Next Generation
July 1996, p.28
Getting Connected
There's no doubt that playing games against a human is infinitely more fun than playing against a computer. Until recently, however, graphic multiplayer gaming was solely the domain of those with access to their workplace's local area network - not the ideal place to play games. But now all this is changing.
Online, multiplayer gaming is the hot new topic of conversation in the gaming industry. But is it really the future? Eventually, will all games be "sold" as downloads, doing away with CDs and cartridges? Where is the technology heading? And what obstacles stand in the way of it getting there? What is certain is that whether you play games on a PC or a console, within the next two years, you'll be playing a lot of your gaming online. In this Next Generation report, we examine the main issues facing online gaming's growth, as well as profiling the companies involved in getting you connected.
There have been multiplayer electronic games since the dawn of computing. Space War!, the first real videogame, programmed by Steve Russell on the PDP-1, was an exclusive two-player game. So was Nolan Bushnell's pioneering coin-op Pong. But a partner with whom to play games was not always available, so inevitably, one-player games soon sprung into existence.
It quickly became apparent, however, that while the satisfaction of beating a computer's artificial intelligence could be immense, the main satisfaction of getting a high score was in knowing that you had not only beat the computer, but also every other human gamer who had played the game since it was plugged in.
Today, nearly all of the most popular games in the arcade feature simultaneous two-player modes (either on one machine, or in the case of driving games, through two or more machines networked together). Just think about it for a second, how popular would the fighting game genre be now if the two-player option in Street Fighter or Mortal Kombat didn't exist?
And it doesn't just stop at two players. Four-player Bomberman is generally considered the best console game ever at the Next Generation office (where news of HudsonSoft's imminent 10-player Bomberman for Saturn was greeted with displays of near apoplexy by some staffers). Games like PC Doom, or WarCraft II which support four PC players or more via local area networks, are famous for their staying power. Name a shareware program other than Doom that's played as often more than three years after its release.
Multiplayer games offer not only important social contact with other human beings, they offer a far greater challenge than computer AI. While almost anyone can eventually learn what a game AI will do in a given situation, and defeat the game (as attested to by victories of 25 to 0 by players who have learned the "shoot-the-corner" trick in EA's NHL '96), humans change their strategies do clever (and capriciously stupid), unpredictable things. They learn, innovate, and generally keep things far more interesting than does the computer AI.
"[Multiplayer] games are an excuse to socialize, an excuse to connect with other human beings," says Jeff Liebowitz, president of the recently announced InterPlay game network spin-off Engage. "It's somewhat analogous to nightclubs or bars. You want a nice atmosphere, a good bartender, good food and all that, but the real reason you're there is to socialize."
Given the truism, then, that the more human beings involved in the game, the better the gameplay experience (generally), we are left with the two questions: How can a solitary gamer be sure of finding a human to play against? And, what is the best manner in which to get the maximum number of people playing at the same time? While two people playing at once on one screen is fine for fighting games, using a split-screen for strategy games like Herzog Zwei or Return Fire can be irritating and, with few exceptions, attempting to engage more than two players (nice for PC games like Command and Conquer) on one screen is impossible. It also requires that the two players be in the same physical location.
For PC users, the solution is simple. Sort of. Playing a game with more than one person on multiple PCs over a LAN is a neat solution, except that very few people actually have access to a PC LAN, except, of course, at work.
What every PC user does have potential access to, however, is a modem. By playing games over the Internet, the number of users who can play is constrained only by technological limitations and by game design. (When you get to 1,000-player Doom, you run not only against real physical limitations on the game server, but you start to lose the important socialization elements present in, say, an eight-person game, in which you can really get to know - and hate - the people you're playing against.)
Playing games over a network isn't anything new, of course. Many of the first text-only games, like Dungeon and Adventure, were played online on dumb terminals attached to mainframes. Multiplayer, text-based fantasy RPGs, known as MUDs, first appeared almost 20 years ago. Although younger players, weaned on graphics-intensive games may not think much of them, MUDs offer extremely rich virtual worlds to explore, realtime interaction with other human beings, and complicated adventures and quests that put even the most intricate SquareSoft or KOEI RPG to shame.
For console users, the story is a little different. While there is no technical reason why a console can't use a modem (indeed, console modems have been promised since the days of ColecoVision), it wasn't until late 1994, with the introduction of Catapult's XBand for Genesis and later, Super NES, that a console modem was introduced in the US. Indeed, the road to online console gaming is littered with failures, most notably AT&T's ill-fated but much-hyped Edge-16, which was introduced at the 1993 Consumer Electronic Show (CES). Set to feature simultaneous voice and data transmission, it was scrapped in 1994, when it was realized that a $150 add-on requiring the user to find someone with whom to play (unlike XBand, which has its own game brokering and news and information service) just wasn't going to be a big seller.
In Japan, however, console modems have had a long history. Nintendo introduced a modem, as well as an online service for the original Famicom, but it failed to catch on, even after it was revamped as a Super Famicom online gaming service.
The future prognosis of console gaming-by-modem may be changing, however. Sega will be introducing an Internet-ready Saturn later this year (using technology licensed from Catapult), and rumors of a web-browser for PlayStation are rife. Nintendo is known to be in discussion with Netscape Communications (presumably regarding licensing the popular Netscape Navigator web browser) and has been known to be pursuing the dream of online gaming for years, both as a profitable spin-off of its successful GateWay system, and as a means of replacing cartridges. (Indeed, the fact that Nintendo 64's disk drive will be writeable, and the fact that Nintendo has made a determined effort to keep the size of its game code small - they currently have to fit on cartridges, remember - both perhaps point to secret plans to take Nintendo 64 gaming exclusively online.)
