Since 1991, I have been running play-by-mail (PBM) rôle-playing games. It started as a way of getting in more gaming than my friends' and my limited schedules allowed; however, it has grown to take on a life of its own. In this article, I intend to deal with the difficulties and rewards of this sort of gaming, as well as providing suggestions for anyone else who wishes to do the same thing.

My style of running PBMs is markedly different from that of the usual commercial PBM games. For a start, there are no turn deadlines. Players put in turns whenever they can; I write the results and send them out, typically around a week later. This is extremely helpful when players are busy, since there is no pressure to get turns in on time; also, as I don't run the games full-time, I can't guarantee fast responses at busy periods. Open deadlines also give a very good measure of how interested the players are in the game; and if players need to drop out for a short time, for example due to work pressures, the system can handle this easily.

A further advantage of this method is that both players and GM have plenty of time to consider responses to actions; this is good for inexperienced GMs, giving them time to look up rule or background points that may be in doubt, and allows more complex situations than would normally be possible in a fast-paced face-to-face game.

Most PBM rôle-playing games I have seen tend to centre on the party as the active group. However, I have found it better for rôle-playing to have each character separated from the others, and probably not even aware of their existence. This has a great many advantages; it means that different players can play at different speeds, the GM doesn't have to wait for every player to send in a turn before writing their results (or penalise the slow ones), there is no feeling of competition with other PCs (something I feel has no place in rôle-playing), and, perhaps most importantly, the players can be doing wildly different things.

After all, there are some types of adventure that are very hard to run with a group. Espionage is the most obvious example - there's always one player who wants to be James Bond when everyone else is George Smiley, or vice versa - but many types of scenario run more smoothly when only one player character is involved. (It's also easier to find or steal ideas for a single primary character than for a group, since this is the orientation of most books and films.)

Rather than trying to remember which PC is due for the limelight this week, every single adventure can be tailored to the individual player and character. Naturally, this is especially useful when dealing with inexperienced or shy players. This mode seems to work equally well with dirigiste or free-form adventures; the free-form mode is remarkably advantageous in play-testing a campaign world. since in my experience players have often come up with totally unexpected avenues of exploration that would never have been mentioned in a multi-player game.

Of course, there are disadvantages to the single-PC game, the most prominent being the lack of interaction between players. The cure for this is simply to provide good quality NPCs in large numbers; this isn't quite such a hard task as it may appear at first, as sufficiently mobile NPCs can be used with multiple PCs, and may even mention PCs to each other. The relaxed schedule also allows the GM to consider NPCs' words and actions very carefully. In some games I have also encouraged players to contact each other with game-related messages. Since the players are not competing, but co-operating, they are usually prepared to share information. However, since the PCs don't usually know each other in game, this does break some of the mood; friendly NPCs, or an organisation for which PCs work, are probably a better method for passing on information.

A seemingly minor idea, but one that seems to help a great deal, is to keep a campaign calendar. Apart from helping the GM co-ordinate and keep track of the actions of different players, it seems to add an extra layer of realism when each turn has the time and date at the top. No special system is needed; I simply keep a file of dates, locations, characters and short descriptions of events.

Another helpful idea is a campaign newspaper. A full newsletter, as seen in professionally run games, is a great deal of work to produce. However, since PCs are often at different time points in the game, I find it easier simply to keep a list of news items, and the dates at which they become generally known, on the campaign calendar; as each player reaches the appropriate time, the news item is included in the turn. "You see in the paper that...;"

There are, however, problems in combining this with the free-form scheduling mentioned previously, in that the character furthest ahead in time will not hear about the actions of other PCs. I have not found an easy solution to this that maintains free-form scheduling. It would be possible to insist on direct correlation between campaign date and real-world date, but this seems excessive; the situation of one PC being far ahead of or behind the others tends not to persist. If it begins to cause problems, the easiest solution is to modify the length of the characters' next inter-adventure breaks, so as to allow some degree of re-synchronisation.

As for writing style, this should be whatever the GM and player feel most comfortable with. I find a second-person narrative works best for me. Given the amount of text involved, however, it is vital for the GM to use a style he enjoys.

Many players seem to have some difficulty in working out just what is required in their turns. I usually suggest at least half a page of text in a turn, dealing if possible with multiple situations. Sample turns are also useful to show to beginning players. A typical turn response is usually at least a page, and can be longer; my longest to date has been four sides of A4.

One of the largest potential problem areas with this system is combat; the game systems I favour tend to have complex combat rules, and even in simple systems combat usually takes some time to resolve. The solution I have adopted is twofold: first, to run games that are comparatively lacking in combat (which I feel does not necessarily detract from game quality); and second, to ask for a general statement of intentions at the beginning of a combat, which should include conditions for fleeing, preferred modes of attack, and so on. In this way, a combat can usually be resolved in one or two turns, only taking longer if something unforeseen occurs.

This tends to make formal "set-piece" combats rare events; what happens more often is a rapid exchange of gunfire from ambush, after which at least one party flees. This has provided an unexpected bonus, as it has helped provoke realistic (combat-avoiding) behaviour on the part of PCs.

This principle can be extended to other interactive situations, in particular conversations with NPCs. To avoid endless series of turns consisting of "I ask him about X" and "He replies with Y", I suggest that players list several topics of conversation - then the character's interpersonal skills are used if necessary to resolve interaction. At least, that's the theory; I have found, particularly with good rôle-players, that conversations can drag out for some time. This is not in itself a problem, but the GM should be alert for signs of incipient boredom.

Good record-keeping is vital to the game quality. As well as the general timeline file mentioned earlier, I keep a complete record of all turns sent and received, and any general notes necessary.

Recruiting to the games can be tricky. I have used a combination of word-of-mouth recruiting from current gaming colleagues, and the occasional Email advertisement. However, this is still a problem area. An advantage, though, is that only one player need be recruited for a game to begin; others can always be added later.

One possible point of contention is payment. Since I am running these games on an amateur, part-time basis, I do not charge for the games themselves; since most of the games have in practice been conducted by hand transfer of turns or Email, I also have no need to charge for postage. However, I do ask that players pay for my postage as well as their own, while I pay my own printing costs.

In summary, these amateur PBMs can provide a very rich role-playing environment, allowing concepts and places to be explored that would be unlikely to be reached in conventional games. I expect to continue to run them for the foreseeable future, and can thoroughly recommend them to all potential GMs.

Roger Burton West