Disorderly Queue

Like any other medium, videogames have developed certain conventions as they have evolved - conventions which were often first set in place by technological limitations of the time, but which can persist long after those limits have been removed by the march of technology. Gaming conventions come to form part of the language of the medium, giving designers (of both software and hardware) touchstones which they know will be familiar and comfortable for their audience.

On occasion, however, there's a sense that a convention has remained in place not because it's useful, or comfortable, but just because replacing it would require thought and effort that developers don't want to commit.

One particularly infuriating example of that exists in the MMO realm, and has had a particularly nasty airing in the past fortnight thanks to the launch of NCsoft's Aion. Much has been written about the game itself, and it's unquestionable that it's a stylish, polished implementation of the genre - indeed, it's arguably the first rival to come close to the quality mark set by World of Warcraft.

You might be disinclined to agree with that sentiment, however, if most of your experience of the game in the past fortnight has involved sitting in queues waiting for a space on your chosen server to become available - an aspect of the game's launch which has coloured almost every piece of coverage it has received.

Aion is not, of course, alone in facing this problem. Most MMOs launch with queues to access their servers - worse again, many of them then end up offering players barren, half-empty servers when the player population falls after the first few months.

The fundamental reason for this problem lies in a basic convention of massively multiplayer games. From the earliest successful graphical MMOs, such as Ultima Online, games have been divided up into "shards", or "servers" - identical copies of the same game world, each carrying a portion of the game's overall population. A game such as Aion (or Warhammer Online, or Age of Conan, or any other normal MMO) launches with a number of servers available, each of which can only handle a certain number of concurrent players.

The game service operator then faces an extremely tricky balancing act. Too few servers will result in enormous queues for the game, which frustrate and upset players - not a good move when your business model relies on them being so pleased with their experience in the first 30 days that they'll happily subscribe for the long term. Too many servers, however, results in some of them turning into ghost towns later on, as the initial surge of interest in the game slows after the first few months.

When the uptake of the game is higher than anticipated, companies can respond to the demand by opening new servers - but this is a blunt, unfocused tool at best, and in many cases, is utterly useless as a solution. Players who have already invested time in a character don't want to start afresh on a new server. More importantly, players who have started the game with a group of their friends will find it very difficult to move en-masse to a new server. Instead, you see the spectacle of a game with a cluster of servers afflicted by huge queues, while new servers sit half-empty.

In terms of the problems caused by the "shard" model, however, this is only the tip of the iceberg. Shards were an effective solution to a technological problem in an era when MMOs were a niche, minority interest. Back then, the chances were that you only knew a handful of people who played MMOs, and arranging to play on the same server as them would be fairly straightforward.

Today, however, MMOs are an important and growing part of the mainstream gaming market, which has itself become a vastly more pervasive and widely adopted hobby. Many people starting out in an MMO will have several groups of friends who are playing, or thinking of playing, and the chances are that they'll later meet other people who play the game. Arranging to play on the same server as all your friends is next to impossible - and discovering that you are already on the same server as a new player you meet offline is extremely unlikely.

Needless to say, this puts a dampener on the idea of playing an MMO as a social game. Bizarrely, it makes more straightforward multiplayer games such as Halo or Call of Duty into more fundamentally social experiences. Certainly, you can only play Halo with a few dozen players at a time, but crucially, you can play against any selection of players from around the world. Any friend who owns the game can hop online and play with you - while with World of Warcraft or Aion, the chances are that if you discover that your friend is playing, you won't be able to play together anyway, since he'll be on another server. End of conversation. You both play a multiplayer game, but there's simply no direct mechanism to allow you to play it together - which flies right in the face of the whole concept of online gaming.

These issues, from Aion's queues to the frustration of being unable to play with friends in WOW, are made all the more annoying because they no longer have a technological justification. Today, it's eminently possible for a game to be created in which a character can easily hop from server to server, playing with whichever group of friends he or she sees fit. It's not hard to envisage a system whereby you log into your character first, and then select a server based on the people you want to play with at that particular moment - instead of logging into the server and picking your character, who is tightly locked to that server, as occurs in games today.

It's been a long time since I could lay claim to the job title "programmer", of course, and there are indisputably both technical and design issues which such a system would create. However, it's increasingly apparent that sticking with the old way of doing things is not a realistic option.

How much further could World of Warcraft grow, if people could simply play with their friends whenever they wanted? How badly damaged will Aion be by the negative perception created by its overcrowding problems? This isn't simply a question of players being a bit annoyed - it's a question of MMO operators leaving vast amounts of subscription revenue on the table, simply because they're wedded to a convention that's over a decade old and no longer justified on any technological grounds. Any developer working on a new MMO who's still stuck in the rut of forcing players to pick servers and stick with them needs to seriously reconsider - it may be "how we've always done things", but this is one tradition that demands to be put to rest.