Running D&D Games - the Role of the Ref

I was recently asked what I think the role of the referee is in a D&D game. I can't speak for anyone else, but I have a fairly clear idea of what my role is at the table, so I'm going to share it here. One important caveat here, I don’t make any claims to representativeness, I am NOT suggesting that how I run my game is the “right” way, or that it is even the way it was intended to be played. I run it this way because it is what works for me at my table. 

The Role of the Ref

To start, here is what I DON’T do:

I don’t see myself as a storyteller. 

I don’t ask for backstories and tailor the game to them.

I don’t have a fixed narrative in mind for how the game will develop. 

I don’t choose the adventure for the PCs.

I don’t force the PCs to stay on the path of the adventure they are on.

I don’t feel the need to have “story beats” to be hit every session. 

I don’t ensure that each PC will have a “chance to shine” in the game. 

I don’t fudge dice, I roll them in the open and the results stand. 

I don’t ensure that every session has social role play or combat. 

I don’t scale encounters to the PC’s abilities or level.

I don’t place magic items for specific PCs.

So you might ask, what exactly is it that I do?

A. Create a game world.

B. Decide on the game world’s reactions to the PCs actions

C. Adjudicate when the rules are ambiguous or non-existent

A. Creating a Game World

I created my own setting for the game we play. It  is a large city. It is primarily run off of tables, so I can roll for things in the moment rather than creating everything ahead of time. To create the setting I read many existing settings, both for D&D and other games, and nicked the bits I liked, then made my own Frankensetting. 

This involves a lot of improvisation. So, for example, I have tables to determine if a building is an Inn or a blacksmith’s, but I don’t have any pre-prepared information on the layout of the inn or the proprietor of the smithy. I have tables to generate NPCs if I need them, but most often I create them ex nihilo on the spot and record the details for later reference. 

When you run a game this way the player’s choices for their PCs essentially create the game world as you play. The Inns, smithies and such that are detailed in the setting are the ones that the PCs have visited. Call it Schrodinger’s Setting, nothing exists in detail until the PCs interact with it. 

I have created flavor details for the setting, holidays and customs, random encounter tables, all of that is pre-created. But the VAST majority of it will not be directly interacted with. 

In addition, I have detailed factions. Factions are groups with goals and interests, usually independent of the PCs. When they arrive in the city they usually become involved with a faction intentionally or by chance.

All of it is schematic. So for example, the city in my setting is run by Noble Houses, the Houses represent a faction in the setting, I have detailed the name, Ward, crest, and function for each House, as well as its leaders (a warrior and a warlock), but I haven’t detailed the hundreds of other people who would be part of the House. It is like this with all factions in the setting, there is some high-level information to give inspiration and flavor, the rest I improvise when needed.

This style of refereeing is not for everyone, you have to be comfortable improvising and avoid too much repetition. I run my games this way as I’m good at improvising, play to your strengths and all that, and I have tables that allow me to generate content when needed. The tables were built to be inspirational. The key advantage to this approach is that it gives me maximum scope with minimum input. Rather than detailing every aspect of the game world, I use tables and inspiration and I invent it as I go. 

It also gives my players complete freedom to go wherever they want. I run a 100% sandbox game, the world is completely open to them, they just have to decide what to do. I know some people dislike this approach, as they feel their players would be paralyzed with options. But it works remarkably well. 

B. The Game World Reacts

Sandboxing is avoided by some as they fear their players won’t know what to do. In my experience this is not what happens. It is just that players are so used to the traditional approach (you meet in a tavern and a mysterious stranger asks you to go on a quest) that they don’t know what to do when the ref doesn’t leave obvious hooks.

But you really don’t need that, as the PCs interact with the world consequences of their actions build up, they make enemies who will oppose them and allies who will ask for their assistance (and offer assistance when needed). You don’t need to lay down hooks when the PCs pass through the game world as eventually it will react to what they are doing. 

Sometimes it happens because the PCs make a decision, e.g. they decide to go purchase mounts and in meeting the animal handler an adventure seed is dropped. Other than that, there are three primary mechanisms that prime this process. One is random encounter rolls, another is encounter reaction rolls, a third is factions. 

There is a misconception that random encounter rolls are all about fighting. Sometimes they are (e.g. if you run into a tiger, the tiger will very likely want to eat you), but they don’t have to be. Random encounter tables are populated by animals, monsters and NPCs, some of which will be potentially violent, others will be less so. As the PCs pass through the city I roll for random encounters regularly, and eventually one will come up.

