This is the third in my ongoing series of columns devoted to the art of gamemastering. In my first column, I laid out the four roles of the gamemaster (judge, world-builder, adversary, storyteller), with judge as the most important role. In my second column, I explained the agency theory of fun, and showed how by focusing on objective rules, honest dice, and player choice, you maximize the fun for your players in the long term. In this third column, I want to turn my attention to that most contentious of subjects - story.
Now, in arguing that judging, not storytelling, is the most important facet of gamemastering, I have not been arguing against a straw man. I've been arguing against the mainstream school of gamemastering in this decade, which teaches that story is the most important function of the GM. This viewpoint reaches its fullest elucidation in the 3.5 Dungeon Masters Guide II. Let's take a look at it now, because it illustrates everything that's wrong with the mainstream view.
This Is Why They Invented Book Burning
The DMG II divides campaigns into two possible structures, "episodic" and "continuity." While admitting to the possibility of episodic campaigns, it notes that since they are just "composed of unrelated adventures" they are best reserved for groups with "spotty attendance" "oddball behavior" and "reactive players who like to have their objectives supplied to them." The real action happens in continuity campaigns, where "one adventure leads to another, creating an overall story arc that builds over time... You will adapt adventures created by others into your story arc." [Emphasis added].
The DMGII never mentions episodic campaigns again. Instead, it moves on to an entire section devoted to how to end your campaign, noting that a campaign with a fixed ending "increases the sense of excitement," adding that "any entertainer knows to leave the audience wanting more." [Emphasis added.] The authors offer several pages of advice on how to close your campaign, of which this was the most telling:
Plan carefully to see that the conclusive scenes of a closed campaign pay off. By centering the entire campaign on a set of victory conditions, you're promising a big, exciting finish. Prepare that final sequence exhaustively. It must be exquisitely balanced... Set up the climactic sequence so your resident tacticians have no chance to reconnoiter or plan in advance. Design the encounter with options that allow you to adjust its difficulty on the fly... When in doubt, fudge die rolls shamelessly, feigning surprise at the results.
The inherent assumption here, of course, is that the GM is a wise, smart entertainer, and the players are his audience and along for the ride - nuisances who will attempt to ruin your ending with their "reconnaissance" and "planning." My prior column has already explained in detail why you should disregard everything written above, but I think it's worth translating the language above for you into how I read it:
Plan carefully to see that the conclusive scenes of a closed campaign arbitrarily end the way you want them to. By centering the entire campaign on a set of victory conditions, you're putting yourself in a situation where have to railroad the ending, so you'd better prepare that final sequence exhaustively. It must be exquisitely balanced so that they almost lose, but then win, no matter what choices they make. Set up the climactic sequence so your resident tacticians have no path except your path, no matter how much they want to choose otherwise. Ignore any sense of fairness to the players, and design the encounter with options that allow you to arbitrarily make it harder or easier for the players so that they end up doing what you want, how you want. When in doubt, lie to your players about what you rolled and just make up an outcome you think is better.
Of course, the moment your players realize you are doing any of the above, the whole effort becomes a sham - they can't be excited about it because they know that they're really just an audience, not a participant. That's the lesson of agency. The players should be participants, not spectators.
The good news is that GMing doesn't have to be this way. You don't have to get trapped into a false dichotomy of "episodic" versus "continuous," nor do you need to force players to drive down your Epic Story Superhighway in order to run a fun campaign. There is a better way.