If there's anything that defines Valve's Team Fortress 2, it's the game's commitment to change.
Compared to its launch in 2007, the game has become almost unrecognizable, as it's grown from a traditional team-based shooter into an ever-evolving free-to-play service with persistent loot, myriad new game modes, and much more.
The game's ongoing content updates have practically become the lifeblood of the hit shooter, as they've given Valve the opportunity to experiment with new ideas and add new content that helps the game stay in line with the latest market trends.
Very few online shooters continue to grow five years into their lifespan, yet Team Fortress 2 has proven itself a rare exception. We recently got in touch with the game's lead designer, Robin Walker, and he said that much of that success can be attributed to his team's unique approach to post-launch content.
The real trick, Walker explained, is to make sure that the core dev team is responsible for all aspects of a new game update -- that includes design, marketing, and even community outreach. During our interview, Walker detailed how he and his team craft their updates for Team Fortress 2, and offered a number of tips to help other developers better support their online games.
The first step to creating an interesting game update, Walker said, is to determine how it will be presented to your players. Team Fortress 2 updates tend to revolve around a specific idea or theme, and Walker has found that it helps the design team identify what makes a new release interesting and unique.
"After each major [Team Fortress 2] update, our team gets together and sketches out a general framework for the next update," Walker said. "The team spends some time discussing what combination of these various goals makes sense for the next update, and then we try to find a customer facing wrapper around it."
In the past, these "wrappers" have been specific character classes, new game modes, and even real-world holidays. By giving the update a single focal point to work around, it becomes much easier to give the release an identifiable label that will grab players' attention.
"Sometimes we’re forced to fall back on something more abstract if the update includes a wide ranging set of disparate features (like the Hatless Update). Occasionally it works the other way around, where we’ve got a customer facing update wrapper already (like Halloween or Christmas) and we can derive team member’s work from the wrapper."
Once the "wrapper" is in place, it's time to determine exactly what the update will entail. And for Walker and his team, designing an update isn't just about creating new content; it's also about creating content that's marketable.
"We think that marketing is a game design problem, and that means we need to ensure each update is easy to market. So once we’ve know what an update’s public face is, we start trying to sketch out how the release will work," Walker said.
"If we don’t think [an update] is exciting enough, or we find the messaging is hard to write, then we know the update needs more [content]. Sometimes that means adding new features to the game; other times it means adding more content outside the game, like a movie or comic that drives our narrative forward, or a meta-game like the Soldier vs. Demoman war," he said.
"The important thing is that the development team itself is responsible for figuring out how the update will generate excitement. That way, these things get properly rolled into the update itself, and don't end up feeling like a tacked on marketing ploy."
Even once the update is ready to go and the marketing plan is in place, however, the team's job isn't over. Once a new update goes live, the dev team needs to determine how their new content has affected the game. To do so, Walker said developers should remain in close contact with their community, as qualitative human feedback can be far more nuanced and valuable than numbers or metrics alone.
"We’re terrified of making decisions driven only by statistics, because they may result in short term gains and long term costs (due to our current lack of hard stats for customer happiness). In the end, we care more about the general customer response to our updates than the statistical results," Walker said.
"Don’t put anyone in-between the team and their customers. The team members must read customer conversations, support calls, and so on, because it’s all data for them to understand better how they’re doing. Each of those interactions is also an opportunity to provide novel and valuable content to their customers. Interacting with your audience outside your product is a game in and of itself, and the team should try to do that here as well as they do in-game."
In all, Walker believes the key to updating and supporting an online game is maintaining a collaborative, unified approach to development. Even if your team is full of design, marketing, and community specialists, it's easier to create a cohesive product if everyone is working together as a single unit.
"I'd recommend that you approach your updates holistically, so that the team members know that they must consider all aspects of releasing an update as a single problem they must solve," he said.
This holistic approach certainly requires diligence, ample communication, and a lot of hard work, but if Team Fortress 2's ongoing success is anything to go by, it can pay off in spades.