Monday, April 13, 2009

GDC 2009 Thoughts

Due to the fortunate aligning of several circumstances, I was able to once again attend GDC this year. In a previous post I already expressed my excitement about conferences in general. This year's GDC proved to be solid, and I even got to spend a couple of days exploring San Francisco to boot. I've been telling people that about 70-80 percent of the sessions I attended were either interesting or valuable and I'm pretty happy with that ratio. Either the conference organizers are doing a better job of selecting content or I am getting better at picking my sessions. In either case, it's a win. While I was originally thinking of breaking this post down into 3-4 digestible chunks, the conference has come and gone. Along with it has gone most peoples interest in the details. So here it is in one big batch.

Here's a brief rundown of the sessions I attended:
Keynote - Discovering New Development Opportunities Satoru Iwata
I've always been impressed with Mr. Iwata and once had the pleasure of meeting him during my days at Nintendo. For someone that manages one of the biggest companies in the world, he strikes me as being grounded and down to earth. He clearly understands the challenges developers face and likes to help teams succeed. This keynote was focused on his own journey in turning HAL around as well as a brief overview of Miyamoto's design process. The main takeaway here is that small prototypes are highly valuable, even if they don't turn into games for a long time. Several companies outside of Nintendo see R&D as a valuable process and a few manage to follow this prototyping process. I think this is probably the exception rather than the norm, however, and many companies either find themselves in full production too early or spend too much money figuring out the core mechanics of a game.

Fault Tolerance: From Intentionality to Improvisation - Clint Hocking
An excellent postmortem on Far Cry 2 that detailed the design philosophy behind improvisational gameplay. Sandbox games demand improvisational gameplay, and as a result, are one of the most demanding genres to work in. Clint broke down Far Cry 2 into its core systems and highlighted which ones encouraged the improvisational style of play and presented alternatives to the ones that didn't. I like how thoughtfully dense this talk was as well as the fact that it builds on design ideas that Harvey Smith presented at a prior GDC. On a side note, I think Clint beats himself up too much about Far Cry 2, which, while flawed, is still one of the best games of 2008. I look forward to seeing how he takes these ideas forward into his next game.

Ups, Downs, Mistakes, Successes in the Making of LittleBigPlanet - Alex Evans/Mark Healy
I think the main thing I took away from this talk is the fact that Media Molecule truly is a collection of amazingly talented mad scientists. With a group of 31 people this studio managed to ship an amazing game in line with each of their publisher milestones while being incredibly flexible in changing their game when things weren't working.

Level-5's Techniques to Producing a Hit Game - Akihiro Hino/Usuke Kumagai
I enjoy many of Level-5's games. Unfortunately, this talk didn't really go into much detail about their production or design process. This talk was labeled as a production talk, but I really think it should have been called a marketing session instead. The takeaway is that design hooks can be used to market the game via press and trailers which can grow the buzz for your game. Their approach is largely focused on the Japanese market where they produce very high quality anime trailers that clearly appeal to their target audience. The closest thing I can think of in the western market is the GTA trailers, which are excellent but tend to focus on the characters and world rather than the design hooks.

Halo in the Laboratory - John Hopson
Usability testing is an area that I find incredibly fascinating. The scope of the usability testing for Halo constituted a large part of the pre-release Halo 3 article that Wired wrote and I was curious for even more details. As I thought, the session reinforced my idea that big games need a lot more user research to get the details right and companies like Bungie and Valve are already using this data to constantly improve their games. Some of the interesting takeaways from this session included low cost ways to do usability testing via a webcam or simple UI mockups that mimic the semantics and flow of your UI.

Keynote - Solid Game Design: Making the 'Impossible' Possible - Hideo Kojima
This was essentially a talk focused on the origin and evolution of the Metal Gear franchise. One of the things I find most interesting about this series is that you can trace most of the elements of pre-MGS4 to the original Metal Gear Solid 2 for MSX, included in the Metal Gear Solid 3: Subsistence discs. What made this talk interesting was learning the origin of the original Metal Gear/Metal Gear 2 game mechanics. One of the finer points that I think must have gotten lost in translation is the idea that Kojima believes western developers use software technology to overcome game design challenges. While I believe he is referring to middleware, I am not convinced that middleware has improved western game quality on the whole.

