How about Archaean, Hero's Journey and Shadowbane? Or Ariandora, The Awakening Project and Ascension, Spectrum of the Shadow, Fireflight Eve and Lineage? (Sorry, that last one's in Korean).

The fact is that there are currently around 40 massively multiplayer graphical online role-playing games being run commercially, plus another 25 in beta and 35 in development - and those are just the ones listed on the Multiplayer Online Games Directory. I know of at least twenty more that have got the finance and are well into being coded.

Now of course, some of these games are actually being programmed by fresh-faced youngsters in their bedrooms, and others are by small development houses buying in a 3D engine. More, however, are well-financed, professional productions by well-known and respected developers eager for a piece of the action. The top ones have development budgets in excess of $8,000,000.

Yes, I thought that might wake you up.

Here's why they're willing to invest so much: Ultima Online has around 130,000 subscribers or so (depending on who you ask); Everquest has over 200,000. At $10 per head per month, that's well into Oh My God territory - and it's not even counting the boxed sets you sell at $35 a pop to each prospective player before they can begin to join in the fun. We're talking severely serious profits here.

That explains why companies are keen to get into the online game market, but it still doesn't justify the massive budgets. To understand why some games are nevertheless attracting them, you need to know a little bit more about how online games affect their users.

Essentially, the first massively multiplayer shared world game you play is the one by which you judge all others. You've seen nothing like it before, it's staggering in its scope, mind-blowing in its concept and utterly addictive. This is irrespective of whether or not it's garbage. There are over 3,500 MUDs in the world at the moment according to MUDlinks, and at least 3,400 of them are drivel compared to the remainder. To the first-timer, though, they simply surpass anything they've seen before. By the time the players realise there are other games out there that are better, it's too late - they're tied by bonds of friendship, loyalty, experience, knowledge, power... People don't continue to play Everquest because it's a good game, they continue to play because their friends do.

The Jesuits used to say, "Give me a child until he is 7, and I shall show you the man". If the experience of MUDs is anything to go by, 7 weeks is more than adequate for players of online games. It even extends to writing games: most of those Everquest clones in development are being written by people whose main experience of online role-playing games is (you guessed it!) Everquest.

So what we're witnessing is essentially a land grab. Once you've got a player, you get to keep them until you screw up (or even if you do screw up - ask any long-term Everquest player). Internet usage is expanding all the while, but it can't continue indefinitely. The time when there are hordes of newbies starting up and looking for games to play may now be upon us, but it won't last for long. When it does, the theory goes, you'll have a hard time prising players away from their "homes" no matter how fantastic your game is.

Even traditional MUDs are getting in on the act. Medievia recently took out a full-page ad in Dragon magazine (i.e. Dungeons and Dragons monthly), paid for by their players. The Skotos people put their brilliant "Why yes. I am God" ad in Knights of the Dinner Table (a comic widely read by role-players). Avalon has been running a successful commercial MUD for years by the simple expedient of capturing players using well-placed web ads before they find out that 99.5% of MUDs are free. Even when most mainstream people expect computer games to look like Quake III, if you can capture newbies for a text MUD you still get to keep them.

Ah, but how do you capture those newbies? Competition means it's no longer OK just to put "something" out fast - new text MUDs and simple-graphics games (à la Codemasters' The Realm) are heading straight for the niche market. In today's climate, you have to have one or more of the following:

• a massive advertising budget (e.g. Asheron's Call)
• a strong reputation from single-player games or other multi-player games (e.g. Ultima Online)
• a large development budget (e.g. Ultima Worlds Online: Origin)
• a big licence (e.g. Star Wars)
• big name backing (e.g. Sony's Everquest)
• lots of pretty, moving pictures (e.g all of the above)
• a huge, captive audience (e.g. Gemstone III on AOL was making $1,000,000 a month before AOL went flat rate)

All of these, except perhaps big name backing, can be bought if you spend enough money. The more of them you cover, the more players you'll get - but the more it costs. A publisher wishing to grab a big slice of the online action would take one of their existing products (e.g. Legends of Might and Magic) or licences (e.g. Star Wars), tag the best 3D engine in their stable to a simple MUD-based server, and spend as much money on three-page fold-out ads in magazines as they can. They'd have to invest heavily in infrastructure too - those rumours of Ultima Online once having more incoming bandwidth than New York City are basically true - but although that may be a necessary condition for success, it's not alone sufficient. Indeed, as writing a game is one thing, but maintaining the servers and Internet connectivity is another thing entirely, it may be wise to pay someone else actually to run the game. Kit costs.

So does this mean that if you haven't started on your persistent world yet, you're too late? Well, not necessarily. This may be a land grab, but you can't grab much of it at once and you don't want to grab any that's unproductive (amazingly, not everyone likes online games). 200,000 players may sound good at the moment, but it's just a fraction of the potential audience. It's possible that a handful of games could eventually see 1,000,000+ players each, but at that stage there'd have to be a risk of fragmentation. In multiplayer games, players only really ever interact with about 250 others at most: 250 is the size of "their" community. It doesn't matter to them if there are 199,750 other people playing the same game, so long as they can play it. In fact, since the game is probably spread across multiple discrete server clusters, they most likely can't interact with the bulk of the playing population anyway.

It's therefore possible that you can find what kind of gaming experience communities of players want, and target them. A game with 1,000,000 players can't keep them all happy the whole time; if you can find something that will keep some of them happier, you may be able to wrench the players away a community at a time. Of the glue that binds the players to a game, only friendship with other players is permanent - game loyalty is chipped away by cumulative administrative errors (real or imagined), and knowledge/experience/power gradually evens out as more players plod their way through the ranks.

So how will things look a few years down the road? Well there'll still be the big, prestige games, and there'll be lots of spin-off games by the same companies who are trying to ensure that if they do lose players, at least they'll lose them to their own games. However, there will remain plenty of opportunity for smaller games to gain a foothold and grow - if they target their prospective players properly. Some of this movement may even come from within the communities themselves: most of those 3,500 text MUDs were created by players who grew disenchanted with their previous MUD and wrote their own (for their community) using ready-to-go engines.

So in the end, we may have something akin to a small number of "black hole" games that suck in newbies and keep them from escaping with their strong gravitational fields, but which in time spew them out into another part of the universe where they can coalesce and form new games of their own.

Don't underestimate the niche market. Remember: more people play text MUDs than play Everquest.

Richard Bartle created the first Multi-User-Dungeon (MUD) in 1978 with Roy Trubshaw, heralding the beginning of online gaming. The game has undergone numerous incarnations, and is still running more than 20 years later. Dr Bartle is the programming director of MUSE, and head of online games at Gameplay.com.