The Making of Hellgate: London

I suppose it’s only natural that the editors at EDGE have transitioned from Diablo to Hellgate: London for their latest “making of” piece, as David Brevik, Max Schaefer, Eric Schaefer, Bill Roper, and other notable Blizzard designers were involved with the development of both titles. And while I thoroughly enjoyed both games, I will controversially say that Hellgate: London was a much more entertaining experience:

Although many of Flagship’s staff had experience working on Battle.net, Blizzard and Blizzard North’s premier online service, Schaefer concedes that they had it relatively easy back in their old home. (It was our office and our business,) he notes. (We were allowed to run it as we pleased, but there were a lot of things Blizzard HQ did from an administrative and HR perspective that we just didn’t have to worry about. We’re game-makers, not businessmen.) So, when Namco pulled out of its commitments to Hellgate, Flagship was given something of a rude awakening. (Suddenly,) he continues, (all this stuff we didn’t have to worry about at Blizzard, we had to worry about. Namco’s goals for reasons apart from Hellgate changed. I don’t want to throw them under the bus, they had their legitimate reasons for changing, but it did leave us out in the cold. We had to operate an MMO, and be an online publisher as well as a developer. And that added a giant, crazy amount of work to the project things we weren’t expert in or prepared to do but we had to do to make an online game. We just started getting spread too thin.)

Compounding this was the fact that, unlike any MMO around (and, indeed, most singleplayer games), Hellgate’s design layout that is, levels, monster placements, items, bosses was, save for several exceptions, completely randomised. This was done in Diablo (and Diablo II) to maximise replayability but, of course, Diablo was a 2D game. Randomisation on Diablo’s scale had yet to be attempted in a big-budget 3D game, so Flagship found itself in need of a unique rendering solution. It ended up creating an entirely new graphics engine. (We did that because we wanted to do a lot of things that no one else was offering,) Schaefer says. (We wanted to do 3D realtime random level generation, and the existing packages at the time just weren’t up for it. And the amount of modifications necessary to make something existing out there work was almost as much work as just creating it from scratch.)

With the custom technology available, Hellgate’s procedural environments weren’t as difficult as you might imagine; Flagship’s directors and programmers, after all, had significant experience coding Diablo’s on-the-fly dungeons. And to the team’s credit they worked very well step out of one of Hellgate’s static hub levels and you’ll find a gameworld that’s not only often as consistent and well-paced as that of any scripted game on the shelves, but that also bears a striking resemblance to one of London’s yet-to-be-gentrified streets. (There was room for improvement,) Schaefer concedes, (but that was more because of the difficulties in producing artwork. When we did the graphics engine, we ended up doing way too many cutting-edge graphics features DirectX 10 features and so forth and that just made the production of artwork incredibly laborious and slow. It’s not hard to make a level that logically works, from a random-generation perspective, but it takes a lot of iteration and evaluation to make it compelling. And we just didn’t have the time to do that because we had so many levels, and so many areas, that we were kind of hamstrung by the laboriousness of the production process.)

The effects of slow asset production on a game with literally hundreds of different monster and weapon types, not to mention randomly generated levels, cannot be underestimated. This, coupled with the extra development time devoted to fulfilling Namco’s former obligations, was, according to Schaefer, Hellgate’s death knell.

(It permeated everything,) he says. (It reduced our iteration on a lot of things. It reduced the amount of time we had to polish at the end. Ideally, we would have had another six months, another eight months or even another year, just to polish what we had and make it slicker. We’d have made the opening sequences a little bit better, and gone over the quests. So much of what makes a great game is what you do after it looks like it’s done. It’s those last bits of polish, and that last go-around. At Blizzard we had unlimited time and budget, and Blizzard’s always taken advantage of that, and will never release anything before it’s perfect. We simply didn’t have that luxury.