However, I also use encounter reaction rolls. In AD&D when you meet an NPC or monster that can communicate, and you speak to it, eventually the ref will make an encounter reaction roll, this roll determines how the NPC/monster will react to the PCs parlay attempt.

This mechanism is the beating heart of sandbox improvisational play. It is one of the cornerstones of old school play as well. The way it works from the ref’s side is fairly simple. I have a set of expectations about how a NPC/Monster will react based on their alignment and the background of the setting. When parlay happens I will free form role play the encounter as the NPC/monster in question.

This stage is exploratory, essentially the NPC/monster is trying to determine if the PCs are going to oppose their interests or not. So for example, the PCs think that a jewelry shop is a front for a cult. So they decide to come back to it at night and check it out. On the way there, they run into a city guard patrol. Now, the guard patrol is going to be suspicious of anyone lurking around in the middle of the night, their goal is to find trouble and stop it from happening. 

In some games this would be the start of a fight, as the party would feel they could take the guards and that they might stop them from their goals. The guards call out the PCs, and they decide to respond rather than immediately fight.  

So they ask the guards a question about where to find a good inn to sleep at as they are tired, lost and don’t know the city. As they look like outsiders and they aren’t being overly hostile, I have the sergeant answer and tell them about the Big Shooter Inn a few blocks away. No need for a roll yet.

Then, one of the PCs announces that she will walk ahead to the inn while the rest of the PCs ask the guards more questions. She is planning on getting into position to backstab if things go badly. However, the fact that one of the PCs left could very well make the guards suspicious so I make an encounter reaction roll. 

It comes up positive, so I have to interpret it right on the spot. Why would these guards be OK with strangers asking questions at night while one of them slips off? I decide they is bored so they will be OK with continuing the conversation, and they aren’t that invested in their jobs anyway as they feel the late night patrol is a lowly task.

Notice that in explaining the encounter reaction roll, I have instantly created some backstory for the guards, and potential for future developments. Now as the conversation continues with the party I have some motivations in mind, and I can make the conversation more detailed and realistic. 

So what could have been an instant fight is now a chance to develop the game world and create potential allies or enemies for the PCs. This combination of random encounter rolls and encounter reaction rolls drives a significant amount of improvisational play at the table. 

It is an important tool for improvisation as it gives you instant inspiration, you have to interpret  the results of the die roll, and that drives you to create aspects of the game world on the spot. You could of course just improvise these without the rolls, but I find the kernel of a die roll is far more inspirational than just pulling something out of thin air.

And that brings us to factions. Other than PC motivated tasks (go to blacksmith) or random encounters, when the PCs arrive in the city their actions usually put them into contact with one or more factions, and most often they align themselves with a faction to gain power or just to survive. So for example, when a thief arrives in the city they often go out and find the Thieves Guild in order to join. Warlocks need patrons to advance so they will often join a Noble House to get one. Priests will look for a temple of their god or pantheon, etc. 

Once they interact with a faction they will often get asked to complete tasks (as an initiation) or become embroiled in shenanigans as a result. And it works very organically. So take the case of a group of thieves who started one campaign in the city. They approached the guild and took on a job as their “entry fee”. Anything they stole would belong to the guild, and it had to be high profile. They cased a place, stole a magic item, and took it back to the guild. That got them in. However, it also earned them the enmity of the NPC who they stole from. In a world of magic, it isn’t terribly difficult to find out who stole your stuff. 

So right out of the gate the party had an ally (the thief who initiated them, their success reflected well on her, and she was their contact with the guild, so she saw them as allies) and a few enemies (another thief who was aligned against the thief who initiated them, as well as the rich merchant they stole from). 

I find that within a few sessions the PCs have generated enough “reaction” from the game world in terms of allies and enemies to sustain long term play. In addition, being aligned with a faction usually carries commitments that they have to fulfill. Faction play is a massive generator of adventures. 

There are two important observations about this system. 

One, it is entirely player driven. Player choices about PC actions drive all of it, right down the line. I don’t direct the players or tell them how to react to anything, they have complete autonomy over their PCs actions, and complete authority over deciding what “hooks” to bite. When I say that I run a “collaborative” game, this is what I mean. The direction of the PCs adventures is entirely shaped by their choices. 

