I'm very curious about your use of metrics in light of the discussion of bad metrics versus good metrics. Tracking the wrong thing, or tracking things, and getting the wrong interpretation of what's going on. Have you found there have been a lot of pitfalls with the way you've approached metrics?
AT: Well, I think that's always the case, right? Everything can be done the right way, can be done the wrong way. But how you track things the right way is a difficult question for sure. I think that a lot of the time, you don't necessarily know what you are doing...
I think maybe the way to avoid having bad metrics is to always tie everything at the end -- even though the problem may be very, very complex, at the end, the conclusions, the cause and effects have to be really crystalline clear to people.
If someone cannot understand what you mean after analysis, you haven't done a good enough job. Part of analysis is not to make things complicated. It's to make things simpler and easier to understand for people to make decisions. If it's not simple, if it's not easy, probably your analysis is not right. I think that's probably all I can say.
You discussed how, with Mafia Wars, the team implemented boss battles, and that didn't really increase the metrics you wanted to increase, versus adding a simple lottery system to FarmVille. You made it sound like these mechanics could be repeatable across multiple games.
Do you feel that developing social games is more about defining successful mechanisms and then targeting different audiences with the same mechanism via different ways of presentation?
AT: You should always look at best practice metrics and mechanics. You always see how they can apply. I think it's one of those things where there's no clear-cut answer. Some things can be applied across games, because people are people, right? People are people, and they have the same psychological drivers.
But some things cannot be applied to other games because game types are different. People's expectations of how a game will behave is different. For example, if I start selling guns or tanks in FarmVille, people are going to say, "What the hell is that?" [laughs] That's beyond people's expectations.
But things like daily returning rewards, that can be applied to both Mafia Wars and FarmVille. Again, it's kind of a case-by-case basis.
With every new platform that comes out, it seems that people start trying to define new ways to go with it, and then a couple approaches become successful, and things start to coalesce around that. Do you think that's happening with social games?
AT: I think that's the [current] stage of development, too. So, right now, social gaming seems to be a couple different categories, but still many, many categories and genres have not been "socialized" yet, shall we say? I think what you said, actually literally, may be right.
It was interesting that you talked about that game [Guild of Heroes] that Zynga worked on, the Diablo-ish game that ended up not really panning out. And the concept that you can't really "socialize" a single-player game experience.
AT: I don't want to say you can't really socialize... [but] It's really, really hard. I think it's really, really hard because you can definitely take a lot of mechanics, but it's really hard to take a whole single-player game, add five social mechanics, boom, voila, you have it.
I mean, that's a simplistic view that many people have. But ultimately that will not really work because people have very different expectations. Again, it's about player expectations, what they want to see with the game. Do your future expansions jive with that? And that's one of the problems which makes single-player games hard to work as social games.
FarmVille, ultimately, is a one-player experience, right? You manage your own farm. You interact with other people, but it's not fundamental to the core gameplay interactions. First of all, it's asynchronous. What separates that from a traditional game in terms of a single-player play path?
AT: Traditional console games, you don't have any interaction unless you go into multiplayer battles, during campaigns. The single-player is self-contained. But like you may not necessarily want to be at a party 24 hours a day, I think there's something cool about the pace and tempo of social gaming interactions.
Because it's asynchronous, you don't have the huge pressure you [could] have, "Oh, I must respond to this person," [instead, it's] "Okay, I can respond to them a little later." But that doesn't lessen the social obligation or social capital that you have.
I think it's precisely because it's asynchronous that people feel "I can come back to the game anytime." Because remember, for social games, unlike MMOs -- a synchronous game -- we don't expect, require, or design a game for users to play hours a day. It's like 10 minutes, 15 minutes a day.
I know there have been some surprises, like session duration not necessarily at first being the same as developers anticipated. Through metrics and through application of design, does player behavior now align with what you anticipate, and can you maintain that?
AT: I think after a couple years, you understand more about player behavior through metrics. One way the metrics can definitely help you is to understand how players really behave, how they really think. Another way is you always ask your user to get qualitative feedback.
Quantitative measurement and qualitative feedback need to go hand in hand. Metrics can tell you how they're doing it, but not why are they doing it. Sometimes you may not be able to answer that -- like qualitative feedback from users sometimes doesn't really represent what they actually do.