has been the main limiting factor, between video cards, processing power, broadband, and the variety of other technical hardware obstacles. So the 3-D camera initiative is at a time Kapor considers just reaching viability. And, FYI - he predicts these 3-D cameras will become cheap and available for mass retail by Christmas season 2009.

6. And So, Philip Steps Down as CEO

The inevitable question was asked, and Kapor replied with a very similar response as the official press release. Well, obviously.

In Rosedale's own words. (full blog entry)

In the words of my company, Involve:
"Philip created a visionary platform that has changed forever the way we think about how individuals and organizations will use the internet. He demonstrated beyond the shadow of a doubt that it is possible to deliver shared experiences in a robust, immersive social networking environment. That’s a huge contribution, not just to technology, but to society."

To paraphrase Kapor from his Life 2.0 speech Wednesday:

Philip Rosedale switching to a Chairman role will "make the company a lot stronger" and be able to face competitor challenges.

And this is where I am in deep agreement with Mr. Kapor.

Rosedale leaving CEO and staying as a visionary role has a strong potential, provided the new CEO isn't a suit-without-a-tech-brain, on taking Second Life from the Beta-ware 1994 Internet type system it is now into a viable, ubiquitous platform. We should applaud Rosedale for being able to realize that Second Life has come a long, long way, and while he and his staff deserves huge credit for how far it has come, a fresh set of eyes and hands needs to examine and mold Second Life.

It boils down to a matter of perspective. Rosedale has been working on Second Life for over half a decade with code, and really most of his life in his head. He has a set of assumptions and biases that are intrinsically unescapable. While I'm sure he spends a great deal of time examining and questioning his own perspective, I'm sure he also realizes that at some point, he'll always be Philip Rosedale and an outside perspective helps tremendously. That's why companies have advisory boards anyway, right?

So, time to look at virtual worlds with a new interface. This emphasizes why it's so important for competing virtual worlds to be out there and offer different feature sets. And while Linden Lab will certainly take a bit of a new direction with its managing, I still have to wonder - when is Second Life going to really play nice with other virtual worlds? With the Internet, even?