Two Approaches to Sharing the Screen

In my last post I discussed sharing the screen with three rotating DMs in a single campaign. A conversation with +cole long on G+ got me thinking about some of the problems with rotating, how we were dealing with them, and what alternatives there were. Here I discuss an obvious problem with DM rotations, and two different ways of handling this problem. They would give rise, I think, to campaigns with very different flavor.

When I refer to DM rotation, what I mean is that everyone in the group plays, and everyone has the opportunity to DM, all in the same campaign. This can be contrasted with co-DMing, where the roles of DM and player are fixed in the traditional way, with two or more people in the role of DM cooking up together the campaign world and adventures. Rotating has some advantages over co-DMing, since playing and DMing have distinctive pleasures, and it's nice to get to regularly experience both in the same campaign. In our campaign, I find that I learn a lot by playing under the other DMs in the group as they work their magic (and sometimes fail to work it). Also, since we spend less time "on" as a DM, there's less burnout, and a more equitable sharing of the preparation. (We're three years in and going strong.) But most of all there's the sheer pleasure of making a thing together. It's a higher degree of creative unity than a traditional fixed role arrangement provides. We build a campaign world together, all as players and, at least potentially, all as DMs. It's a lovely thing.

The two approaches I have in mind arise as responses to a special problem that DM rotation presents in light of two elemental joys of D&D. The first joy arise from genuine player agency. In the kind of game I like to play, a great deal of the fun comes from the fact that the world is open to the players. As a player, I want to be deciding where I go when and what I do when I get there. This is why, for me, the ideal campaign will always have the structure of a sandbox. It need not involve a literal hex crawl across an open map--maybe it centers on court intrigue and planar travel--but it needs to involve equivalent combination of structure and potentially infinite open endedness. The world is our playground with the limits set only by our shared imagination.

Another of the great pleasures of D&D is the uncovering of secrets. By design, an informational asymmetry exists between the dungeon master and the players. This is the source of great fun. When my group stood on the parapet of Submerged Spire of Sarpedon the Shaper, looking down beneath the waves at the tower below them, there was a sense of tension and anticipation on their part. They were embarking on a thrilling journey into the unknown. As DM I shared in that pleasure from the other side. It delighted me to see them deliberate about how to enter the Spire (e.g. should they take the stairs down from the bell tower, or plunge down to the base of the keep and enter from another level? etc). Similarly, knowing about the mad laboratories in the inner precincts of the dungeon was fun for me from the get go, and it would have been fun even if they had left the dungeon without having discovered them, or had left for the moment to return in a year's real time. From one side the pleasure in question involves involves cooking up secret wonders, and mapping undiscovered countries; from the other side it comes from the sense of anticipation, from exploring, probing, discovering.

The problem with rotating DMs is that these two elemental pleasures come into tension under this arrangement. For example, suppose I cook up the Submerged Spire and the players partially explore it. During some other DM's turn, the players (including me) decide they want to go there. How's that going to work? If I have to share the map and key of the Submerged Spire with the other DM, then I reveal my secrets to him. But now I'm also a player going into the Spire. I have to pretend I don't know what's coming and that's no fun. Indeed, the point generalizes; it's not just about maps and keys. Suppose another DM sets in motion a chain of events. We have learned that the white pods of the Moon Lich have crashed into the Screaming Hyena Jungle, somewhere near the lost City of the Archivists. Is this an advance force, scouting for an attack? Does the Moon Lich seek something slumbering within the lost city's precincts? Presumably the DM who introduced this event knows or has some ideas. But suppose we don't figure it out for now, and we switch DMs. And suppose then we head off to the jungle. You see the problem.

