This is an archived post. Information in this may be out of date. More info…

ARCHIVES.estonbond.com

© 2005-2009

On Spymaster’s virality

What started as a crazy idea I had because I’m now not-so-secretly-obsessed with superspy films and the luxury aesthetics of the Craig-era James Bond is currently battling Google Wave for the top spot on Twitter’s trending topics. By just about every metric I know, Spymaster, a game I built with my colleagues at iList, has become a wild overnight success.

As far as I know, we’ve been the pioneer of using Twitter as a social gaming platform at this scale. With being a high-profile pioneer of an application of a social technology, it’s expected that you’re going to get both the wild support and the criticisms of your technology’s detractors. While the huge majority of the messages I have received about Spymaster have been insanely positive, with people explaining how much fun this little creation has given them, of course I’ve had my detractors. I’ve been called the man that ruined Twitter, a “social media f–ktard,” and some other, more profanity-ridden things from those who feel that the rest of the Spymaster team and I have completely destroyed the user experience of Twitter and everything they once found good about it. It all comes down to the notification system that tweets your given Spymaster actions, such as this one from ReadWriteWeb’s Jolie O’Dell. Some backlash is to be expected, and I’m not all that worried about it. Former Digg Architect Joe Stump was the first person to really start it amongst the digerati, which eventually led to MG Siegler’s backlash post on TechCrunch. Like Joe said, though, it was bound to happen. And Joe, if you’d like to hate anyone in particular for Spymaster’s notification system, you can hate me. I came up with the idea to incentivise people in-game. But please, if there’s one thing I could say right now to you and the Twittersphere at large: don’t shoot the messenger.

Uncovering a flaw in Twitter

As Joe very respectfully said — and I completely agree with him — Twitter is in dire need of a way to filter content from followers. Even in the complete absence of social games on Twitter, this is something I’ve been dying to do anyway, and it’s not an issue of spam: it’s that much of Twitter’s social hierarchy, no matter how asymmetric it is in practise, really isn’t in a real-world scenario. Follower reciprocity is basically psychologically required by Western culture in most of the cases where you actually know the person. At the risk of possibly outraging nearly everyone I follow, there are few people on my followers’ list that really just don’t say anything all that relevant to me. Having filters would also help with those issues where you still feel like you’re supposed to follow your ex on Twitter, but you don’t really want that real-time reminder of your breakup.

Previous to working with the iList guys, I was a Product Designer at Facebook, and Feed filters found their way into the product shortly after the launch of News Feed. Thankfully for me, they’ve saved plenty of frustration (or depression) from the psychologically-obligated reciprocity of people I haven’t talked to since middle school (or exes.) As Twitter matures as a platform, failwhales notwithstanding, we’re watching this happen as the demographic broaden. As much as many wish it still were, Twitter is no longer the playground of the digerati, something I’ve had to painfully recognise as former Twitter superpowers such as Kevin or Veronica fall to the likes of Diddy or Ashton Kutcher. With that, we’ve found Twitter’s trending topics go from news-junkie subjects and obscure technology releases to current network memes. It’s the byproduct of the growth of the network as both a messaging service and a platform.

Yes, we’ve found a flaw in Twitter’s user experience, and I fully trust the guys over there such as Doug and Vitor to fix it. Spymaster has caught Twitter by surprise, Stump by surprise, and probably myself by surprise most of all in how quickly it’s spread, but none of us mean any harm in it. The one thing I do take offense to, though, is being called out as being spammy because of the fault of an interface I have no control over.

I’m not trying to point fingers. I’m trying to explain my perspective: as an interaction designer, I am trying to maximise the user experience of those playing Spymaster. That’s what my job is and what I’m paid for, and it’s what I can control. I want to create a compelling social game, and with anything social you need to get your friends involved. Somewhere in the product process, I laid down a basically we-do-it-or-I-quit directive on what we checked by default and chose the three I thought were the most publicly relevant (and least noisy) of actions Spymaster can generate on the public timeline: assassinations, wiring money to users, and spymaster level increases. Why? Assassinations are something that people want to brag about anyway: other friends that are playing Spymaster find it funny if you’re ganging up on a friend in your social circle (and I’ve seen this happen firsthand quite a few times.) Wiring money rarely happens because it’s so hard to get a Swiss Bank account unless you’re at a really high level, and level increases happen rarely enough as you progress but give other Spymaster players (in the world at large) a good strategic count of how you’re progressing. These were things I thought I’d find the most strategically optimal to be able to know about in near real-time from other players in my spy ring (and my opponents’ spy rings, for that matter.)