Two, this system works best when I decide (using the dice where appropriate) what the game world’s reaction is to the PCs actions. I know there are games where the PCs decide on the narrative and how it will play out, but that’s not the system I’m running. The appeal of doing it this way is that the game world feels independent of the PCs, they don’t decide how it will react to them, I do. This mirrors an aspect of the real world, in that you make your choices and the world reacts to those choices, often in ways you didn’t predict and cannot control. In short, running the game world like this creates immersion, it feels real. 

Additionally, as I rely on dice rolls rather than my own whims, the world seems impartial to a significant degree. My players know that the dice are a big part of how the game world reacts to them, so they don’t feel I’m making it easier for them, or that they can make it easier, and this gives them a sense of accomplishment. It could have went south, they know that every time they have an encounter, so when they manage to make it work it feels earned.  

The combination of setting, factions, random encounters, and encounter reactions can sustain an almost infinite amount of game play. None of it is predetermined, there is no fixed “story” to tell, and it can develop in ways that surprise everyone at the table, players and referee. That is one of the greatest strengths of this style of play.

C. Adjudicating the Rules

I’ve been running games off and on for about 35 years. In the last year I logged about 300 hours of table time running a house ruled 1e AD&D game. In almost EVERY session I can remember, at some point, I had to make a ruling on something that wasn’t in the rules. 

Take an example from a recent game. One of the party members had been possessed by a ghost. The party magic-user had cast an anti-magic shell on himself. The possessed party member attacked the magic-user. One of my players asked, “would the anti-magic shell push out the ghost?”

Good question. The spell description says the shell is impervious to “all magic and magic spell effects”, possession is clearly magic. However it also says that it “prevents the entrance of charmed, summoned, and conjured creatures.” Is a possessed PC a “charmed” creature? Would it just be shut out of the shell like a protection from evil hedges certain creatures?

Whether you think the answer is obvious or not, it is clearly the kind of thing that might require a ruling at the table, as it is not obvious from the text.

“If I drop on the guard from above, how much more damage will it do”?

“Does speak with animals work on alien animals in a space ship?”

“Can illusionists read magic, as they can read and use magic-user spells ‘which contain spells usable by illusionists’, but they don’t have the spell read magic.”

“If I use a one handed weapon with two hands, can I do any extra damage?”

“If my magic user holds up a shield, does it lower his AC, as magic-users cannot use armor and shield, does he take a penalty?”

“Can the vines around the waterfall be used to tie together logs to make a raft?”

“How far can I swim in leather armor”

“Does the monk’s ability to knock aside missiles allow her to knock aside the poison stinger of a giant hornet, or alternately, does the monk’s ability to take no damage on a save and half on a failed save halve the damage from a giant hornet’s sting, does it reduce the time of incapacity if the poison save is failed?

“Is a Brownie familiar, an “enchanted or summoned creature” hedged by a protection from evil?”

You get the idea.

Any TTRPG that doesn’t have a “universal mechanic” for everything will run into this problem. TTRPGs tend to be open ended in this sense, they have constraints, but compared to board games the possibilities are almost endless.

But that’s not the only problem. 

There are of course “stripped down” games that are “rules light”. They will still require adjudication, but less of it. D&D, however, is “rule heavy”, in the sense that there are rules for a LOT of things. Spells are a great example. Spells are essentially mini-rule sets for a specific application, and there are hundreds of them. And any time you have that many spells, the number of ways they can interact is almost limitless. Add rules for magic items, and right there you have a world of rulings ahead of you.

But it’s worse than that!

D&D isn’t really a fantasy game, it’s a pulp fantasy game, so it draws from EVERYTHING. D&D has ray guns, six shooters, alternate planes of existence, time travel, teleportation, psionics, interaction with deities and flumphs. 

I used to play Top Secret, a spy game. It also had many occasions when rulings were needed, but nowhere near as many as in my D&D games, because Top Secret didn’t have magic or superpowers or gods walking the earth. It had guns, cars and bombs. And for the most part the PCs were stuck on Earth and could only move from place to place by conventional means.

Because D&D has so many fantastic elements, D&D models an entire universe, not just a world. The scope of that is incredible. You may only do it bit by bit, but as you campaign you will build a universe around you. That means the game will always throw up situations that stretch the rules.

Now, one can argue, perhaps fairly, that a certain percentage of these cases will be cases where:

The ref has forgotten an existing rule that covers this situation

The ref is misinterpreting a rule and doesn’t see how it applies

That seems fair. But even if you account for that, say by dropping ½ of the cases in last year’s games that I mentioned above, that still means it occurred in approximately 75 sessions. That’s still pretty significant.