It seems to me that there two ways of handling this problem, each of which comes at a cost. The first involves sacrificing some player agency. Here the general approach is to collaborate on a campaign by dividing and conquering. Each DM is encouraged to take responsibility for some portion of the setting by introducing locations, central NPCs, and chains of events that are his own. While there is room for endless cross fertilization, borrowing, and mutual influence, these elements, once introduced, are the presumptive property of the DM who introduced them. If you're going to the Submerged Spire, then I'm your DM. If you're going to try to go up against the minotaur Malveraux who slaughtered Pierre's character, then Winston is your man. This can be accomplished on the side of the players by having a norm that they will only interact with something that belongs to one DM during that DM's turn. This is a cost because the balkanization of the campaign world introduces an artificial limitation of player freedom. (This is the approach my group uses. It works well, and is a viable solution, but I definitely feel the loss of agency that it introduces.)

The alternative approach that I've been thinking about this week resolves this tension in a different way, by sacrificing, to some extent, the pleasure of keeping secrets. Here the central concept is campaign canon. Something is canon if it's an established fact that has been encountered in play. For example, if the players have been to the second level of the Submerged Spire of Sarpedon the Shaper, then the map, layout, and contents of that level are canon. Similarly, if we have learned that the white pods of the Moon Lich have landed in the Screaming Hyena Jungle, then that too is canon. On this second approach, what passes from DM to DM is canon only. Supposed the PCs don't explored the bottom level of the Submerged Spire that I created during my turn. Suppose that they return at a later time, during another DM's turn. At that point, the DM works with the map and location as it exists in canon, but is free to put whatever he wants beyond the last closed doorway. Then, when I go as a player into my own dungeon, I have no idea what I'll find. Player agency is thus perfectly preserved. But it does come at the cost of wiping clean the geography of the undiscovered country that each DM may have created. The secrets count for nothing when the DMs rotate; canon is everything. Longterm slumbering secrets would not really be a viable option, since everything is always up for grabs. Here we see the same problem resolved in the opposite direction from the divide and conquer approach.

I haven't tried this second alternative, but I think it could be a real blast. It would be great fun to see another DM take something I made and spin it out in a direction that I never anticipated. I suspect that it would naturally produce some interesting dynamics. For example, it would certainly encourage DMs not to prepare too far in advance. I suspect it would also lead DMs to introduce evocative hooks that he would enjoy seeing other people develop. I bet there would also be a premium on elaborating canon in a surprising way, and I might expect a competitive dynamic to develop with DMs vying to take one another's material in a strange and wonderful direction. But then I don't know, since I haven't tried it.

The Protocols

Getting Started

On either approach, you will need to start with what I call a campaign frame. The campaign frame consists of three things.

(a) A paragraph long description of the setting that captures the aesthetics, genre inspirations, and main themes of the campaign world.*

(b) A campaign map. This can be a hex map for a sandbox, or a looser world map of some kind.

(c) A barebones description of the campaign base and its immediate environs.

*For example, "A sword and planet campaign set in the final days of the dying sun. Vast graveyards from the Draftsmen Wars dot the landscape, their technological horrors slumbering, as a dwindling mankind turns ever inward, languishing in highly mannered, hierarchical societies. Imagine it as sort of a fusion of Jack Vance's Dying Earth and Rifts."

The Divide and Conquer Protocols

(1) Everyone is a player, and everyone has the opportunity to DM.

(2) A schedule will be fixed in advance, giving everyone a chance to DM who wishes to do so. Alternatively, volunteers can be solicited towards the end of a DM's rotation with preference being given to those who have not DMed before, or have not done so recently.

(3) Once a DM has introduced a major setting element, including an adventuring location, NPC, or chain of events, that element is considered his.

(4) Players may do anything they like in the world, except interact in a significant way with the property of a DM who is not behind the screen without first seeking his permission.

(5) A DM's typical rotation will have an expected length suitable to your schedule and the number of DMs who wish to participate. In our group, the expectation is 6-8 sessions per rotation. When the upper end of that limit is being approaches, players and DM will voluntarily allow adventures centering on the current DMs property to wind down.

The Canon is Everything Protocols

(1) Everyone is a player, and everyone has the opportunity to DM.

