Triple Town devs on finding the fun in free-to-play game design

What's the key to Spry Fox's (Triple Town, Realm of the Mad God) free-to-play success? In a session from GDC's free-to-play summit today, David Edery, Daniel Cook, and Ryan Williams suggested that devs need to focus on prototyping early and often in order to refine ideas that can yield online games that will last for years.

Finding the fun -- quickly

"Finding the fun, for the successful games that we've made, historically doesn't take very long -- a few weeks, a month, two months tops," said Edery while discussing a failed prototype. "We spent six months banging our heads against a wall even though the warning sides were there. We should have set a timer, but we didn't, because we were so enamored with the [core mechanic]. Falling in love with your ideas can definitely hurt you the most." Cook explained: "We're looking for a core, tight, robust mechanic that we can explore for years and years on end. What we're doing is we're sending ships out into an empty ocean, looking for land. Sometimes you find a little island, and sometimes you find a giant continent of gameplay. We're making games as services, and they have to last for years. What you need is something that people can engage in for hours every week, for years on end, and keep it fresh and changing as it goes. So when we say you have to prototype until is fun, what we want is this rich, robust world of fun." Working with tech and tools that facilitate rapid prototyping is key to their iteration speed, Edery said. "We tend to be highly intolerant of any terms of tech setup that slows down your short-term. We need to be iterating on a daily basis. We don't need to invest in tech that will pay off later on; we've found that we can get those benefits later on when we need it. So we usually work in environments like Unity and Flash. It does not take an amazing engine that you've crafted for ten months to make a game that others can enjoy." Cook identified three rules of thumb that he used to identify potentially successful game ideas: Momentum ("Is the fun increasing from iteration to iteration, month to month?"), size of playspace ("How do we add stuff?"), and the robustness of the fun ("You could spend a week making it horrible on purpose, and it'd still be fun").

Broad game design potential = broad monetization potential

Once you've iterated your way to a promising game concept, you need to see if it's the kind of thing that will actually make you money. Edery used Triple Town as an example: "One of the risks when you use a process like ours is, you'll come up with a game that's really deep and enjoyable, but it doesn't really make any money after people have played for hundreds of hours. This has happened to us two or three times by now. Lots of people make the assumption that [Triple Town] is a masterful financial success, and it's not. Our revenues tend to top out to 4, 5, 6, 7 bucks per person...The joy of the game comes from doing better without paying. Usually, the reason you purchase performance is because you're showing off to other people [in multiplayer]." "In this regard, Triple Town is a shallow experience. It has one thing going for it, and that one thing is not something that makes you want to spend money," Edery said. "If you can come up with three or four categories of things that people might wanna buy, you're probably okay. But with Triple Town, we didn't ask ourselves that question in the beginning." "When we tried to add them to the game later," Cook added, "the design was already so tight and functional that it just didn't make sense. Single player games are our vanity projects now. We don't expect them to make money."

Release early

Edery attributed some of Spry Fox's f2p success with their willingness to release a game earlier in the dev cycle than competing studios would feel comfortable -- almost as an extension of the prototyping process. "You always learn more when you put it in front of a natural population of users, just playing it on their own," Edery said. "We release our games six months to a year earlier than our competitors would. We'll release in Canada, first."

Bring art in late

The rapid prototyping phase is not the right time to become emotionally invested in a concept -- and bringing well-developed art assets in early made it hard to maintain the necessary distance. "Once you've seen the art, it's like a sugar rush," Cook said, "This is what the game's going to be about! But it doesn't work so well. It kind of destroys the whole prototyping process. The artist was making a lot of cool stuff on a regular basis, and we thought we were making a lot of progress. But every day, we had a choice between thinking, 'Ooh, we're doing good because the art looks good, let's talk about how to make that happen,' and the hard, grueling mathematics and abstract structure of the game design. And each time these two butted heads, the art tended to win. Art creates emotional investment, and you don't need that in the prototyping phase. Now we prototype in a crude fashion, and think, if it's fun when it's crude, it'll be more fun with those emotional hooks." For Gamasutra's full GDC 2013 event coverage this week, check out the official GDC 2013 event page.