In 2001 Anarchy Online entered the scene with what may be regarded as the "worst launch ever." By 2003 Lineage was still dominating in the East but the West was still not really paying attention. Around this time the MMO industry was starting to coalesce a body of common wisdom, encapsulated in Gordon Walton's GDC talk "10 reasons you don't want to make an MMO." There was an awareness of the need for huge development teams and large server farms to support these games. Meanwhile, Runescape launched in 2001 having been mostly developed by one person in a short amount of time and the industry still ignored it, pushing on into bigger and bigger blockbuster budgets (leading to Adam's great next slide which read simply: WTF?).
By Q1 2005 World of Warcraft had launched and reached 1 million subscribers, going on to rapidly become the biggest MMO in the world. Now the fear is that WoW is too far ahead and no one can ever catch up. But the question really may now be, according to Martin, whether our goal should be to beat WoW or to beat Facebook. The common wisdom says that "it would take at least $150 million to compete with WoW," but a look at Facebook reveals a platform built by a few students without a publisher and no startup capital. In light of this, are MMOs even a sensible business model anymore?
Martin advises to start with the basics: what would Google
do? What would Amazon
do? What would some of these other huge brands who routinely deal with more than 10 million regular users do? Let's back up even further and ask what we mean when we talk about Web 2.0, Game 2.0, even by what we mean when we say "internet." Kahn and Cerf arrived at a good working definition back in 1999 when they said the internet is "global information services at very low cost." Information service is at the core of the internet, which is inherently global. So what is the internet for, besides porn? (The next slide shows a clip from "The Internet is for Porn") We might say knowledge, but if you look at Myspace it is not at all about knowledge.
Today the question "what is the internet?" is practically synonymous with "what is the web?" Martin says we need to be concerned with this question if we have any hope of understanding where games technology will be in 10 years. The web is a reflection of ourselves: "the web is us." The first wave of individuals pumping data about themselves into the web was crap like Geocities and horrible blinking banners but it has turned into something better. We have stories of individuals and their lives, plus a good smattering of what we think we know (aka "news"). It reflects our hopes and projects, reflects our fears and allows us to have fun.
The next question is then: what drives the web? And equally importantly: how do we make money from it? It seems clear that the web isn't driven by commerce because too much of it is offered for free. It was started by the government but not driven by it since then either. It's also not an oligarchy in which a handful of rich and powerful people control it. Instead it's driven by the need for people: a need for validation, self-expression, competition, interaction, experimentation and even finding a mate.
Web 2.0 is even more expressly about social interaction -- but the key is it's social interaction "with a purpose." So how do people socialize "purposefully"?
- discovery: users need a way to find other people
- charcterization: need a way to learn a bit about those people and find out if you want to play with them or not
- interaction: chat, forums, blog postings, etc.
- build relationships: users need ways to meet those same people again over time and keep interacting with them
- manage: once we've built these relationships we need a way to keep up with them
Games excel at the interaction layer and tend to focus on that. They are also good at the "build relationships" aspect because of the deeper experience and the ability to connect with someone in a way that recognizes common touch points in your backgrounds and interests. The web does extremely well at the discover, characterization and management stages -- and this is exactly the stuff that should be built into more games.
But the problem with saying that relates to the "good, fast, cheap" triangle and the rule that you can only "pick any 2" of these. Budgets are already huge and now we need to add more layers to games -- how do we achieve this? Martin appeals to Guy Kawasaki's famous quote, "don't worry be crappy." When you create a new feature you always have two options: release it now (it will be crappy) or release it later (it will be more polished but will have taken too long and cost too much). Because of the web ethos, users need something now
and they expect to find it. If you're not offering it, they'll go and find it elsewhere -- and what they find may actually be even worse than your crappy untested feature.
