GDC 09: Dan Teasdale on Rock Band 2

In less than four years, Harmonix has successfully birthed not one but two extremely successful franchises that have both moved to a yearly format. First there was Guitar Hero, which Harmonix developed in 2005, and them moving on to MTV in 2007, where they developed Rock Band. Even though it was bright and early Friday morning, Harmonix designer Dan Teasdale had a pretty nice turnout for his GDC session, entitled “Dirty Deeds Done Dirty Cheap,” where he talked about the challenges of designing games with a sub-yearly development cycle.

To start, Teasdale mentioned two stereotypes of game the game industry: that publishers love franchises with yearly installments, and that developers have the exact opposite sentiment — that yearly games are the devil, that you can’t innovate and push the industry forward with yearly timeframes. From here, Teasdale launched into an exploration of the Harmonix design process, which he broken down into several main categories.

The One Question

The first challenge is how to maintain one design vision with a group of 100 people. As Teasdale put it, it becomes sort of a “miracles to-do” list: read people’s minds and get everyone on the same page. And the way they accomplished this was with what they called “the one question” — something that the design team could always keep in the back of their mind, and keep them aiming at the same target. For Guitar Hero, it was “Does It Rock?,” and you can see that mindset permeate every aspect of the game, from song selections to the artwork.

But for Rock Band, the challenge seemed a little tougher. At first, the question was “is this different compared to previous games?” But that felt vague, and the team was stumped until someone pulled out a video of The Who’s “the Kids Are Alright,” and Teasdale showed the breakdown section in the middle of “Won’t Get Fooled Again” where Roger Daltrey just screams “Yeaaaaaaahhhh!!!!!” before the band comes back in. It’s lights, it’s screaming, it’s rock and roll. And this summed up what Harmonix wanted the game to be: “Is this an authentic band experience?” And so that helped steer the direction of the game, where you’d create a band, play arenas and basically live the life of a touring band.

As development continued, the team picked out little things that would help sell the authentic band experiences. It meant no powerups, no minigames, and no guitars on fire, but things like big rock endings, solo bonuses, and being able to cover for and rescue bandmates from failing.

Content vs. Design

Teasdale spent some time stressing the importance of truly separating content from design when it comes to making a game on a short schedule. He used the example of Guns ‘n Roses’ “Shacklers’ Revenge,” which arrived at Harmonix on July 2. The game was supposed to go to general master less than two weeks later, which represented a problem, since the song would be connected to various aspects of the game: world tour setlists, challenges, etc. But Rock Band was set up in a way that the developers could just throw new content at the game and the mechanics would sort it out. In this case, “Shackler’s Revenge” contained tags like “rock,” “00’s” “Los Angeles” and the game automatically arranged where it needed to go. And so getting new GnR turned into a huge win without having to tear the game apart.

Playtest Feedback

Teasdale started this topic by talking about some stats Harmonix had on what difficulty level players were playing at. As Harmonix expected, the largest amount of people had completed campaigns on Easy, then a large amount had completed Medium, and then only a tiny percentage had completed Hard or Expert. But what surprised Harmonix was how small a percentage of people had completed any campaigns at all, compared to how many people were playing the game. To Harmonix, the World Tour was the core of the game, and only a tiny fraction of their players were bothering to finish it.

And so Teasdale stressed the importance of learning from your community. Initially, there were plans to have a 4v4 competitive band mode in Rock Band 2, but it turned out virtually no one was playing the other competitive modes (like Score Duel or Tug of War). But as it turned out, people liked competing against scores, and so the Battle of the Bands challenges was born, where there was a new daily event for players to chase high scores, a feature that required relatively little effort to implement and offered a bigger payoff.

Of course, trying to learn from your community can be a double-edged sword. Teasdale posted a screenshot of a ScoreHero post soliciting requests for Battle of the Bands ideas, with one particular post droning on forever about what was broken with the game instead of just offering what was asked for: a cool theme and songs to go with it. The reality, of course, is that some segment of your audience isn’t adept or interested in following instructions or offering proper feedback (usually lumped under the label of “insane”), and so there’s the challenge of finding feedback that’s actually useful.

The Future

The challenge for Harmonix now, Teasdale said, was how to keep innovating while still churning out titles to satisfy its publisher. Obviously, there’s the Beatles game coming out later this year, although details on new features are still slim. (Later in the Q&A, someone asked if Harmonix had considered new instruments like a keytar for future games, which Teasdale wouldn’t comment on.) Not only does Harmonix have its own games to compete against, but Teasdale put up a comical screenshot of the future landscape of rhythm games, showing the insane barrage of Guitar Hero games in the pipeline. Teasdale mentioned the Rock Band Unplugged for the PSP, which marries Amplitude-like gameplay with a Rock Band narrative, and a desire to keep improving the current games, such as plans to add star-tracking to RB2 via a future patch.

Ultimately, it was an interesting look as to how a successful developer looks for ways to keep pushing their games forward while still working within the confines of a system where they need to produce titles in a relatively short timeframe.