(2) A schedule will be fixed in advance, giving everyone a chance to DM who wishes to do so. Alternatively, volunteers can be solicited towards the end of a DM's rotation with preference being given to those who have not DMed before, or have not done so recently.

(4) When DM rotations occurs, all that is carried over to the next DM is campaign canon. Canon is defined as fact that has been established through play. DMs are free to treat what is not yet canon in any way they like.

(5) Players may go anywhere and do anything with any DM.

(6) The length of a rotation can be fixed by arbitrary fiat, or by social consensus. I would suggest starting with the presumption of 6-8 sessions per rotation.*

*In theory, using the Canon is Everything Protocols, you could rotate DMs as frequently as you like, even every session. I suspect that this would produce an experimental quality to the game that would probably be a buzzkill once the novelty wore off. It would also be a bitch to rotate in new PCs every session. How did they get there?

Sharing the Screen

My face to face D&D group rotates between multiple DMs in the same campaign. We've been doing it for almost three years now. It all started when I moved in next door to Winston. He's the man behind the screen. Winston heard that I played D&D from my wife, and told me that he was reconvening a group of neighborhood folks who used to play. He had been homebrewing a campaign world for the group. He offered us three things: a map of the campaign world "Elmontare" (el-mon-TAR-ee), a minimal description of the setting, and a first adventure to set the tone. He explicitly conceived of this as a framework. From the start the plan was to switch off DMs. Everyone would play, and everyone would have the opportunity to DM. Through our collaboration, and shared adventures, Elmontare would become a living breathing thing. We would build it together.

When he gave the minimal description my heart sank. Owing to some sort of apocalypse past the distant edge of human memory, an advanced, decadent, and evil civilization was destroyed and the region reverted to pristine wilderness. In this wilderness, our people grew from infancy to maturity. We have no name, for we are the only humans of whom we know. We are hunter gatherers who live in relations of social harmony. There is no a currency, and our economy combines stoic independence and sane communal arrangements, with barter around the edges. It is largely matriarchal, being ruled by a queen, advised by a female priesthood (druids) devoted to our sole deity, the All Mother. In short, the setting was about as far from Vancian shenanigans and murder hoboism as you could conceivably get. There's no such thing as money, and the only (human) spell casters are druids. Oy vey.

But our first adventure convinced me that this could work. At the age of 17, those who chose to enter the ranks of the warriors (i.e. adventurers) underwent a secret ritual initiation. After a long day at the sweat lodge, the elders fed us brew laced with potent hallucinogens, and, while we were each suffering our own private spirit visions, we were taken on a wild ride through the air to some wilderness region beyond the edges of our tribal civilization. We awoke on a raft floating down a river, wearing only loin clothes. A set of instructions informed us that we could only return bearing the pelt of a great grizzly bear. So after psychedelic visions we were plopped down with absolutely nothing in the middle of a wilderness hex crawl. It took us a full SEVEN sessions to find and kill the bear. The adventure was very challenging and had the eerie flavor of a primeval wilderness filled with inhuman dangers. It was a blast. I learned a lot about how to do a wilderness adventure from it, but that's a topic for another time.

Noah is the dude with the hat

Noah, the next DM up followed suit. His adventure had us working on loan to another clan during fishing season, providing security for the fishing operation. Once again, a definite wilderness terrain was introduced that we got to know over several sessions. After the group blew it through some poor tactical choices, the fishermen were slaughtered, and we ended up waging a guerrilla war behind enemy lines. The world was developed through the introduction of a clan structure to our tribe, and some delineation of the economic basis of our existence, as well as by the introduction of a well defined wilderness area, and some new humanoid threats.

I was up next, and found myself facing quite a challenge. I doubted that I could pull off the kind of wilderness adventure that they had done so well. And the setting was just not the sort of thing that got my creative juices flowing. I never in a million years would have come up with it. My first apparently bright idea was to take the whole thing in the direction of Imaro. I asked myself why our society had no wizards. Maybe there was a taboo against magic, and a lot superstition about witches. Cults of freaky, demon worshipping witches as villains in a tribal setting seemed like a lot of fun. But from the beginning, I could see that the players didn't like it. Although I thought of myself as acting in good faith, in reality I had done something very different than the second DM. What I had done was introduce elements of paranoia and bigotry into the society, and internal sources of evil. This wasn't building on the frame as Winston had presented it; it was changing the basic premises to make the setting something (different) that I wanted it to be.

