the online game Hat

the online game

The hype has died down but the work has only just begun. Half a decade into the much heralded 'multiplayer gaming revolution' and still no-one knows exactly where online gaming is headed. Edge charts the possible courses

In 1996, the computer industry market analyst Datamonitor predicted that the European online games market would be worth some £900 million in the year 2000. Yet 12 months later it revised its forecast to £600 million by 2002. If even professional industry watchers are confused, what chance has the average gamer got of understanding the online market?

Anyone who's played anything from Quake to Hearts on an office LAN knows where the buzz surrounding online games comes from. Playing with other human beings illicit a thrill that conventional games can only rarely manage. This has been obvious for years - witness how Gauntlet and Street Fighter II quickly became spectator sports when they first appeared in the arcades. Indeed, the Japanese make no distinction between multiplayer games, connect 'em ups like Pokemon and Internet games, which underscores the fact that the presence of the internet merely ups the ante in the online stakes.

Screen shot
Ultima Online

Today, enough entrepreneurs have been won over by the promise of Internet gaming to continue throwing money at what remains an embryonic industry. For now, these players can be broken up into four distinct factions: Multi User Dungeon (MUD) operators, the so-called 'aggregators' (such as Wireplay, DWANGO and Thrust World), Internet Service Providers and individual software publishers. When considering this proliferation of different online services, it's important to remember one thing. Compared to the overall size of the boxed games market, online is virtually non-existent. The odd commercial success like EA/Origin's Ultima Online is the exception that proves the rule that hardly anyone will pay to play online.


The rules of the game

Richard Bartle invented the first Multi User Dungeon back in the early eighties while teaching artificial intelligence at Essex University. Here are his golden rules of online games design

A great game is not enough
The players are the point, not the game. A coin-op simulation may suffer in comparison to a technological no-brainer like Hearts.

Being thrashed is no fun
The debutante Quake player who finds himself thrashed into submission will probably never return.

Users must be able to wander about
Not everyone is brilliant at gaming, so a truly democratic crime experience should allow all players to take part at their own pace.

More isn't always merrier
Where's the joy in a community of thousands? Bartle suggests over-populated games should be sub-divided into groups numbering not more than 250 people.

Not just playable but replayable
If there's a logical conclusion to your game then users will finish it and move on. Even worse, shared knowledge means that new recruits will take progressively less time to get through.

Don't fix times for play
Despite failing phone charges and ISP rates people are still aware that the clock is always ticking. Players must be free to enter or leave the game at any time and still enjoy the experience.

Make it easy to meet other players
To reiterate, online games rely on player interaction. It's best to make interaction central to the game concept.

State of the art doesn't stay that way
A technological marvel in the offline business can only stay that way for a few Months. Low technology online games might go on for decades.

Online games need love
Because online games should be organic they need someone to look after them. Work on an online product never stops.

The most popular MUDs have a few hundred members. Wireplay claims 23,000 registered users but is lucky to have even one per cent of them playing at any one time. Ultima Online (flawed, but arguably the premier example of a tailor-made online game) is reckoned to attract up to 3,000 players at its peak.

But stand these next to the casualties and the confused state of online gaming is apparent. British startup E-On folded last November, US-based Internet games service TEN laid off 80 people a month earlier, and rival Engage reduced its staff by half in August. Jeff Leibowitz, president of Engage, said at the time: 'It's simply a fact that the market isn't there. The internet market just hasn't developed the way anybody expected it to.'

However, Colin Duffy, Wireplay's head of games, sees the sector at the start of a classical economic curve. 'All new products, whether TVs, PCs or videos, pursue the same curve. You start at a crawl being ignored by all but a few early adopters. Slowly the mass market catches on until the curve rises sharply and you're into rapid growth. I'm certain online gaming will be no different. We just don't know how long the process will take.'

Duffy's case is persuasive, but the cynical observer might feel the online market has had its chance. The Internet has been in the public consciousness for three years - five for the more clued-up gamers. It's almost as long since Doom first set office LANs ablaze at lunchtimes. The path from early text-based MUDs to Carmack's stunning Quake code hasn't always been smooth, but the public have had plenty of opportunities to be tempted online.

Arguably though, Quake World is a new phenomenon, the biggest success so far for online games. And it's achieved an unprecedented level of success in spite of one huge and persistent handicap - latency. Like chewing gum on the shoe, the latency problem (or lag or ping) is an irritant that follows wherever remote multiplay goes.

Latency describes the time delay between a player's actions and registering that action on the screen. It's caused by the need to communicate the action to the central server which runs the game (or to each individual player in games without a server). Latency affects games differently. While irrelevant to a turn-based game like Bridge or Chess, latency can render 'twitch' games like Quake unplayable - when one player's screen freezes, the other sees an immobile foe and fills him full of bullets.

Latency isn't an issue on office LANs and only rarely a problem when two players connect via a modem. But across the Internet, latency hits hard. The journey of every single packet of data can be hindered by connection quality, by bottlenecks due to narrow bandwidth, by error checking at the modem and by complex features running within the game itself. There's also the perennial problem of users' different machine specs. And even the speed of light is a bugbear if the signal goes international.