So the theory is that we can get rid of the "good" part of the triangle if we can learn something: expect failure and deal with it. The web is built on an explicit assumption of failure so we're equipped and prepared for that model there. The internet is still fundamentally unreliable, but it's very good at working around its own failures. Martin suggests a new mantra: "Embrace the Chaos" -- this requires a lot of letting go. His advice is to have faith that chaos is good for you, despite the lack of obvious guarantees. Whereas traditional programming is built on the principles of ACID (atomic, consistent, independent, durable), the web is built on the principles of BASE as introduced by Google (basically available, soft-state, eventually consistent). ACIDity creates trust but kills innovation, whereas BASE promotes innovation and still allows a certain level of trust. The trust comes in the "eventually consistent" part of the equation in which all your data coalesces -- and provided you can get your vendors to agree to this model you have a non-obvious guarantee of trust and you can successfully sacrifice the "good" point of the triangle. In other words, learn to love the internet at its basic level. Once you embrace the chaos and run with it, a lot of things become possible.
Martin ends with a summary:
- understand the web: the web is us
- don't worry be crappy (development costs upwards of $100 million is just not necessary for all properties)
- drop (some) quality from the triangle
- embrace the chaos
- build BASE not ACID systems
By 2010, Your Game (TM) could then easily eclipse Facebook/Myspace. Question and answer
Q: Isn't the comparison you're making unfair because Facebook/Myspace users aren't paying customers? Most MMOs are paid while these sites are free.
A: Of course they're not free. Those adverts that are blinking away are stealing little parts of your soul (laughter) and that's really in fact what advertisers are paying for.
Same questioner: But advertising can't pay for that forever -- can it pay for my 250 million player game?
A: If your dev costs were small then yes. If it's WoW scale, then no. For something like Runescape
, it's achievable now. But keep in mind that business models won't even be the same 2 or 3 years from now. Advertising will be a part of it but not the same part anymore. It's more valuable to us to focus more on the general guidelines and to work out each quarter what can come next.
Q: Are Facebook/Myspace more destinations or platforms? Do we need to beat them or join them?
A: Myspace became a destination and that became a problem when Facebook came along. Myspace never had any defensible value -- anyone could have recreated it. I think Facebook has that problem as well; it's less a platform than it wants to be. Those folks are looking for ways to bolt on and inject more value into the system; it's so indefensible they have to move away from it. Games are much more defensible but we haven't been very good at getting them out there and getting people to use them.
Q: Isn't it also an unfair comparison in terms of time commitment? MMOs require so much more time and investment to get enjoyment out of them, and we're talking flat web product vs. whole client. Is it even realistic to imagine we could get those kind of numbers? It seems like we're comparing apples and oranges.
A: On a direct financial comparison it's obviously unfair. But the valuable comparison is in the size of the community and the exposure. The larger your footprint, the more business models you have at your disposal that become viable and the easier it is to guarantee you'll get your revenue.
Q: What are your thoughts on sites with game elements like Kongregate
or even eBay
? How does that fit into what you're talking about?
A: I agree absolutely. One of my favorite examples is Counterstrike -- one of the first MMOs. Counterstrike as it was played versus how the developers made it. The way people actually played it turned it into an MMO, and the advent of things like Ventrilo pushed that further -- it's almost like you can't stop it. Players will find whatever random tools they can and collaborate with each other to create the game they want to play out of the components you've given them. eBay is definitely a game -- but not a very good game! You could do much better (laughter).
Q: Isn't there a cultural psychology you're fighting too? The culture of video games is much more hardcore, and the casual consumer is naturally intimidated by that. How do you fight that psychology so casual users are more likely to give MMOs a chance?
A: That makes me think of Richard Bartle's recent Guardian article where he said "get over it, we've won." The oldest people who grew up with video games are now in their 40s. People who think games are nerdy and geeky are diminishing. That problem is receding and so even if we did nothing it will go away. On the other hand I agree we haven't done enough to make the experience easy and enjoyable for players with more usable UIs and so forth.
Q: How do you convince ACID companies to change to BASE?
A: Failure usually works. It generally takes only a few people going in the right direction before other smart people in this industry pay attention and start trying it out. Other people's success and failure can serve as a model too, not just your own.
Same questioner: Are there any good examples of this yet?
A: All the web-based multiplayer games starting to crop up: Dofus, Flyff, Runescape, Habbo, Club Penguin
are all starting to change things.