Many of us started gaming at school, when getting a few friends together on a regular basis was a fairly easy proposition. It's less easy now; we have other commitments, building a world is time-consuming, and getting a six- or eight-person group all to be present can be practically impossible. I hope to demonstrate how a new model of player/GM relationships can make the running of a long-term game easier in such circumstances.

The traditional gaming model has a single game master (GM) who has sole control over the world and its contents. This is, of course, quite hard work; moreover, players can feel excluded from the world-creation process, and may be hesitant to run their own games due to the burden of creation involved. Some games have tried to modify this, encouraging players to construct world sections, produce game-related props, and so on; however, this has always felt somewhat forced, particularly when it is connected to the giving of in-game rewards.

The premise of distributed gaming is that there is no sole authority over the world; any player may stake out and develop a section of it. There is a co-ordinator role, but this should be less demanding than solo world-building.

Distributed gaming is not intended for initial world layout; the basics should be set up in advance, either by a single designer or by a group, to ensure that a strong flavour is maintained. However, details can and should be left open. The aim should be to produce a short document that gives a good idea of the type of world being built and the sort of games that will be run in it. The background sections of most traditional games, as opposed to later supplements, would be a reasonable top end of complexity.

Once this is done, the distributed gaming set-up can commence. Members decide among themselves who will take which areas: not all areas need be taken at first. Generally, about three areas per member seems to work well: for example, in a cyberpunk setting, one might choose one geographical location, one corporation, and one "other" area (such as the Net, or how the financial system works).

Following this, players generate characters. They may, indeed should, generate more than one, to allow participation in a variety of adventures. Character generation should be slightly restricted, in that PCs should be able to work with each other on short acquaintance; this lends itself to the idea of a professional organisation of some sort, which hires out teams of competent specialists to its clients, and the sort of PC who would be prepared to work for such. If the settings used are geographically diverse, PC should also have reasonable mobility.

GMs will typically run games that deal with one of their own areas. Games should be designed in an episodic manner, such that one session is long enough to complete one adventure; at the very least, the end of the session should be in a situation in which PCs may leave, and new ones join, the party. This is not to say that there cannot be overarching themes, but they should not rely on individual PCs.

Games are scheduled on the world's master calendar. The game date may not be retarded, though adventures without major consequences may be played retroactively. The calendar is split into units of a convenient size; for a modern game, this might be a week or so, while fantasy games might work better with longer periods.

A typical session starts with the members present deciding who is to GM; the GM then specifies the date of the game (the earliest free slot on the calendar) and the sort of characters wanted, whereupon players select their characters from those available and play proceeds normally.

Normally only one game takes place per calendar period. The GM for that period is responsible for writing up a report, in a manner appropriate to the game world, which should include any news generated by the adventure. It should also carry other events as decided by the GM; it may also have items that rise from the aftermath of earlier adventures, or that foreshadow games to come. These latter two should be provided by the relevant GMs; the period-GM acts as collator and editor. The news report should be distributed to all players via the co-ordinator, who also keeps the master copy of the game calendar.

It is possible for multiple games to take place in a single calendar period, though this should be rare, as it invalidates the separation and GM responsibility that the period system provides. Two exceptions exist to this, however. Introductory games may be played detached from the main timeline, and attached later; they should not have any effect on reports. Neither, normally, should private games (one player, one GM), typically dealing with a PC's life between adventures, which are likely to overlap with conventional games involving other characters. (A powerful PC, taking actions which will have a significant effect, should of course be scheduled on the calendar as a standard game.)

Players are responsible for their own character sheets and notes, since they may play with a number of different GMs.

Once an area has been developed, the background material for it should be made available via the co-ordinator to other GM members, who can then use it in their own games. Responsibility for further development, and integration of new and contributed material into the background document, rests with the original owner, who has overriding authority for the area.

Members may at any time take responsibility for an unclaimed part of the world; they should confirm this with the co-ordinator, then start developing. Players who wish to start GMing should find a significant amount of source material available, and may easily run a game without needing to do much world-building or waiting for players to generate new characters.

The co-ordinator should decide, with members, how to allot in-game rewards such as experience points. This need not be rigidly fixed, but should remain fairly consistent between GMs.

GM members may of course conspire to create larger plots, in the manoeuvres of their power blocs. The principal weakness of distributed gaming is that long-term plots do not flow out of it as easily as with a single-GM system. In particular, it is hard to have a secret that affects a large part of the world; however, a secret that is restricted in scope is easily set up, and might be discovered over an extended period by a number of different groups.

It may be appropriate to rule that the co-ordinator is the final arbiter of the world; for example, if the world was originally designed by the co-ordinator and is now being made available for distributed gaming. The co-ordinator may never overrule a GM running a game, but may confer with the GM if the GM seems to have missed something vital.

Overall there is potentially a three-tier system, with the co-ordinator, GM members and player members. However, in practice it seems likely that most players will wish to be GMs at some point; moreover, the job of the co-ordinator could be exchanged among members.

The distributed gaming system obviously requires a fair amount of trust among its members, but seems well-suited to the development of game worlds and long-running games. I hope that it allows groups with limited time to continue to play and develop RPGs.

Roger Burton West, with thanks to Sean Desmond