The consensus seems to be that no more than 150 milliseconds leg is a reasonable target. It is achievable - though not always. According to Pete Hawley, a producer at GT Interactive and respected Quake World player, latency often makes Quake unplayable. 'It depends on your Internet Service Provider, but I'd say 50 per cent of the time there are problems with ping times. Players tolerate it because of the competitive element in the game but, really, it's pretty frustrating.'

Screen shot
Diablo on

Developers have come up with various software techniques to minimise the tyranny of lag. VR-1 uses a predictive motion algorithm to calculate the likely path of objects on screen for smoother movement. MPath's MPlayer monitors the player's connection and displays red or green to indicate whether it can handle the requested game. Meanwhile Jack Mathew's QSpy finds the fastest available Quake World server and re-routes the player to it.

Screen shot
id Software's Quake

Of course, latency is not the same problem for every player - causing all manner of 'Netiqette' fury. Play by ISDN and ping times are reduced, giving those lucky enough to have this faster access - LPBs or low ping bastards - something of an unfair advantage in play. (An ISDN is a dedicated, purely digital, additional line to your local telephone exchange).

But don't put too much faith in ISDNs or in new modem technologies as the big solution. Over the next few years, a number of 'broadband' devices will enter our homes, such as ADSL and cable modems. These will provide enormous increases in bandwidth and should revolutionise the delivery of music and video across the internet. But they're unlikely to reduce latency to any greater degree than an ISDN line. Broadband modems may offer more data but not necessarily faster delivery of this data. Only improvements in the Internet's underlying architecture can seriously crush lag.

Needless to say, latency is a crucial issue to aggregators who need to attract gamers away from 'free' sites such as Quake World servers. Hence, DWANGO's claim of 100ms ping times and Thrust World's 135ms. Wireplay, BT's dial-up service which runs across the public service telephone network to a dedicated server, projects itself as a nationwide LAN. Unhindered by many of the Net's obstacles the service claims 105ms ping times on Virtua Fighter and 65ms lag in the lab.

The other key issue with latency is whether the solutions to it are scalable. If 100ms can be achieved with ten users can it be sustained for 1,000? At some point, the cost of solutions like adding extra servers becomes prohibitive.

The fact that latency excites such intense debate hinges on the presumption that everyone wants to play 'twitch' games online, But there's simple argument to suggest that they don't. In a 'twitch' game the same players usually win. And while it's fun to be an ass-kicker, it's no fun to be the ass kickee.

This is why there's a vocal caucus within the online community demanding a new type of gameplay. Just as storytelling on film is different from storytelling on stage, they reckon gaming needs to be considered differently when on or offline.

For many, community-based persistent worlds are the template. Daniel James, head of MUD creator Avalon, claims that, 'MUDs are the historical backbone of the industry, Their features are present in most online games. People play them for years and when they sign off it's like leaving home.'

Although Ultima Online and 3DO's Meridian provide an illustration of how to make persistent worlds visual, the text format still has its adherents. Steve Cooke, of Ogalala, raves about the inclusive nature of words. 'Show non-gamers an arcade game and they have all kinds of preconceptions. They don't with text. And if you can persuade them to play for five minutes they get hooked. With improving voice recognition, text has the potential to define a lot of online gaming in the future.'

Others believe MUDs as they stand now are a cul-de-sac, defined and confined by their RPG roots. Pete Hawley: 'I see their appeal but they're not for me. You walk in and someone says 'thou hast entereth' and you just think 'Christ, no thanks'.'

Jez San, managing director of Argonaut, is another doubter despite being 'Jez the Wizard' in the original MUD. 'The bullying by experienced players in MUDs is just as serious as it is in 'twitch' games like Quake. And gameplay is flawed because I think most gamers want to be able to go online and feel they've played something to the finish.'

Screen shot
Warcraft 2

Of course, MUD proprietors can make one great claim - they're profitable. Clem Chambers, pioneering founder of Online Plc maintains that every game his company makes recoups its costs within four weeks.


Online options

MUDs/Persistent worlds
Multi User Dungeon operators have two distinct advantages over their rivals in the race to shape an online future - they've been doing it for years and they're making money. There are said to be around 800 persistent worlds on the Internet, many located within ISPs' own domains. Most are text-based like the original MUD but there are some visual worlds like Ultima Online.

The aggregators
This unwieldy term has emerged to describe companies which act as a single site for online games or a dial-in network for boxed games with a multiplay option. The UK's 'aggregators' are:

  • BT Wireplay
    operating since October 1996. On Wireplay, players load up their offline game and click on a Wireplay icon to dial straight into the server and meet other players. The proprietary software can be downloaded from cover disks or the website. Wireplay offers over 30 games and charges 2.5p a minute evenings and weekends, 6p all other times.
  • Thrust World
    Unlike Wireplay, Thrust World is Internet based. Its servers match players who pay £17.99 a month for unlimited usage (which includes other internet access).
    The US service which made its name with Doom death matches is beta testing in the UK. It hasn't finalised a pricing model but in the US charges $1 an hour.
  • European Gameszone
    A brand new gaming service set to launch in the UK, Sweden and Germany which promises little lag, chat zones and a service tailored towards Europe.