Let’s talk about spam, baby

I watched Facebook explode from the inside during its days as a nascent platform, and I fully remember applications that were wildly spammy; no lecturing is required to make me remember things like Zombies/Vampires/SARS and their variants (which I still despise, because they have absolutely zero entertainment value internally.) At Facebook, we worked really hard to combat spam whilst still maintaining both a development environment that allows the platform to thrive at the creation level and people to have fun at gaming in the social environment. Of course, we ran into a lot of problems, and we learned a lot of things that I think Facebook still wishes they could take back: users were forced into the invite process. Users published tons of hidden notifications without their consent to their friends. In the end, a lot of Platform players felt completely violated. I’ve learned what players don’t like, and I’ve actively avoided those things in optimising for the player. If we want to talk about a real flaw Twitter is going to have when they move forward that Facebook dodged masterfully, it’s that their open platform has absolutely no way to police this type of behaviour. I’m sure someone else with a truly evil and enterprising intent will come along and do something like this and make Spymaster’s day of being a trending topic make what is currently the Sean Connery of Twitter games look like David Niven. I am a well-intentioned messenger of something that could really be bad. If I wanted to drop a viral nuclear weapon, I could have. I have no desire to do that, because such work is anathaema to my very existence as an interaction designer.

Furthermore, we are limiting Spymaster’s virality even further because we were afraid of it being wildly noisy. Things like tasks, which are rapid-fire actions in the game’s UI to progress quickly, we quickly found to dominate public streams. Spymaster public notifications are throttled internally, something which Siegler himself noticed on a Black Market shopping spree.

I agree with Joe that this was all inevitable, and honestly I’m trying to set in place best practices for that community before it gets bigger. If Spymaster can set the UX precedent, I’ll be happy. If I was worried about nothing but virality, I would force people through invite processes before I let them touch the game. I would send out tons of public notifications without a user’s consent. I would do all sorts of things that are remarkably self-serving to the application that I have explicitly chosen not to do on the grounds of integrity over popularity, and I am thankful that the application has become as powerful as it has without a coercive user experience. From those in-game, I haven’t heard a single complaint about Spymaster’s virality. I’ve done all of the things I know to make the virality of the application completely user-controlled. I learned one more thing from Facebook, which is granular control, and it’s something I took with me on Spymaster. Nothing on Spymaster is sent without a user’s direct consent. As far as I’m given the liberty to call the shots on Spymaster’s user experience, I won’t let that behaviour happen, either.

One last thing: let’s get something straight about Spymaster strategy

Also, one thing I need to clarify before it gets too large is that although Spymaster does incentivise you for your notifications (and it is an intentional in-game strategic move,) it is being seriously shortsighted about Spymaster’s models and gameplay to assume that is the only way you can excel at Spymaster. The 8% money increase you get from all of Spymaster’s notifications is a little over half you’d get from choosing to simply be American CIA out of the gate. It’s nothing compared to the increase you’ll get from even the cheapest Safe House or doing the lowest of tasks. There are so many ways to play Spymaster that the heavy notification route is one of them, your friends be damned. I was actually surprised at the psychological effect of this; some people have outwardly defended their notification-increasing actions as part of the game’s strategy. If that’s the way people want to play, I’m not going to stop them; they’re having fun and it’s the player’s user experience I am controlling. They can alienate themselves if they feel it is worth the reward.

I won’t make these decisions for Spymaster players. That’s a mission they choose to accept, self-destructive or not. As long as my players are happy, I’m doing my job.

Comments?

Comments are closed on archived posts.