Making Great Games

FarmVille: What Went Right and What Went Wrong

Excerpted from Making Great Games: An Insider's Guide to Designing and Developing the World's Greatest Video Games (Focal Press)

By Michael Thornton Wyman

Read more Insight articles
Mediabistro On Demand photography and design video tutorials. Request your two-week free trial.

FarmVille is nothing short of a phenomenon. Estimates have suggested that greater than 1% of the population of the entire world are playing FarmVille. Essentially a stripped-down, realtime farm simulation game, FarmVille was created by the San Francisco, California-based social games developer and publisher Zynga, and launched in June 2009 as an application on the social networking website Facebook. Microsoft’s MSN Games launched FarmVille on its own site on February 4, 2010. As of this writing (spring 2010), more than 20% of all Facebook users play Farm- Ville, making it the most popular application on the service, with over 88 million monthly active users. The Facebook FarmVille fan page boasts over 22 million fans.

In FarmVille, players manage a virtual farm by planting, growing, and harvesting virtual crops and other plants and raising livestock. After creating an avatar, players receive a basic plot of land on which they can begin planting their crops. Various crops have varying costs as well as growth times, which take place in real time. If not harvested when ripe, crops will be ruined, a mechanism that entices players to regularly visit and tend their farms. Players work to amass virtual currency, which they can use to buy more land, seeds, livestock, and a host of decorative elements for their farms. Players also earn bragging rights through the accumulation of ribbons earned for various achievements in the game.

I had an opportunity to sit down with Mark Skaggs, Vice President of Product Development for Zynga and Creative Director on FarmVille, to chat about what worked well and what he and his team gleaned from the process of creating Facebook’s most popular application. “Our original idea was to make a real time strategy (RTS) game for Facebook,” recalls Skaggs. “But after spending about a month exploring the idea, we realized that the Flash development platform couldn’t support all of the things we wanted to do. We knew we had to simplify our goals, and then it occurred to us that we could simply focus on the first part of a traditional RTS game — the resource accumulation and management aspects. That’s how we arrived at the idea of doing a farm game. When we saw some other farm games out there that people were playing pretty regularly, we knew there were at least some players enjoying this kind of thing. So, we set out to make the best farm game that we could make.”

FarmVille allows players to tend and grow crops on their own plot of land e eventually raising livestock and harvesting trees in addition to their basic crops.

Data Points
Developer: Zynga
Publisher: Zynga
Release date: June 19, 2009
Release platform(s): Facebook
Development engine(s) used: Flash, YoVille Avatar Creator, Zynga internal statistics server, internal monetization page, etc.
Game development timeline: ~ 6 months to initial launch; ongoing with updates and refinements
Development team size: 10 (on core team)
Development budget: $40 million
Awards, honors, sales thresholds, etc.: FarmVille has garnered multiple significant awards, including 2010 Game Developer’s Conference Best New Social/Online Game Award; Academy of Interactive Arts and Sciences Social Game of the Year Award; DICE summit’s Social Networking Game of the Year Award
Through FarmVille, and special virtual crop sales designated for this purpose, Zynga was one of the largest contributors to earthquake relief in Haiti, contributing more than $2 million to this particular cause.

What Went Right

Small, Experienced Team
Skaggs points to the small, highly skilled, and highly focused team that Zynga put together as a key success factor in creating FarmVille. “We basically acquired a team to work on FarmVille with us. In doing so, we got not only a great team of talented individuals and all their experience, but also the technology base to work from.” Zynga acquired a company called MyMiniLife, who had been working on a decorate-your-house type simulation for another social network. “At the time, there simply weren’t that many people with social games experience. With their code base and their talent, we were able to really hit the ground running and be super-efficient with a relatively small team.”

Keeping their team small allowed Zynga to take advantage of something almost opposite to an economy of scale — it felt like scaling down enabled the team to tighten feedback loops and keep the entire process hyper-efficient. “FarmVille was actually a lot smaller than some of the typical products many of our team had worked on, so we could take advantage of that smaller size with our shared knowledge of the game development process. One small team, with one simple goal: make the best farm game we could. Because everyone had games experience, we could speak in a kind of shorthand with each other and get stuff done quickly.”

Keeping the core development team small allowed them to alleviate, or avoid altogether, some of the issues that we have learned about plaguing larger teams. This stands to reason, as many of the core issues boil down to breakdowns in communication, and with a smaller team, effective communication is naturally much easier.

As Little Reinvention as Possible
Another successful practice that increased efficiency for the FarmVille team was the mantra to not reinvent any wheels as they built out the game. “We had multiple teams working on multiple titles at our company, and we simply grabbed pieces of code from other teams in the interest of shipping the game quickly,” began Skaggs. “For example, our avatar creator was borrowed from YoVille. We’ve got pieces from the Mafia Wars team, the Poker team. Behind the scenes, a lot of the server code — transaction infrastructure, those kinds of technologies e we were able to leverage from other teams as well.”

