There is an idea floating around the role playing game world that went unchallenged for a very long time. Rule books for many games have described an approach to play that we almost take for granted. Yet if we were to stop and consider what it was we believed, we would almost certainly realize that it was internally contradictory, impossible on its face.

Most readers will agree that in a standard role playing game, the referee, or game master, has complete control over the story, and that the character players have complete control over their characters, who are the main characters in the story.

As far as anyone knows, Ron Edwards was the first person to point out the problem in this idea. If one person has full control of the main characters in the story, how can another person control the story? The story is presumably about what the main characters did. If the character players have full control over what the characters do, then the referee cannot have control of the story; conversely, if the referee has full control of the story, then the players don't really have any control over what the characters do.

Although Ron Edwards is the designer of quite a few games, most notably Sorcerer, Trollbabe, and Elfs, and is a recipient of the Diana Jones Award for his contributions to role playing, he is probably best known for what is popularly called GNS. Although that is very important to role playing theory, it is not part of this article; indeed, it is only a small part of Professor Edwards' own work, although it has been the most discussed portion.

Probably this conflict, which has been dubbed The Impossible Thing Before Breakfast, does not appear to be a problem to you. Most gamers respond to the assertion that these texts are in direct conflict by claiming that they are not, because you must understand them in context. Every gaming group that is functional has found a means of resolving the conflict, and most gamers will happily tell you what the text really means. The problem is that what the text really means depends entirely on who you ask. There are at least four ways of reconciling these two concepts. These are expressed as play styles, specifically as referee styles, the way referees run their games and players respond within them. The four methods are not entirely compatible with each other, to the degree that players transplanted from a group that uses one method into a group that uses another may find it very difficult to play, because he doesn't understand what he is expected to do.

The hobby would benefit immensely from clarifying this confusion. Unfortunately, to some degree the major publishers are in a quandary. Were they to state which of these approaches to play they intended, they would undoubtedly alienate the majority of their player base (not one of these represents the majority of players). Further, because the confusion has existed so long, many rule books and supplements for major games have used different approaches in different places. A game may illustrate one approach in its core rules but use a different approach in its modules, without indicating that this is different. Fortunately, independent publishers and theorists are addressing the problem through new games that recognize the need to clarify this distribution of credibility more precisely.

The first common referee style is known as Illusionism. For reasons that should become apparent, this style is generally regarded as dysfunctional. The illusionist referee takes the statements that he is in complete control of the story very seriously, and he controls the story completely. Whatever the players do he adapts and counters as smoothly as he is able so that his story will be told; but he never lets it be known that he is doing this.

Meanwhile, the players are unaware that their actions have no impact on anything. They sometimes succeed and sometimes fail, and it all seems to fit together. The amount of power the referee has to control events through the credibility given to him enables him to completely hide the fact that nothing the players do will ever matter. Yet if you were to catch the illusionist referee off guard and honest, he can tell you what fate awaits each character, what is going to happen to them and to the group, how different story lines are going to develop and ultimately resolve, who will win and who will lose, where the final encounter will occur, and possibly even which character will deal the fatal blow. No one dies in this game without the referee deciding it; no one succeeds without him willing it; no one fails without him directing it.

The critical distinguishing factors of illusionism are that the players have absolutely no impact on what is happening in the shared imagined space, but they don't know it. The referee tells them what to imagine. They make suggestions for what they want to have happen, but he guides these into his plan, allowing those he can accommodate and negating those which would derail his story, all the while making the players feel as if they are contributing when they are not.

When illusionism is recognized, it frequently results in the dissolution of the gaming group. Many players actually want to impact what happens to their characters, and the knowledge that they cannot takes out all reason to play. However, some groups shift from illusionism to participationism, the second recognized referee style on our list.

Participationism, identified and named by Universalis co-creator Mike Holmes, is not structurally different from illusionism. The referee still controls everything meaningful that happens in play, while the players add nothing but color. The difference is that the players know it, and are quite happy to let the referee tell his story about their characters.

I illustrate participationism thus. There is a little girl sitting on the knee of her grandfather. "Once upon a time," the grandfather begins, "there was a princess." "And the princess's name was Alicia, right grandfather?" interrupts the little girl. "That's right, Alicia," the grandfather replies, "the princess' name was Alicia." The little girl has contributed nothing meaningful to the story, but she has been allowed to feel that she was part of it, and so listens more eagerly to the story because of her otherwise meaningless contribution. In participationist play, referees are using all the same illusionist techniques to lull the players into believing that their choices make a difference, but the players know this and are happy with it, because the referee is going to create some great stories in which they are going to feel like heroes. They can't look too hard at their victories or their successes, because ultimately they did nothing, but they had fun pretending they were doing it, just as they might have fun identifying with the hero in a movie over whom they have no real control.