So I gave it up and switched gears entirely. What I needed to do was to make the game mine, without introducing something that would stretch or alter the minimal premises of the framework. Winston had shown me on the map where the coast line had extended before the apocalypse submerged half of the continent under water. Ancient evils slumbering under the sea--I could work with that. My solution was to plop down the Shattered Isles from my Wilderlands setting off the coast of Elmontare. They were a set of islands that were once the geographical high points of the evil civilization that was destroyed in the apocalypse. The PCs went wavecrawling and explored the demon haunted ruins of an evil Aztec type ancient civilization. The group has been to the Shattered Isles twice now, the last time diving beneath the waves to explore the Submerged Spire of Sarpedon the Shaper. It's been a big hit. The lesson is that you can always find ways to introduce something that suits your fancy, perhaps by locating it in a unexplored region (across the sea), or in the distant past, or, in my case, both.

Over time, some very nice effects have emerged from alternating DMs. The first thing is that we learn from and inspire one another. I can't tell you how much I've learned by watching Winston DM. The rotation also introduces a nice competitive culture. When I took the PCs under the sea to explore a freaky vivimancer's castle, I laid down a gauntlet. Winston responded by taking us into the heart of the apocalypse, traveling through the strangely altered wilderness, and into the endless dwarven mines that had been ground zero of the radioactive comet strike that ended the previous civilization. Since we're only DMing three months a year, we have loads of time to prepare. And it's fun to sweat it a little bit, like, "Man, what crazy shit am I going to bring to top this?"

There are also certain accidental effects of the arrangement that are worth mentioning. Perhaps the most significant is that adventures don't tend to take a completely open-ended, player driven form. This is because we each have some ownership over the themes we've introduced. When it's my turn, it's just not cool for the players to go after the minotaur priest Malveraux that Winston introduced as part of a brewing threat, even though one of our PCs died on his gore-drenched horns the previous session and we're itching for revenge. As a result, there also tends to be a bit of hard-framing to get things started. That's not to say that there isn't tons of meaningful player choice. If Winston introduces a bear hunt wilderness crawl, or if I induce the PCs to go on a wave crawl by dangling some juicy hooks, after we're off and running it's all maximal open-ended player agency. But there's a certain amount of, "It's my turn; here's what we're doing." This isn't inevitable, but I suspect that there is a natural inertia to the arrangement that will tend in this direction. It's likely that the DMs are going to stake claims to different parts of the world, including different areas and different trajectories of events, and this will necessitate a certain amount of steering at the outset of a session.

Other accidental effects include a sort of episodic feel to the game, which necessarily involves rotating party members, and adventures with different styles and themes. We also play fast and loose (Gygax would hate this!) with the campaign timeline. It's not entirely clear that the adventures are happening in the sequential order that we're playing them, although to make strict sense of player advancement, equipment, and so on, they would have to. No one is keeping track of how many years its been since such and such happened, or what season it ought to be. We often pick up old themes from a (real life) year ago as though only a month of game time had passed. Again, I think you could avoid this if you wanted to, but there is a certain natural tendency for things to go this way. I, for one, don't find it to be a bad thing.The combat round is famously abstract in early D&D. Why not do the same with the campaign timeline?

The last three years have convinced me that sharing the screen is one pretty great way to do things. My advice would be to give it a try if the opportunity presents itself. Don't shy away, even if differing tastes and interests involves figuring out how to fit your style into a setting very far from your default. If you're going to share the screen, you have to be willing to give up a lot of fussy control. But if an obsessively over-preparing, Vance loving, swords and sorcery guy like me can do Elmontare, then anything is possible.