Media coverage of the talk was fairly scanty, and we’ve been after video or at least audio of the presentation ever since. Thankfully, Dave’s new company, Gazillion Entertainment, has posted the full presentation on YouTube. The forty-minute talk is broken up into three parts, and while the audio is kind of echo-y (turn up the volume) it’s quite watchable.
In part one Dave talks about the early days at Condor, when they were making console ports and desperately pitching every developer with their idea for a dark dungeon RPG. After some early struggles, Dave and the guys at Condor bumped into another small development studio at a a trade show, where they discovered, to their mutual surprise, that they’d all been working on different console ports of the same title. Those other guys had recently upgraded their company title from “Silicone and Synapse” to “Blizzard” and when they secured some more funding they commissioned Condor to go into production of that dungeon RPG. Condor ultimately became Blizzard North, the RPG evolved into Diablo, and after hurdles involving a transition from turn-based to real time and a late-game change over to allow online play via the just-created Battle.net, amazing success followed.
Part Two briefly covers the creation of Diablo II, and ends with a PowerPoint slide saying, “And there would be a Diablo III…” Dave doesn’t say anything more about it, and leaps right into a discussion of World of Warcraft and other big-budget MMORPGs. It’s an interesting transition, since as we so recently learned from Max Schaefer, the early version of Diablo III was planned as an MMORPG with a monthly fee. This seems very relevant to Dave’s talk, since much of his Diablo II segment focuses on the problem of Diablo II; fantastic sales, but no ongoing revenue, while they had heavy ongoing costs to maintain Battle.net servers for all the Diablo II players.
Click through for the video to parts two and three, and a some analysis of Dave’s insights into why games (like Diablo III) take so damn long to make these days.
The rest of part two largely focuses on MMORPGs, starting with WoW, and continuing to other massively-budgeted projects that tried to “Out WoW, WoW.” And failed. Dave shows a slide of the “2007 to 2010: MMO Graveyard.” Headstones include Tabula Rasa, Age of Conan, Warhammer Online, Tabula Rasa, and… Hellgate: London.
Dave gives a high-level postmortem of Flagship Studios, talking about how they tried to do way too much, built everything from scratch, how the staff ballooned to over 100, and how they did a poor job managing “Development, Finances, Expectations, and Subscription Messaging.” Which about covers it.
During this segment, at around the 11:00 mark, Dave does a great job explaining why giant games (like Diablo III) take so long to make. It’s largely caused by the size of the teams, and the fact that everyone has their own specialties. In the early Diablo I days (and on smaller dev teams even today) everyone did multiple things; wore a lot of hats. The concept artists were also technical artists were also animators, were also programmers, etc. Now there are such large staffs that everyone does their own thing and it’s all driven by a committee. Dave’s example of adding a chest to Diablo dramatizes this.
Paraphrasing: “Let’s add a chest to the game. Awesome, sounds like a good idea. Artist runs off, makes a chest, four frames of animation, sprite driven, cut it out, into the game, in three hours. Works great.”
Today, adding a chest is a huge project. Concept artist spends all day making 27 pieces of concept art for a chest. Then there’s debate; which ones match the game theme, and background, etc. Then you need a model artist to work it into game form, then texture artists to texture it, then animators, then a tech artist, then programmers are required to put it into the game, then special effects artists, and on and on. As Dave says, “…all of sudden it’s two weeks to make this frickin’ chest! … It’s really hard to do anything now that used to be a simple task.”
He’s talking about how things worked (bogged down) making HGL, but clearly his description applies to to any similarly-huge development team. It works for Blizzard since they’ve had so much success (and have the infinite revenue stream of WoW subscriptions) that they can take forever to make games. Hence Diablo III. But this massive team development and compartmentalized working goes along with the endless iteration, and that’s largely why their games take so long. Because it takes 5 people two weeks to put a chest into the game, and they probably “iterate” that three times over the course of development.
Is the final chest (or whatever) that much better, given how much longer it takes to make? Of course not. But it’s just sort of how the development process goes on mega-sized projects. Which is a major reason Max Schaefer and Dave Brevik are trying to keep the team sizes smaller and the development much more focused and nimble at their new companies. I talked to Max Schaefer about this very issue last week, in terms of how they’ve been able to develop things in Torchlight so much more quickly (you’ll hear that in part two of our interview, which will be posted next Monday in the podcast.)
Part Three: Dave talks about the rise of Facebook games, and how they’ve proven that F2P can be very profitable. He also stresses the importance of accessibility, and how browser-based games are becoming better and better, and will soon include 3D graphics and full screen play nearly as good as downloadable products. He feels that this is a big time of change for developing, and that new technologies are improving accessibility and making small team development very viable.