The Internet service Providers
Games sites are usually available on an ISP's own service and they are generally the most popular destinations among subscribers. However, publishers are generally uncomfortable with being so dependent on the ISPs. In fact, 3D Realms withdrew from America-On-Line when the service asked for a fee. There is a belief that ISPs prefer customers to be web browsers than gamers because it makes data traffic easier to manage.

The Software publishers Software Publishers have the cash and resources to Produce multiplay software which could forge lucrative new markets. But they also have a vibrant offline market and a strong retail base which they do not want to upset. This makes online a risky proposition especially since, as one developer puts it, 'online games publishing and conventional games publishing are about as similar as movie making and sheep farming.' All the same, online-only games are coming out, with Sony's Tanurus, Red orb's Extreme Assault, Subspace and Sole Survivor - both from Virgin, all recently making headlines.

Chambers believes the ability of these sort of games to attract and keep players is underestimated by aggregators and ISPs right now. 'Content is queen and wires are king at the moment,' he says. 'One router company is worth five times the value of all Net content providers put together. But this will change. it happens in all media.'

Screen shot
Studio 3DO's Medidian 59

If Chambers is right then online will be all about content in five years time. And because of the nature of the Net (there are no warehouses, or lorries, no distribution or point-of-sale worries) everyone thinks they can control the supply of content.

Some software publishers privately admit that they, too, aim to host their own sites rather than pass games, and potentially large profits, over to the aggregators. It's certainly a tempting opportunity for them - players buy boxed games and then pay a monthly site fee, with the added bonus of all that data mining to boot. Yet there are problems, not least of which - as Ultima Online has shown - are the ongoing costs of keeping abreast of the constant emergence of hacks and bugs. And what if a publisher has a big online hit, sets up a number of servers to cater for the demand and then follows up with a series of flops? Idle equipment costs money.

Naturally, Wireplay's Colin Duffy, speaking for aggregators, has further doubts about the viability of such publisher-led ventures. 'There are 20 million Websites out there. Just because it's easy to set one up doesn't mean anyone will visit it. Wireplay is a branding exercise. Our aim is to be a magnet for gamers, like the online version of Dixons.'

For all the bluster and partial predictions of the online community, ultimately no-one knows yet which business mode s or gameplay styles will predominate. But go along to any online games conference and it's immediately clear that there's a burning passion to make online happen. Keep your phonelines free, because the revolution is still coming. Just as soon as someone finds the right map.


Pay to play

Much of the debate about online gaming centres on how to make money. This currently boils down to the choice between subscriptions or pay-per-play.

Both arguments have their merits. If subs are used clock watching is prevented, collection is made easier and users can budget accordingly. These factors convinced Thrust World to adopt the subs model. Geraint Bungay, head of marketing at Thrust World, says, 'The trouble with pay-per-play is that it's not conducive to immersion because you always have one eye on the time.'

Yet pay-per-play proponents (notably Wireplay), argue that their system is fairer and doesn't alienate new and uncertain gamers. Having said that, Wireplay has now introduced a credit card paying system to run alongside its original paying model.

A couple of companies have proposed fresh solutions. Scottish developer, Vis Interactive, mooted a new approach for the online version of its fighting game HEDZ. Players would win or lose their HEDZ (basically their stock of fighters) in one-on-one combat. If they ran out of HEDZ, they'd have to buy more from Vis. Sega's similar, but more general approach, is Limited Edition Digital Objects, or LEDOs for short (see E53). Here players purchase the core game, then buy extra weapons, vehicles or equipment as they see fit. Of course, all these financial models struggle against the free gaming on offer at, say, id's Quake World and Blizzard's But as Wireplay's Colin Duffy says, 'Quake World is awesome. But it only works because it's Quake. Anything else and people wouldn't put up with it.'

Screen shot
Vis Interactive's Hedz


Imp: Argonaut's online gamble

For Imp, Argonaut has gone against the received wisdom that online games must be communal and persistent. Jez San, Argonaut's managing director, explains: 'I don't believe people want a substantially different or inferior kind of gaming on the Net.'

San claims that a 3D game - heavily inspired by the seminal two-player C64 game, Spy V Spy - combines elements of classic arcade amusement: 'A bit of Pac-Man, some R-Type, some Domberman'. And although San is making no claims for originality he believes the game will transcend latency to a degree never seen in online 'twitch' games before.

'Most Net ready games you see now are actually developed for a LAN which is why you get lag,' explains San. 'We thought about the problems of the Net right from the start. Instead of gunfire you drop bomb to kill each other and there's necessarily a delay involved. And during it the server can synchronise all the clients.'

As for bullying, San says it's important to allow players to play in small groups and choose their own level of opponent. There are also chat rooms which enable players to share experiences after play. For now, distribution isn't finalised, although it will sell as a box but may not include a single-player game.

Richard A. Bartle (
16th May 1999: edgapr98.htm