By taking advantage of working code bases that handled both player-facing and ‘behind-the-scenes’ aspects of the social application, the FarmVille team was able to save a lot of time, and a lot of headaches, in bringing their game to market. “Basically, if something worked that did what we wanted or needed to do, we used it. Doing this allowed us to build and ship the game as quickly as we did.”

The FarmVille team leaned into and used other Zynga technologies, such as the Avatar generation engine from YoVille, in an effort to streamline their production process.

Great Art
FarmVille turned out to be just a gorgeous game,” Skaggs told me. “I think this is one of the things that went really well for us. We had a fantastic Art Director, Craig Woida, who came out of the traditional game business. Also, we had established a pretty clear vision for the art style of the game e fast and breezy e and having such a strong, clear visual style in mind for the game from the beginning made the entire process easier. The convergence of us knowing exactly what we were looking for style-wise, and finding Craig, an artist who could create that style, as well as lead a team to create that style, was a big win for the game and the way it ultimately came out.”

FarmVille offers a unified, compelling visual style, described as “fast and breezy.” This style was established early in the game’s development cycle.

Putting the right team together is always the goal, and luck is an important component whenever you wind up with a team that really comes together and gels. Skaggs agrees that luck was a factor, but also feels that their focus on bringing on the right people made a big difference in the assembly of the FarmVille team. “We were fortunate across the board, and made some really smart hires in some key positions.”

Tapped into Entire Company
“The core team for FarmVille was only 10 people,” explains Skaggs. “But people from all over the company helped us and contributed to the game becoming such a great title. The whole company was behind getting it done well.” Aside from sharing technology, Skaggs and the team made a point to reach out and include perspectives from across the entire organization. They had folks from across the entire company play builds, soliciting ideas and feedback from everywhere they could. “We got some great ideas from unlikely sources, and having so many different people pounding on what we were doing allowed us to keep improving the quality of the experience.”

This success factor assumes, of course, that testing was given a high priority during the game’s development cycle and that the team utilized testing feedback to help course correct and finetune the player experience. For informal testing, by looking outside of the core teamand tapping into the entire company, the FarmVille team was able to greatly expand their ‘test bed,’ and the overall quality of their game benefitted from this approach.

Launched on the Amazon Cloud
Another decision that proved critical to FarmVille’s success, especially early on, was to serve the application utilizing Amazon’s Cloud, instead of Zynga deciding to buy and staff their own servers. “All Zynga products have launched on the Amazon Cloud,” relates Skaggs. “Basically we’re renting instead of buying boxes ourselves. What this allows us to do, though, is to scale and ramp up to meet demand almost painlessly.”

Skaggs told me the story of FarmVille’s initial launch to illustrate this point. “Our plan was to do a ‘soft launch’ on Friday, June 16, 2009. We just thought, OK, let’s turn it on. Just for families and friends — in fact we called it ‘family and friends weekend.’ We, the team, had made an internal bet to see if we could get 2500 players by the end of the first 24 hours. Well, in actuality we wound up with 25,000. On the next day, we had over 100,000 players. Day 3, we saw 500,000 players and on day 4 we crossed the 1,000,000 player threshold — all this with no announcement, no cross-promotion in our other games, no advertising, no nothing.

“It took us a while to grasp just how successful the game was, even after this incredible launch. I guess we had all seen Internet phenomena that experienced explosive growth and then petered out just as quickly.” Eventually, the game’s success pushed the boundaries even of Amazon’s Cloud. “Fortunately, Bing Gordon, a games industry luminary who happens to be one of our investors, was able to contact Amazon CEO Jeff Bezos and pull some strings for us,” remembers Skaggs, “so they were able to pull in additional resources and keep us afloat.”

As players succeed in FarmVille they can take advantage of new opportunities in the game, including the ability to collaborate with Facebook friends to raise a barn on their plot of land.

What Went Wrong

Not Staffed Well for Success
“This is one of those great problems to have,” Skaggs said with a chuckle. “But when the player numbers started exploding — and kept exploding — we had to scramble across Zynga to try to keep up with the game. We had to hire a bunch of people quickly,many within the core team, but also in customer service, server techs, you name it. We had to really staff up in order to keep up with the demands of the game and also feature flow.”

Unlike more traditional game launches, social games are the gifts that keep on giving. New players continue to flock to the game, but current players also expect to expand their experience and discover new things as they continue to play. It can be a mad dash to try to keep up. “We had to initiate so many people so quickly into the Zynga culture. We also needed external help e for example, we eventually got a technical team in India looking after our servers so the American team could get some sleep at night. Looking back, it is clear we weren’t ready for our own level of success, but then again, how do you predict that kind of thing?”

Ongoing Leveling Curve Issues
The leveling curve in FarmVille did not work out as well as Skaggs and his team would have liked. “With an RPG or RTS game, moving up fromlevel to level is such a key part of the player experience, and getting that balance right is so important. At the same time, it’s a different balance in the social games space — it’s not exactly head-to-head competition, and also a very different type of player profile.

