From swords and sorcery to Playdom

Who was there: Raph Koster, whose frequently changing business card once carried the logo of Sony Online Entertainment (where he worked on Star Wars Galaxies) and has since featured his startup Metaplace, which was acquired by social developer Playdom, which was in turn acquired by Disney.

What they talked about: Koster--who was pinch-hitting for colleague John Donham on short notice--began by talking about the reasons he and Donham jumped ship from SOE in the first place to start Metaplace and go into social gaming. While Koster acknowledged that making money is nice, he said there are also draws to providing entertainment to the largest audience possible, building and strengthening social ties, and making games that the people in his life--wife and kids among them--will play.

The first lesson Koster said he and Donham learned is that with few exceptions, console games are niche. He said 20 PlayStation 3 games have sold more than 1 million copies, while 219 Facebook games had 1 million users in the month of August alone. As a result, traditional console game developers have been making massively polished games for "almost nobody."

"All of your assumptions about--Hey I know my audience and I know what people find fun--need to be reevaluated," Koster said.

Koster started with assumptions about Facebook games--namely, that they're shallow and simple and only played for five minutes at a time. Koster said that more than half of Facebook gamers play for more than 30 minutes a session, with an average of more than one session per day. Some 35 percent will purchase virtual goods at some point, Koster said, and more than 25 percent spend more time social gaming than watching TV.

"These people are not casual," Koster said. "They're hardcore."

While Koster said the assumptions about Facebook games are largely wrong, they get one thing right: The audience for Facebook games is different from the audience for traditional games.

One of the big bonuses for Facebook games is how easily and quickly they could shift concepts. Koster took a swipe at BioWare for the five years it spent on Star Wars: The Old Republic without having released anything and contrasted it with the way Metaplace determined its projects. Koster said he took each of the company's game ideas and made $150 Facebook ads for each. Based on the click-through response and the demographics of those people, Koster could determine which games the company would work on.

"It's much healthier to have your dreams crushed after 48 hours rather than five years," Koster said.

As a designer, Koster wasn't thrilled with the idea of using demographics to determine gameplay because it ran counter to his designer instincts. But, eventually, he got over his reluctance and found it to be a boon.

"Metrics are just a way to play-test faster, quicker, and with more people," Koster said.

Koster touted a site called Usertesting.com that would run monitored focus tests for the company and allow for quick iteration. Koster said if they wanted to know whether or not people wanted a certain feature, they would put a button for it in the focus test version of the game. Even though the button didn't do anything, if lots of people clicked on the button, they knew it was a feature worth actually developing. The team even had Usertesting.com run play testing on competitor's titles, just so they could see what features they had that worked best with players.

Another shift traditional developers need to get over is the idea that instead of creating an online game now so that it can make money in future years, social gaming is a business that makes it possible to make money on from day one. The key traits to focus on for social games are "virality" (a measure of what percentage of users go on to invite other players to the game), retention, and average revenue per user. If developers can figure out how many new users they can add each week, how long they'll stick around, and how much they'll spend on average, it's easy to figure out how much money they can expect the game to bring in and plan accordingly.

Once a developer has those numbers, it's possible to tweak the numbers and figure out what brings in the most money overall. For instance, it's easier to run a sale than redesign a game to be more addictive, Koster said, though they might have similar impact on the ultimate revenue.

Another lesson Koster said was learned on the way to social game development was that specialization is bad. He said if people go into social games with every person filling just one role, they will fail. AAA game developers who may be the most talented level designers aren't as valuable to a social developer as an average level designer who has also managed a retail store and has some understanding of how to bump up revenue.

"For the first time in quite a while in a large portion of the game industry, specialization is bad," Koster said.

After eight years of Everquest, Koster said SOE had added a ton of features. But there was no way to know which ones added to retention rates.

"There are probably features in EQ that should have been removed years ago not because they were bad, but because the environment had changed."

For Metaplace's My Vineyard, Koster put in achievements and found that people were actually recommending the game to their friends slightly less. It took seven iterations of achievements before they started to see any appreciable improvement on the game's virality. Koster said the team kills two-thirds of the features it creates every week as a result of this testing, and that's a significant improvement over the way it worked before.

Koster also talked about how every lost user is an opportunity to learn, and it's better to tick off and lose 5,000 users because it will better serve the 1 million-plus other users. That was another hard lesson, as Koster recalled looking at a subscriber number as the end-all, be-all during his days with massively multiplayer online games at SOE and jumping into the forums at a moment's notice to retain even one unhappy player.

Although he relies on metrics as a guide, Koster cautioned developers in the audience not to be overly reliant on the numbers.

"Metrics are not a cure all," Koster said. "They are for optimizing. If you are in the vicinity of something good, they will take you to that something good. The metrics are awesome, but you still have to be just as creative and awesome a designer as you ever had to be."

The single biggest thing that AAA people have trouble getting a grasp on is that Facebook wall post spam is not spam to the people who play the game. The players have real emotional, social investment in what people derisively call spam. People show up every day "because they know their friends are counting on them to water their cabbage," Koster said.

Koster stressed for the audience that it's no longer enough to be a creative developer. Those who want to succeed in games also need to understand the business side of things and they need to be able to keep up with the constantly changing landscape. Flight sims and other genres exploded and became niche in five-year cycles, Koster said, but genre cycles on Facebook are turning over in a matter of months.

Quotes/Takeaways: "We're very used to treating this as a creative endeavor… the biggest difference here is that much more so even than in service-oriented MMOGs, you're running a business concern on the first day. You can no longer be a designer who doesn't understand money."--Koster on the need for social developers to wear multiple hats.

"The biggest thing involved in making this leap is mental adjustment and learning new habits. I can tell you it's not like this is hugely different, but those differences matter a lot."--Koster, summarizing his presentation.