Well, I made it through day 4 of the GDC. I always find the second-to-last day of any conference the longest, since you've already made initial contacts with everyone and seen enough tutorials and sessions to make your head explode. But, enough about me, time for info on what happened.

This was the second Expo day, so there were a lot more people in the conference who were in on just Expo passes (admission to only a few conference sessions and full access to the show floor, but much cheaper). Fortunately, only one of the sessions that I went to was an Expo session, so it was the only one packed beyond the gills with people.

There were quite a few sessions and round tables today, many of which conflicted on the schedule, so it would have been beneficial to have a second set of eyes and ears to catch the other presentations. However, the folks who run the conference are offering a CD with the talks in MP3 format on their website and here at the conference. I ordered a similar CD for E3 last year and was pretty happy with it.

Chris Taylor, president of Gas Powered Games, makers of Dungeon Seige (published by Microsoft) spoke in the first session of the day about "Lessons Learned from Dungeon Seige." The talk was probably the most entertaining one so far. Overall, I think that Raph is a better speaker, since he is both entertaining and informative, but Chris's talk was a great way to start Day 4. Full of lots of anecdotes and tidbits of help. Although much of the discussion was on topics that had been discussed fully elsewhere (as seems to be the case with most of the post-mortem talks--see later in this missive), it is interesting to see the evolution of the industry in how people think about certain techniques and mechanisms. The key ideas that Chris presented were the conflict between a large game and a well tested game and the importance of linking information internally in such a fashion that people browse it instead of reading it. The idea here is the manner in which people meander around the web is a good thing if they are wandering around information that they need to know (like internal specifications). This is much better than people having to go directly to a specific document, explicitly check it out from a source code control system and read it. Then decide if they want to find another document. Interesting general thought on corporate information culture.

The second session I attended was "Breaking the Rules of the Game", by Eric Zimmmerman and Katie Salen. Despite the absolute, bar-none worst slides I have seen at this show (or any other show that I've attended in the last 2 years), these two gave an interesting talk. It was all about the way that people play games IRL and on computers and how design needs to take into account these player types and accommodate, or at least react to, them. It was an interesting pairing to Raph's talk from the day before on Social Networks.

Third up was "Proven Strategies for Self-Publishing on the Internet: Real Life Stories of What Does and Doesn't Work"... damn, what a great title! Unfortunately, it wasn't nearly as good of a talk as it was a title. First thing that should of tipped me off was its structure. It was a panel discussion with 6 people and the moderator was from Forrester Research. Next thing that should have tipped me off was that when the introductions were made, all of the people except one were selling games that didn't take advantage of the network. Unfortunately, I spent 35 minutes in the talk before I decided to bail, and all I had to show for it was some statistics that I don't find particularly interesting.

The first session of the afternoon was another Gordon session, this time a round-table about Service in MMOG's. This was a great group of people. Next to me was one of the lead game designers on Shadowbane and around the rest of the table were people from many of the already shipping games in just about all of the MMOG genre's. This covered everything from Airfight to World War II online. The discussion was good, with lots of people (including me) chiming in with different points. Although few people took up my "some of the same rules for CS as those used in ISPs apply," it was a jumping off point on a number of the later discussions.

There were many interesting views, including the difference between games like Magic, in which characters have value on day one, and the MMORPGs where characters usually don't have value for some time.

Clearly, everybody agreed that customer service is the number one issue in an MMOG and that the first week to first month is the most crucial time in getting the message out about how your customers will be handled and what they should expect to get from customer service.

The last session of the day was another post mortem about Halo. The guys were interesting to hear talk, but in the end, it was like any other game development post mortem, some nice anecdotes, a few do's and don't's (do put your team in one place, don't open your mouth if you work for Microsoft).