How I Spent my Spring Break: A Report on the 2000 Game Developers Conference
by François-Dominic Laramée

Despite Air Canada's best efforts to the contrary, I have survived another edition of the Game Developers Conference.

This year's GDC was held March 8-12 at the San Jose convention center, and was by far the biggest of the innumerable (two) editions of the show my poor starving freelancer self was able to attend. As usual, several trends emerged from the talks and the expo floor, some of which came up again and again, in lectures by people from every corner of the industry, to an almost surreal effect. While it is far too soon to make intelligent guesses as to which of these trends will revolutionize the way we work and play, and which will turn into next year's vaporware buzzwords, I have never been one to refrain from reckless prognostication, so I'll jump into the fray and tell you what I think are, in no particular order, the 5 most important concepts discussed in San Jose.

At best, this article will establish my reputation as a visionary and bring in heaps of praise and lucrative consulting contracts for years and years to come. In all likelihood, it will give you plenty of reasons to seek me out at next year's show, so you can stand beside me and look smart in comparison.

We're big

How big?

Really big. The game industry's total sales will overcome those of the movie business this year.
Really, really big. Jon Peddie Associates predicts that there will be about 1 billion internet-enabled, 3D-enabled devices in the market by 2004. This includes PC's, consoles, set-top boxes and possibly other things we don't know about yet. With DVD drives being included in PlayStation 2 and X-Box, we can start thinking in terms of VCR-like (90%+) penetration of the home market.
Too big. In 1999, according to numbers quoted by Hasbro Interactive's Tom Dusenberry, no less than 7,590 games were available on the PC market, and only 199 of them sold 100,000 units or more. On the console side of things, the situation is not that much better: 2,250 games were on the market, and 288 sold 100,000 copies. To rise above the crowd, the big boys may spend the rest of the industry into oblivion: Yu Suzuki's RPG for Dreamcast, Shenmue, required the services of 300 people at the end of the project, and it is not hard to believe that it cost 50-60 million dollars in development alone. Who can compete with that?

Other things to keep in mind:

At the present time, 60% of online game players are women and 79% are 25 or older. However, by 2004, kids and teens will account for 60% of the total online player market.
Only 16% of the PC games on the market are 3D-enabled, and a mere 2% actually require 3D hardware to function. Despite all of the hype, there is still a great deal of 2D product being shipped, and most of the best-selling PC games are either pure 2D (Deer Whatever, Who Wants to Be a Millionnaire) or would work equally well gameplay-wise without 3D technology (Sim City 3000, Age of Empires II, Baldur's Gate). Of the top-ten PC sellers of 1999, only Half-Life and probably Microsoft Flight Simulator qualify as must-be-3D games.

Unlimited capacity

Both the PlayStation 2 and the X-Box promise developers the Holy Grail of effectively unlimited 3D capabilities. The numbers are staggering: 1 trillion operations per second on the X-Box, 300,000 polygons per frame at 60 frames per second using RenderWare 3 with only one of the vector units on PlayStation 2, etc. Even if you apply a healthy dose of skepticism and cut these numbers in half, the net results still boggle the mind.

For developers, this changes everything. Instead of making feature decisions based on hardware ("this won't work"), they will have to make them based on design choices ("we don't have time or money to do everything the machine can support, so what do we really want?") Low-polygon character models will soon be a not-particularly-pleasant memory, like smallpox, powder-blue polyester suits and the New Kids on the Block. The technological quantum leap also suggests that these machines will have longer life cycles than their predecessors: it will take many years before anyone even approaches the X-Box's limits. From the dark side, publisher pressure may drive development costs even further into the stratosphere. ("The machine can take it, so please add these 1,200 additional features, won't you?")