10 Things Great Designers Exhibit - Gordon Walton
Like any top 10 list, the content of this one will probably be debated by many people in the industry. The short version is: passion for games, breadth and depth of knowledge, problem solving and analytical skills, flexibility, KISS, player empathy, continuous learning, team work, a positive mental attitude, and clear communication. I agree with the bulk of this list as well as the general order of priorities, but there are a couple of "things" that I feel were not on the list. Chief among these are sufficient technical knowledge and persuasive abilities. I find the first to be valuable because of how well my programming background is serving me in my design experience as well as my belief that our work will continue to get more technical as game engines mature. The second seems like a no-brainer to me because designers that can't convince others why an idea might work well will have a hard time with their self-confidence as well as their ability to push an idea to the point of awesomeness. While one could lump this skill under clear communication, I have met plenty of people that can communicate ideas clearly but cannot convince others why they are worth pursuing.

Experimental Gameplay Sessions - Jonathan Blow, moderator
Inspirational and excellent. The EGS panel did a fantastic job selecting interesting indie games that all explored different ways to shake up what we think of tradionally as our game "play space".

Dirty Deeds Done Dirt Cheap: Design Lessons Learned from Rock Band - Dan Teasdale
Harmonix is a company that has come to truly understand cutting edge design by iterating over several games that continuously improve their understanding of the music game genre. One of my favorite talks of GDC 2007 was the Harmonix session on building games from Frequency to Guitar Hero where they shared lessons learned along the way. The 2009 talk was largely focused on the evolution of Rock Band and once again, many valuable lessons were shared. The takeaway is that we really need to understand what it is we are trying design within the constraints of what our audience wants and how much enjoyment they will get out of it. While this may seem obvious, it still seems like a number of features creep into games solely for the purpose of tacking another bullet point onto the back of the box.

Dead Space: How We Launched the Scariest New IP - Chuck Beaver
An excellent post mortem on the development of Dead Space. I salute the team for making it through an 18 month greenlight process. The Dead Space team did a good job of breaking the mechanics down into buckets of conventional, evolutionary, and experimental categories and knowing where to put the right amount of resources behind each of these areas. A number of my colleagues feel that Dead Space is the evolution of the survival horror genre and based on the reviews of Resident Evil 5, I suspect that many others feel this way too. The presentation touched on things like the outdated control scheme behind Resident Evil and how EA did their own usability testing to push the controls into a more modern direction. In addition, the session also did a great job outlining elements that effectively create a believably scary atmosphere and where the best bang for the buck is within this type of game.

Lionhead Experiments Revealed - Peter Molyneux
I thought I had an idea of what to expect for this talk - an series of Lionhead prototypes. I wasn't that excited about it initially but it turned out to be more engaging than expected, primarily because it outlined a way of structuring an ongoing prototyping process within a larger company. One of the most interesting things that Lionhead is trying to do is to create a tool that makes game content and assets easily shareable for prototyping purposes. I think this would be a huge win for many companies because often times team members have a hard time visualising how gameplay might work with placeholder graphics and executives certainly don't like to look at boxes.

The Iterative Level Design Process of Bioware's Mass Effect 2 - Corey Andruko/Dusty Everman
Rigorous, thorough, and highly focused on constant improvement. Those are the main thoughts that came to mind as I listened to the speakers outline the level design process that Bioware uses. I'm not sure how many companies have such a well defined development process, but I think it must help the team understand gameplay objectives and where things are headed. One of the most interesting facts they mentioned is that the game was fully playable in some form as of Christmas break 2008, over a year away from ship. I find this to be mind boggling because the projects I worked on were largely focused on getting a single piece of the game highly polished and then cranking through the rest of the game based on that polished slice. While this approach definitely works and generates a lot of excitement throughout the team, I imagine that having the whole game playable so early tends to get a lot more eyeballs across the entire spectrum of content.