These two approaches rely heavily on illusionist techniques. Such techniques are not bad or dysfunctional in themselves, and participationism is considered functional (because everyone accepts that this is what they want to do). All referee styles use them at some point. It is as possible to use such techniques to enhance player crediblity as to restrict it. Illusionists use these techniques to disempower player input on the factors that really matter, leaving them only meaningless decisions. It is as easy to use them in reverse, to disempower player input on the meaningless choices and empower them when it matters. One of the important revelations of recent theory discussions is the recognition that nearly all techniques can support nearly all types of play, if used well.

Many modules are designed for a referee approach known as trailblazing. When I first identified trailblazing as a unique approach, I dubbed it module play while trying to understand it, and then renamed it trailblazing in reference to the concept of someone marking a trail through the woods for others to follow. This solves The Impossible Thing by an unusual agreement between the players. In a trailblazing game, the referee has set up a scenario that is usually fairly linear, and has set up clues for the players to follow to reach his desired end. Thus he is in complete control of the story, because he created it and decided how it should end. However, once play starts, he is limited to revealing his scenario information and adjudicating the results of player character actions. The players are free to do whatever they wish. The caveat is that they have agreed (usually implicitly) that they are going to follow the clues and attempt to reach the end of the adventure as the referee has planned it.

To some degree, such adventures are about the challenge of completing them successfully. A group that manages to find its way to the end congratulates itself for figuring it all out; the referee also deserves congratulations, as he was able to create a trail they could follow. If the players lose the trail, they attempt to get back to it without referee intervention; if they can't, they've in some sense lost the game, and it's time to create another scenario.

As mentioned, many modules follow this approach: the players are expected to begin at a certain point, move to another place where they solve something, then through several other stages in sequence finally to reach the end. The only thing that genuinely keeps the characters on this adventure is the promise of the players that they will attempt to finish the module. Many referees design adventures this way, because it's what they find in modules.

From the outside, it may be difficult to distinguish illusionism, particpationism, and trailblazing from each other. In each case, the referee has created a story and the players are following it. The illusionist referee has locked the players into his story without telling them this. The participationist referee has their consent to locking them into his story. The trailblazer is dependent on their good faith effort to follow his clues so that his story will be told. Yet in the end, it is almost always the case that the player characters have lived the story which was prepared by the referee. The fourth approach to play, known as bass playing, is completely different from these.

Ron Edwards identified and named bass playing, with reference to the role of the bass player in a jazz or rock band. The bass player sets the beat, probably the mood, the key, and the changes in the music, but he almost never plays the melody. That's given to the other instrumentalists to provide. In the same way, the bass player referee sets up the world, the mood, perhaps the situations, but then falls into the background and allows the players to improvise, he merely supporting their efforts, bringing changes when it will work for them, and keeping it moving at an acceptable pace.

In this resolution of The Impossible Thing, the referee controls the story in the sense that he sets up the world and the situation, and so decides what the story is about and where it begins; but he does not control how it ends, or how it reaches the end, because that depends entirely on the choices made by the players expressed through their characters. In the end, the story will be as much a surprise to him as to anyone else. Bass playing is the freest and most interactive approach to play. However, it demands that the players not expect the referee to tell them what they should do. The players should do what they want to do, to make things happen that interest them. It is thus in some ways the most surprising approach and most difficult to implement. Quite a few independent game designs attempt to encourage this approach to play.

That is not to say that other approaches are not valid. Trailblazing and participationism are common and well-enjoyed by many players. What matters in game design theory is how to make a game fun. These methods have all succeeded at least for some players.

The four methods described here are to some degree the sort of Platonic pure forms that don't exist in reality. Most people blend these at some point. The distribution of credibility within the game is constantly in flux, constantly being renegotiated as play progresses. It is likely that your style falls between two of these, or draws elements from them, or even that you shift where you fall according to what you are running. (I know that I do.) Also, since these are abstractions to some degree, there is a lot of debate about whether there really are four. Some would assert that illusionism and participationism are really the same thing, or that trailblazing is just participationism that works well. It is assumed that there are other styles that have not yet been identified, but no one is certain how to find these. Theory is constantly advancing and expanding, with new ideas being added each month, and old ideas being reexamined and reformed. This series has attempted to focus on ideas that are fairly well established at least among Forge theorists, but it is always possible that something new might be recognized even as this goes to press, and something here will be at least incomplete, at worst obsolete. Still, the theory surrounding referee play style and the solutions to The Impossible Thing Before Breakfast impact not only independent game design but how players and referees can understand and modify their own approach to play within their own gaming groups.