Each of these machines comes with a caveat:

Microsoft promises the X-Box for Fall 2001, but the company has a questionable record when it comes to delivering quality product in version 1.0 and on time. However, X-Box is based on well-established technology, so I wouldn't bet against it.
During Phil Harrison's keynote address, Naughty Dog's developers suggested (rather forcefully) that programming the PlayStation 2 was every bit as excruciating as reported earlier. If the middleware market doesn't materialize (and fast), many smaller developers will be tempted to abandon the platform and wait for the DirectX 8-based X-box instead.

Episodic distribution

This was the single most pervasive idea at the show. Everybody talked about it, from Sony to Microsoft to Wild Tangent to Hasbro. (It made me feel pretty smart, considering that I suggested the exact same scheme to a broadband interactive TV project more than five years ago. Wait a sec while I pat myself on the back. Ahhh.)

There are several key reasons why distributing games as short episodes and/or time slices has everyone excited:

People are used to episodic formats in TV, magazines and other media. This is a natural.
Online distribution works better for smaller downloads.
You can bring a product to market faster if you only need to produce a single level before release and add more when/if demand requires it. You can even keep a winning product alive for years by pumping out new content as long as people buy it.
It is probably easier to get 500,000 people to pay $5 five times each than to find 100,000 who will pay $50 once.
Episodic distribution reduces risk, as a product that bombs and is taken off the market after 2-3 trial episodes will cost less than an equally unsuccessful 50-level CD-ROM marathon.
Except for the subscription-based games like Ultima Online and Everquest, gamers pay essentially the same amount for a game whether they play 10 or 500 hours. Imagine how rich Carmack would be if multiplayer Quake II cost even 10 cents per hour to play?

Broadband

Of the U.S. households with internet accounts, 40% of those who have access to broadband connections (cable modems or xDSL) have adopted it. This rate of penetration is faster than that of the VCR.

Consoles are getting into the fray as well. PlayStation 2 will not ship with a modem (although it can accept a third-party USB device) because a broadband add-on is coming next year. X-Box will ignore traditional internet access entirely and ship with a 100 MB/sec Ethernet card. A massive installed base of broadband-enabled devices will make online distribution of demos, levels and even full games mainstream (especially for the X-Box with its 8 Gig hard drive) and will allow developers to create whole new genres of entertainment (i.e., 11-on-11 multiplayer football, massively multiplayer X-Wing dogfights, etc.)

Mass-market pricing

Roller Coaster Tycoon was the best-selling game of 1999, and it came out at $29.95; Who Wants to Be a Millionnaire sold 600,000 copies in a month; Deer Avenger made the top 10. When a game is inexpensive, it can become an impulse buy, and once the impulse bug has hit a few hundreds of thousands of people, word of mouth takes over. Frogger's sales doubled from Year 1 to Year 2, and nearly tripled from Year 2 to Year 3. Yummy.

Low price points also make simpler games economically viable. Few people would spend $60 on a 100-level puzzle game on CD-ROM, but they will happily shell out $15 four times for the basic 25-level game and three add-on packs. Selling a piece of an episodic game online at 3-5$ also plays to consumer inertia (who will bother to cancel a $3 subscription?) and may have the additional side effect of making piracy obsolete (who will bother stealing a $3 game?)

Conclusion

Yes, the GDC is expensive. Yes, the food provided by the conference was unspeakable crud. Yes, the free shuttle service was woefully inadequate, with buses packed tighter than neutron stars going right by my hotel more often than not in the morning. Yes, I had to wait in line for an hour and forty-five minutes at registration because I never received my badge in the mail as promised, even though I paid online on December 23rd and showed up early. Yes, I do think that the conference is 20-40% overpriced. Of course, I would have appreciated if they had accepted my proposal for a lecture on speech interfaces and given me a free pass. I would also like to increase the value of the Canadian dollar, cure smelly feet and beat Roy Jones Jr.

Still, there is no substitute. See you there next year.

François-Dominic Laramée, March 2000

Discuss this article in the forums

Date this article was posted to GameDev.net: 3/15/2000
(Note that this date does not necessarily correspond to the date the article was written)

See Also:
General