But I would maintain it happens in almost every game I run. 

Why does all of this matter? 

People criticize old school games for having a ref that has “ultimate authority”, this is seen as an open invitation for god-complex DM’s abusing players in the game. Gygax didn’t do anything to help with this, he very much advocated for an adversarial style of play where the DM was “out to get” the players, and where DMs came to expect players to try and “game the system”. 

And I’ve seen that in action. I’ve played in games where the DM was just being a jerk, making arbitrary decisions to the detriment of the players and being inconsistent in the process. 

However, making rulings on these sorts of things is not something that is best left to “group consensus” for a number of reasons:

1. As the ref I’m the most familiar with all of the rules for the game. I can be mistaken, or forget something, but at the end of the day very few players have taken the time to get to know the rule set like I have. So I’m best equipped to make that ruling.

2. Any game that has lasted for a long time will have a set of house rules, a few or a lot, and I’m also most familiar with those.

3. Many of my players are utterly uninterested in making these sorts of decisions, they came to PLAY, not to RUN THE GAME. For these players, it is preferable for me to adjudicate and move on.

4. I’m fairly fast at making these sorts of decisions as I have had to make many of them, so we waste a lot less table time.

5. It adds to the sense of immersion and accomplishment when I make these decisions instead of the players. When they make adjudication decisions the game can seem less challenging, as they can adjudicate in their favor. If the game world is too mutable to their desires then it feels like they are writing a story, not going on an adventure.

6. When the adjudication doesn’t impact all PCs the same way, it adds to the impartiality if I make the decision. So for example if it is an adjudication about a class ability then it will only apply to some players, and leaving it to the players can compromise their decision.

7. It is not uncommon for refs to enjoy these sorts of decision making processes, making an in-game ruling is a lot like game design in real time, and that’s something many refs get a buzz from. But many players just find it overwhelming. Particularly when they realize that their decision will impact future play. I have many players that want no part of that.

8. Sometimes the players just can’t make up their minds, that actually happens a lot. Players have varying degrees of investment and knowledge of the game, and there are often interpersonal issues at play behind these decisions. Being the final arbiter allows the ref to bring the process to a close when it is taking up too much table time.

9. There are often situations where the PCs can’t be aware of all the factors in the decision making process, as the NPC/monster or environmental factors in play are unknown to them. To keep that mystery in the game it is better for me to make the call.

10. Game worlds have flavor, atmosphere, a sense of presence. Sometimes adjudication of rules can impact this aspect of the game. A ref has the ability to shape the game worlds flavor and atmosphere, and this also creates immersion and a sense of a living world that the PCs explore. It is much easier to maintain this when the ref can make rules adjudications. 

Having said all of this, the ref has to maintain a degree of impartiality, otherwise the game can seem unfair or “rigged” against the players. So there are steps I take when adjudicating a situation that isn’t covered by the rules. 

First, I tell the players what my adjudication is, and ask if there are any questions or concerns.

If there are any questions or concerns, we talk it out. They mention precedents they think would lead to a different ruling, and I make my case for my interpretation, or change my position based on their input. I’d say that the division on this in my games is about 80/20, e.g. about 80% of the time the players agree with me and we move forward with my call, and about 20% of the time I change my mind.

Then I remind them that any decision we make at the table is now the rule for our game, and that rule applies to NPCs/monsters just like the PCs. So if we rule that dropping on a foe from a height and striking them can double damage, then their foes can drop on them from a height and do double damage as well.

I then codify the rule in our House Rules document, and it becomes a part of our game.

I’m sure that this process could be made more democratic, or handed over completely to the players, but for the reasons I mentioned above I prefer to be the one who has the final say. 

Conclusions

So that’s the gig. I create a game world, I decide how it reacts to PC actions, and I adjudicate “corner cases” where the rules are non-existent or ambiguous. I think this is plenty for the ref to do, and is both engaging to me and rewarding for my players. It has helped me to sustain 4 concurrent campaigns over the last two years (now 5 as of this September) without mountains of prep, and allowed each of these campaigns to run as sandboxes where the adventure is directed as desired by the players. 

When I read about D&D campaigns where the ref tries to “hit story beats”, incorporate PC back stories, direct PCs to the “big events”, balance encounters for the PCs and “ensure that each player has a chance to shine” I just feel exhausted. I have managed to run engaging, rewarding and immersive games for my players without adding any of these elements to the game. It’s not for everyone, but if you are interested in trying D&D in a different way, it might be for you.