As you level up in FarmVille, you unlock new crops and buildings, and these are important to motivating players to continue, and some of our levels were just taking too long to complete. Instead of taking a day or two maximum, some levels were taking folks five days to complete. Fortunately we are able to course correct — we do so all the time — but there is always a cost to your current players, and I think we could have done a better job tuning the game and tailoring our level curve to our target audience.”

This is another issue, it seems to me, that relates directly to testing. As we are starting to see in case study after case study, high-performing teams test extensively in an effort to tune the interactive experience of their game to the target audience’s expectations and abilities. Despite this focus, though, until the game is released more widely it remains really difficult to nail this down. Only more extensive testing can help alleviate this issue; I am still waiting to hear from the team that felt they did too much testing during the course of development of their game.

Soft Launch Became the Real Launch
Skaggs discussed above the remarkable numbers of players who flocked to FarmVille from the absolute beginning. What was planned as a soft launch, for friends and families only, saw more than 1,000,000 players within four days. “We simply couldn’t undo over 1 million players playing every day,” says Skaggs. “And there were things we needed to have in the game that we were planning to include in the ‘real’ launch that we hadn’t yet implemented. For example, we had to furiously add lines of code to track how people were playing. We had to wire up the metrics tracking pieces after the fact, and metrics are really important to us as a company and to how we do business. Nowadays, no more games launch at Zynga without metrics tracking code — this is an important lesson we learned from the FarmVille launch.”

The FarmVille team knew they had a hit on their hands when player numbers crossed the 1,000,000 mark on day 4, after a ‘soft’ launch without advertising or cross-promotion.

Triggering Facebook Failsafes
“Again, I don’t know how you predict a problem like this,” explains Skaggs, “but early into that first summer we started bumping into some Facebook limits that hampered our progress. Like some of their failsafe code that attempts to automate the process of dealing with issues within your application. The actual numbers don’t matter, but let’s say Facebook has established a system to make sure that all applications are functioning correctly, and they verify this by tracking communication ‘errors’ like timeouts and the like. Well, they establish thresholds that seem reasonable for 50,000 people using an application, and then FarmVille comes along and we’ve got 10,000,000 people using the application. Suddenly a normal ratio of timeouts and other errors, which are simply a part of life online, trigger alarms in the Facebook system and the game is shut down.”

This is the kind of problem all game developers would love to have — their game is too popular and overloads the system or platform. “Facebook was excellent to work with, and eventually we got it all figured out, fortunately,” concludes Skaggs, “but it was no fun at the time.”

Should Have Made Changes Faster
“I mentioned some of the issues with our leveling curve earlier,” began Skaggs. “And there were other things as well — technical glitches, or infrastructure changes — basically, changes we knew we needed to make, but that we delayed over concerns about the installed base.” Having millions, and then tens of millions, of people, using your product adds a lot of weight to decisions to change anything. What if I make it worse? What if I delete something that is attracting many of these fans?

At the same time, according to Skaggs, it is a mistake not to go ahead and make the changes you realize that you need to make. “We definitely should have just gone ahead and made the changes. If you wait, you’re not really servicing the new players, or the current players. If something should be changed, go ahead and change it.”

Making changes to a world with tens of millions of players can be daunting, but once you realize that there are changes to make, you have to go ahead and bite the bullet, advises Mark Skaggs, FarmVille’s Creative Director.

Wrap Up

I asked Skaggs if he had any parting thoughts, looking back on the process of creating Facebook’s most popular application. “I would say the biggest takeaway is to make sure your application has all the metrics and analytics tracking code in place before release, in order to keep track of what’s going on with it. Coming off the FarmVille experience, we now have a Zynga metrics team as well as an entire stats group — these are centralized resources that service all the teams. We are an extremely metrics-driven company. But this is one thing that is kind of cool about this space, and Zynga specifically. Instead of wondering what’s going on out there, or even if a button should be red or blue, we can track and analyze the statistics, and do a whole lot of A/B testing along those lines as well. So, be well metric-ed and know what’s going on before you go out the door.”

Skaggs has blogged about an underlying design philosophy he and his team employ at Zynga, which he calls ‘Fast, Light and Right.’ He explains: “Speed is very important in this space. Don’t think of design that you’ll do in 6 months; figure out what players want, and do it quickly. If what you’re working on is not getting your game ready for launch, or post-launch support, than you’re just wasting your time.”

‘Fast, Light and Right’ — the design doctrine employed at Zynga that has helped make FarmVille the most popular application on Facebook.

Don't miss the next Insight article on Get the free newsletter in your mailbox each week. Click here to subscribe.

Excerpted from Making Great Games: An Insider's Guide to Designing and Developing the World's Greatest Video Games by Michael Thornton Wyman. Copyright © 2011. Used with permission of Focal Press. All rights reserved.