Until the last couple of years, however, a number of factors precluded the playing of graphic, fast-action games over the Internet. First, despite a few proprietary games, such as those offered by the ImagiNation Network (which was started by Sierra in 1990) and the online service Genie, there was no infrastructure in place to coordinate the playing of graphic games on the Net. That infrastructure didn't develop until developers like Id included network support in their products.
Second, not enough people were online until today to justify the cost of supporting commercial games over the Internet (most MUDs are free, hosted by university systems and maintained by volunteers). "The biggest impediment to online gaming is getting people online!" says Liebowitz. "Only 5% of the population is really online, and 95% of the people who could use personal computers and modems simply don't." This is the real challenge, getting people comfortable with being online.
"And only maybe 10% to 20% of these online people play games," Liebowitz continues. "On the other hand, when you do play a game, you're online for a pretty long time. I would suspect that around 40% of the total time spent online is spent playing games." Ironically, the 40% figure is the same number quoted by Microsoft about the percentage of time home PCs are used for playing games.
Third, game configuration hassles, already legendary on the PC, increase immensely when you also have to configure your game with you Internet service provider. "Configuring the PC for the Internet connection and actually just getting things to work is a huge impediment," says Alex Beltramo, of the Total Entertainment Network (TEN).
The largest issue, though, has been that of network speed, or - more specifically - latency. Latency is a measure of not how much information can be transferred in a certain time (this is bandwidth), but the minimum time it takes for even the smallest message to be sent, received, and understood. Imagine pulling a move in Street Fighter Alpha and not having it execute on screen for half a second and you begin to understand the magnitude of the problem.
Some games work fine despite extremely long latencies - chess, or turn-based strategy games (both of which can be played by mail) for example. Others, like flight sims (which can work out - given your direction and velocity - where you should be at any time), can function with higher latencies. But fast games, like fighting games or Doom-style action games, require near-instant response. Typically, it is estimated that "arcade style" (for want of a better description) games require latencies of less than 0.25 seconds. A standard Internet connection, however, generally comes with a latency of between 0.5 and 1 whole second.
The solutions to beating the latency issue are numerous. Some networks, like DWANGO, do it by bypassing the Internet altogether in favor of local, dial-up servers. Players have little more latency than on a LAN, but if they're more than a local call away from one of DWANGO's 23 U.S. servers, they'll have to pay long-distance phone changes in addition to the standard connect fees.
TEN, one of the first gaming services that will be accessible through the Internet (it also has DWANGO-style dial-up servers) has invested a massive amount of effort in beating latency. It has an online agent, "Mr. Bandwith [sic]," who checks your connection and lets you know which games you may want to avoid. You may be restricted to playing gamers close to you geographically, to keep down latency (which increases, obviously, with real physical distance). "By doing lots of little tricks, there are lots of different ways we can chip away at the latency issue," says a TEN spokesman.
Others, however, see the latency issue as a paper tiger. Liebowitz simply doesn't see it as much of a problem at all. "Very few games require low latency," Liebowitz says. "We played Descent in test mode with a one-second latency, and the play experience was excellent. Probably only with 10% of all games is latency a factor at all, and of that 10%, really only the top players would notice a latency."
Even for these players, however, Liebowitz is confident that the latency problem is temporary. "We've talked to companies in the telephony business" he continues, "and the cable business, and the modem business. Every one of them is spending tens, if not hundreds, of millions of dollars trying to solve the issue."
Although there are still a number of technical issues to be ironed out before playing a multiplayer game online is as easy as putting a CD into your PlayStation and turning it on, it's clear that online gaming isn't going to go away.
1996's most popular PC games - Command and Conquer, WarCraft II, Duke Nukem 3D - are designed for multiplayer online gaming, as are many of the most promising titles to be released through 1997, including Id's Quake and Blizzard's Diablo. Even classics like Civilization have been revamped for online play (CivNet). And by the end of 1997, online gaming will likely be an option for console gamers as well.
What remains to be seen, however, is just who will be providing and selling the experience. While their current gaming selections are fairly lacking, can the big online services (such as America OnLine) beat the challenge of the technically-superior dedicated gaming networks (such as TEN and Mpath) by simply virtue of their large installed user base?
"The typical online services are not the model for the future," argues TEN's Alex Beltramo. "To appeal to gamers, you really need a narrowcast service, so I don't see a lot of competition coming from them."
Of course, once all the technical problems are solved (as they inevitably will be), and the market shakes down to two or three main game service providers (as it inevitably does), the emphasis will be back on those who really count - the game designers themselves. And let's just hope they're up to the task of realizing online gaming's potential.
DEDICATED GAME SERVICES
Total Entertainment Network (TEN)
(http://www.ten.net/)
With $12 million in venture capital, as well as exclusive deals with Apogee (a.k.a. 3D Realms, the Duke Nukem 3D people), Spectrum Holobyte (Falcon 4.0), and SSI (Panzer General, Renegade II), TEN (created by the merger of Mac gaming site Outland, and Planet Optigon) is poised for greatness in the gaming service world.
TEN is probably the best marketed gaming service out there, and it has been signing deals for games at an astonishing pace. By going straight to developers and evangelizing its software development kit and APIs, the company has gone a long way to solving the ease of the use problem, too - you will be able to automatically install and configure the TEN client from within any TEN supported game. The service will also include record keeping, profiles of players, and hooks for social interaction (chat boxes and e-mail, although not Internet e-mail, which is odd). The big question for TEN will be its ability to work over the Internet, with all the latency problems that come with it (there will be direct dial-ups in larger markets to increase speed). With three T3 lines worth of connectivity to the Internet, although, any delays will not likely come from TEN's end.
