Game Developer – October and November 2005

November 7, 2005

Resident Evil, Puzzle Pirates, and the Top 20 Publishers loom on the cover of Game Developer’s October issue. It’s been staring me in the face for quite a while, all throughout my diseased tenth month.

SpeedTree is the software on display in Skunkworks. The SpeedTree components allow for quick generation of vegetation, specifically forests. There’s not only a CAD program, but an API, Max and Maya plug-ins, and a model library. A piece of software I’m actually not only familiar with but have had the chance to play around with (bless you GDC), it was interesting to see a full-on review. It’s an intuitive setup and a product that’s shown up in several interesting games lately. If you’d like to see what I’m babbling about, they have an amazing demo called Trees of Pangea at their site. Well worth the download.

The Top 20 Publishers is a series of short rundowns on the most powerful organizations in the business. Though the #1 isn’t a big surprise (its initials spell Exploding Aardvarks), where other folks placed in the order was. Nintendo above Sony in slots 4 and 5? Take-Two only #10? Interesting choices. I agree with the assertions made (despite some good sales and great games, Sony’s revenues have slipped this year), but Codemasters being on the list honestly surprised me. Lucasarts is number 20, a generous decision, I think.

Daniel James looks damned good in a waistcoat. That’s one of the interesting tidbits of information I gleaned from “A Pirates Life”, a Q&A with the Three Rings designer and developer. I’ve followed Puzzle Pirates since days of yore (Beta), and their recent successes make my heart leap with happy times. More interesting to me was some of the talk of their newest game, Bang! Howdy. It’s a short time cycle title, a drop in and play online situation. They also talk indie game life in an ocean of large and hungry fish. More variations on the talk they gave at GDC, and all interesting.

Yay! Postmortem! “The Graphical Style of Resident Evil 4” is a look at the distinct visual and design decisions made by Capcom in making the best horror title of the year. The analysis pointed towards workflow efficiency as one of the major elements that allowed them the success they achieved. Improved technology and a willingness to re-examine gameplay elements supported the skeleton of the beautiful corpse here. Bad things: The combat emphasis made handling art assets difficult, as did the use of a deep field of view and the use of multiple cameras in certain sequences. Really, though, their wrong elements seemed more like challenges than the actual mistakes we see in other postmortems. Which makes sense, because the quality of RE4 overall is mind-meltingly high. My biggest regret from the early part of the year was the fact that I let the game slip past me without reviewing it. Ensaddening.

Meanwhile, the November issue of Game Developer has a Combine soldier from Half-Life 2 menacing me from the cover. Squee! UW-Madison inhabitant James Paul Gee talks serious games to coincide with the Serious Games Summit this month. Good stuff, as Gee discusses the ways in which all sorts of game types and environs can be put to serious use. He finds Yu-Gi-Oh and Pokemon interesting because of their use of somewhat advanced language, fulfilling a teaching element while rotting their minds with CCG madness. He also calls the government on their hypocrisy, denouncing violent games with one hand and developing America’s Army and Full Spectrum Warrior with the other. My favorite sentence from the interview: “I think the deep problem [with the public view of games] is that the powerful part of the public is the baby boomers. They’re in control of things, they’re the right age, and they just don’t understand games at all.” So true. Thanks again for being unique, Mom.

“Hovering on a Handheld” was a surprisingly accessible discussion of Wipeout Pure’s physics development. The PSP title plays fast and loose with traditional physics engines in the interests of fun gameplay and a brisk framerate. Essentially (the programmers who bother to read me start wincing) the ships are the only thing that are properly kept track of with anything resembling real physics. Even then, the collision detection is only done on a frame-by-frame basis. Handling is accomplished via two anti-gravity cones, one at the front and one at the back of the ship. It was awesome to read an article penned by a developer willing to sacrifice physics realism for a higher fun factor.

The cover article. “Scaling the Cabal” is an overview of the design process Valve used to create Half-Life 2. The original title was created by a process they dubbed “the Cabal”. The Cabal approach is a shared resource design system that incorporates some elements of eXtreme programming. The original game was crafted by a single group working at a brisk pace. The scale of Half-Life 2 required that a multi-cabal system be set up. The title took six years to make, and was crafted by several small groups working together to accomplish the entirety of the game. Each group was four or five people, with half being level designers, half programmers, and with one systems engineer. The designers acted as customers for the programmers and vice versa. The engineer was there to extend the functionality of Source for whatever ideas the cabal came up with. Artists, writers, and sound engineers were used as shared resources, organized into their own cabals. In the end there were six design teams creating content for the title, with six different ideas of what the game should be. They ended up having to create a cabal cabal to form the picture of the game into a cohesive whole.

Their practices were equally as interesting as their process. Keyframing prose was a practice they used whereby they strung story elements out in a fairly vague fashion. While they knew that Freeman would get from A to B in this section of the game, they kept what exactly happened there ambiguous until gameplay decisions were finalized, and then the story was written around the game. In production, level layout was accomplished using nothing but orange textures. In this way, designers and artists wouldn’t become attached to the way things were coming together during the design process. Testing, then, could concentrate purely on gameplay. They used a layer of human-readable abstraction in their scripting to allow for transparency of assets used, as well as allowing changes to be made to data and design without massive rebuilds.

Their biggest problem, it seemed, was consistency. With so many cooks working at the broth, the Cabal cabal had to sit down and hash out what exactly it was that they were creating. A representative from each cabal came together and gameplay elements began to be evened out. Interesting stuff done by one group in one section was twisted and examined to see if it would fit equally well into another section. Team-wide playtests ensured everyone knew what everyone else was doing, and led to some interesting cross-pollination of ideas. Once things began to even out, the designers did iteration after iteration on design, changing elements until just a few weeks before the launch date. (Must be nice to know that everyone will have to deal with your download client). The article was fascinating, easily one of the most interesting I think I’ve ever read in Game Developer. Though it sounds chaotic and sort of backwards, the cabal process must have been very rewarding to participate in, and it was really nice to be able to get a feel for it.

%d bloggers like this: