In my limited experience, the GM provides a story to the players, and their "main" characters mostly react to the environment. Having a look at a couple of random highly-voted (and IMO high-quality) answers on this site, confirms the same impression:

As a GM, it is your job to create an enjoyable game.

one of the great strengths of role playing games is the fact that the players can actually influence the story

I was going to ask a question on Writers.SE about plot writing for one-offs, but something was not right. The question was coming out clumsy.

A similar question to mine was answered that the main character should experience:

Conflict, Rising action, Climax, Denouement. No subplots, not many secondary characters.

So I attempted to explain that in RPG, we have several equally-important main characters that drive the plot.

Wait, NO. The GM drives the plot! So what is "main" about the main characters?

I have heard about narrative-heavy games, but never been exposed to one, nor do I see many question on this site about such.

Why are RPG plots/scripts different than those of a book or a movie - in the aspect that not the main character(s) create the story, but the environment around them?

Thank you everyone for the wonderful answers. Here are my personal conclusions:

  • Speak with future players/DMs if they too like sandbox play with elements of drama (failing that, sit at home and play Skyrim).
  • Try different systems. Try FATE.
  • Play with people with imagination, who enjoy using it.

16 Answers 16


The key concept to understand the difference between RPG's fiction and written fiction is that of authority.

In a written story, trivially the author has authority over main characters and the environment, so she can optimize the sequence of events (the plot) to heighten the emotional impact for the reader (if she knows what she is doing).

But let's not forget that RPGs were invented to mix the experience of war games with the emotional appeal of written fiction.

In war games, the environment is fixed, and each player has authority over one and only one character (whether an army or a single person doesn't matter). To mix this with the kind of flexibility the story in a book can have, the most natural step is to simply give authority over environment to another player.
Thus the concept of Game Master is born: it's the single person who has the responsibility of setting-up a situation, gather the reactions of other players, and let the game mechanics decide the result.

There is a problem though: since dice have no sense of aesthetics, it can often be the case that a satisfactory resolution of a conflict is spoiled because excessive bad or good luck. The Game Master is then given authority over rules, so she can bend or ignore them for the sake of the story. This is the modern understanding of the GM: the player with authority over rules and environment, with the implicit responsibility of the plot.

However, once you explicitly recognize this, you can start playing with the structure.

You can, for example, remove the authority over rules, keep the GM's authority over environment, and come up with game mechanics that automatically steers the story in an interesting and balanced direction. Games like Dogs in the vineyard or Apocalypse world retain a traditional role for the GM (minus rule-bending) while employing narrative conflict resolution (the so-called narrative games).

Or you can have chance-based conflict resolution but share the authority over environment and characters between active and non-active players, usually in a turn-based structure (the so-called master-less games). If I'm not mistaken, Polaris, Shock and Dirty secrets follow this structure.

Other games employ both: Fiasco, for example, has a turn-based shared authority with a bare-bone narrative conflict resolution; it is thus in the category of master-less narrative games.

All these approaches produce different kind of stories, aimed at different kind of emotional impact: written fiction is (supposedly) maximized for the passive consumer, who has no responsibility. RPGs, on the other side, optimize for immersion and responsibility, and regulate the unfolding of the plot by mechanic means, sharing authorities in different ways.

  • 6
    \$\begingroup\$ "But let's not forget that RPGs were invented to mix the experience of war games with the emotional appeal of written fiction." This alone would have been enough. Nice to have you aboard this site! \$\endgroup\$
    – Vorac
    Commented Oct 23, 2013 at 15:53
  • 8
    \$\begingroup\$ Another mix is removing the GM's responsibility for plot aesthetics entirely, moving them back toward a referee role and letting story be an emergent aspect of playing. This results in explorative sandbox play (though it doesn't make GM storylines impossible, just no longer protected). \$\endgroup\$ Commented Oct 24, 2013 at 0:53

The reason that constructing plots (for a GM) is different from constructing plots (for an author) is simple but profound:

The GM is not in control of his protagonists.

A game is not a book, even more than a book is not a movie. The players are in control of their characters.

The Conflict / Rising Action / Resolution structure is still at the core of dramatic storytelling. But in a novel, it's perfectly possible for that conflict to be internal, with the character's personality designed around the plot, as much as the other way round. And the author can control the character's decision-making - their poor decisions can come at dramatically perfect times, in plot-convenient ways. (So can their good ones, but plot rarely turns on those before the climax!)

A GM, however, even in a "classical" non-narrative roleplay, has little say in the personalities of his protagonists - or even in elements of the story structure around them. (An author can sideline a central character with an off-page errand for a chapter. A player will quite reasonably resent have nothing to do all session.)

Instead, the GM has to create the plot around the established backgrounds and motivations of the characters. This requires more attention to worldbuilding, and to generic plot motivators which can be adapted to drive the adventure forward whatever the players do, without overriding their efforts. (If the players can't affect the outcome, you are not playing a game, just conducting an usually slow story reading.) Likewise, the GM does not decide how many primary characters there are - like a guest writer for a TV series, the GM must work with exactly the established cast.

It is still perfectly possible for the main characters to drive the story. But the GM cannot force the characters to behave in story-driving ways, only invite them. This means that the GM cannot rely on character motivation, until he knows the characters and their players so well that he can predict their actions in a given situation, and design situations that bring the characters into the spotlight.

Thus, the story is constructed with the flexibility to continue working even if one of the characters surprises the GM by doing the opposite of what is expected.

In a modern narrative-centred game (FATE, say) this is even more true; the GM doesn't even have fixed control of the universe outside the protagonists. In a game like this the GM's role is to create a consistent universe which drives the story, but not to tell a story - instead, the GM mediates the group who are telling the story.

The point of this is to make the characters drive the plot more, not less - giving the players the ability to adjust the universe lets them adjust the plot in the way an author would, to bring their characters into the forefront.

  • 4
    \$\begingroup\$ +1000; this should be on the first page of every "how to be a GM" article. Not understanding this is probably the leading cause of poor GM-ing, particularly railroading. \$\endgroup\$
    – user1861
    Commented Oct 23, 2013 at 15:43


RSConley's answer comes close, but misses an important detail that, in my opinion, answers historically the central question: "why?"

When Dave Arneson was invited to demonstrate Blackmoor to the Lake Geneva group, only Dave could make the trip from the Twin Cities. In the Twin Cities Dave's role was arbitrator, not antagonist or master; the players had each other for that, by collaboratively controlling the game and playing the world's inhabitants themselves and providing in-world antagonism to their fellow players. Dave merely refereed those player-driven conflicts. But the Genevans didn't get to see that game "mode." In the mode they saw, one person played "against" the rest.

Blackmoor, of course, was Dave Arneson's famous early Chainmail-based campaign, a precursor to Dungeons & Dragons. In Blackmoor some players in the Twin Cities group played traditionally-heroic PCs fighting for good and Law: Dave Fant controlled Blackmoor Castle and Duane Jenkins played a baron ruling an area near the town of Blackmoor, for example. But others played more nuanced or antagonistic personalities: Dan Nicholson "the shady Merchant of Blackmoor," John Soukop the evil wizard in Blackmoor's dungeon complex. And some played monstrous characters: Fred Funk commanded a horde of orcs as King Fred I of the Orcs. (Peterson, Playing at the World, pp.68-72)

In these games Dave played the "referee" role as was commonly understood in wargaming: arbitrating conflicts between two players.

Lake Geneva group sent an invitation in the fall of 1972 for a Blackmoor demonstration--but only Dave could make the trip. This set the stage for a concept of the game which prevailed in Lake Geneva in the early days:

In Blackmoor, as it was played in the Twin Cities, most of the Baddies were nominally under the control of players.... Running the game for the Lake Genevans, however, Arneson needed to embody all the forces of Chaos himself. (ibid., emphases mine)

So the Lake Geneva group, whence derived the traditions of D&D, whence derived many of the traditions of TRPGs, had an experience with the following characteristics:

  • they played relatively-minor heroic characters--no doubt in the course of a campaign these characters could grow in stature/reputation, but they started their careers as "nobodys";
  • a referee (nowadays "GM") presented them with a setting whose details he knew immeasurably-better than they--in the Twin Cities game Dave had originally designed the castle, town, and dungeons, but the players and he constructively added enough details that the setting was "theirs," not "his";
  • the referee played the roles of all non-player characters, particularly the antagonists!

Sound familiar?

If two or three more of the Twin Citiers had been able to make the trip, might we have seen their model prevail through the 70s and 80s? Might they have all played Chaotic forces "against" the Genevans' characters, and would D&D have been born as a team game? We'll never know.

The "why" of how it started this way is an historical accident, I believe.

  • 1
    \$\begingroup\$ Having just read Rob Kuntz' short book that is about Dave Arneson's genius, there is some truth in this answer but it's hard to tell about the what ifs. \$\endgroup\$ Commented Apr 26, 2017 at 13:34

Who Drives the Game Forward? Whoever Steps Up

The fiction of tabletop RPGs is, at its foundation, collaborative. Rules and social structure shape who gets to say what about the imagined space, but the core appeal is that it's shared. Roleplaying games thrive on fruitful creative interaction between all the participants, mediated by useful structure.

Since it's a shared space, anyone can drive the game forward. All that's required is the ability to step up and make bold fictional statements — whether about the world, the situation, or just your own character — that engage the other players directly.

This really isn't the sole occupation of one person. Games flow best when all of the participants1 are involved in driving the story.

The Traditional GM Has To Step Up

A pattern in most RPG of the past and present is that the GM fills many roles. Oftentimes, the GM is:

  • The main authority2 on the setting and environment. It's right there in the term "Dungeon Master"— the GM is the person who defines the environment we are exploring in play.
  • The main author of the situation. This is an outgrowth of setting authority in traditional RPG structure. What's going on right now is treated as a feature of the setting, which is offered up to the PCs through vehicles like adventure hooks.
  • The player who controls the antagonists.
  • The main authority on the rules. Older D&D often refers to the Dungeon Master as a "referee;" anyone can read the rules and attempt to apply them, but the GM is the main authority on how to interpret the text in play, and makes on-the-spot "rulings," which involve both adjudicating interactions of existing rules and inventing new mechanics for peculiar situations. (Sometimes the GM is even expected to override the rules for the sake of realism or satisfying dramatic progression.)

So, traditionally, the GM wields a lot of power and does a lot of work. If they aren't doing this work, it's just not getting done at all — a team-oriented game like D&D is not really designed for the non-GM players to become the antagonists, for example. The game will just fizzle and flounder without the GM's active input.

This concentration of responsibility and authority means that typically the mantle of GM falls on the most invested player in the group, the one who's most willing to put in the groundwork to make the game sessions happen in the first place. Also the one whose head is full of ideas to share with the group — ideas that end up driving the game.

What Does Everyone Else Do?

Well, they're playing, too! They can be involved in the game in several ways:

  • Creative Back-and-Forth

    If they step up as well, then you get a healthy back-and-forth where the GM stuff and the protagonists feed into each other. The protagonists and antagonists interact. The situation adapts to their motivations and their motivations adapt to the situation. Their actions change the world. Et cetera. Everyone is driving the game; the GM just has more responsibility to keep up their part (see above). Whether it's a purposeful series of dramatic events or more of an afterthought to a sandbox hexcrawl, the overall "plot" is very much the result of shared effort coming together in play.

  • GM-Lead Storytime

    If the non-GM players don't quite step up, then the focus shifts to the situation on its lonesome. The PCs still have to navigate it, but the relationship is much more one-way: the PCs' main role is to react and witness. The GM's scenario becomes the totality of "the plot."

    Note that this behavior isn't necessarily a problem. It can be a letdown for the GM if they were hoping for more stuff to bounce off of, but there's nothing fundamentally busted about a player treating RPGs as cool storytime with audience participation rather than shared improvised fiction per se.

  • Disaster!

    If the non-GM players want to step up but the GM won't let them? That's illusionism and rampant social dysfunction (but I repeat myself). This state of affairs quickly leads to frustration, boredom, and hostility at the table.

In my personal experience, the most common pattern is probably some mixture of players stepping up and sitting back depending on their particular mood, which results in the GM driving the majority of the plot without becoming the true sole author. Most of the game is determined by the GM's choices in defining the scenario, but the PCs still make critical choices or go off on their own side plots from time to time.

(So, what happens if you're GMing and you "write the plot" ahead of time? Well, now you've made the assumption that players won't push their own contributions very strongly, so you're locked into "GM-Lead Storytime" or illusionist "Disaster" mode, depending on the group. Make sure you know which one you'll be getting3.)

The GM's Role, More Generally

Okay, but it's 2013 or later! Roleplaying is, like, 40 years old and "traditional" RPGs don't represent the entirety of the hobby anymore. In the modern day, not every game calls upon the GM to be the sole author of the setting and situation, nor do all of them assume the GM will play the role of referee, teacher, or host. So, what are GMs doing in these new-fangled games?

For example, if we take strong character-driven play on the part of everyone else for granted, does the GM still need to "drive" anything?


Yes, because we're all supposed to be driving the game in our different ways.

Yes, because if the answer was no, this game wouldn't need a GM player at all.

In particular, the GM is typically the player with the best view of the big picture of play. A common paradigm is for the GM to be responsible for the game's overall objective4: the other players can focus on advocating for their PCs while the GM makes sure that the overall theme and flow of the game — Is the setting contributing to the desired mood at the table? Are the character's beliefs being suitably challenged?

The GM is a "bass player," setting the tone and rhythm for the other participants, the foundation on which we all build our fiction.

The GM is an instigator, introducing elements to spotlight and challenge the protagonists in ways that serve the larger themes of the shared story.

These are very much active, engaged, story-driving roles.

1 - Or at least most of the participants — I think there's actually a pretty viable niche for playing your PCs as "supporting characters" in RPGs, though it's not discussed often.

2 - "Authority" itself is kind of a loaded concept. It shapes how we think about "who says what" in subtle ways. I'll use it here because it's an accepted term and that's a different conversation I'm not entirely prepared to have.

3 - This is why trying to learn how to GM from published adventures, especially the "story-driven" ones, is perilous: they're pre-written for no one in particular. At best, you get a bit of canned prep you can reuse and repurpose. At worst, you are now exploring a story completely devoid of protagonists within the framework of a game completely devoid of choices.

4 - Related context: The Object of the Game vs Your Character's Goals, Recipe vs Game.


First of all, It is impossible to bucket all of the role playing games every conceived or played into a single group and then say that in this group, the story is driven by environment (GM) rather than characters.

It is very possible that some role playing games are in this category, but I am sure many are not.

Regardless, of whether it is a role playing game or not, all stories require a plot, a story line, or a narrative of some kind to progress the reader or player through. It is my experience that this is a combination between the main characters and the world around them.

To more specifically answer your question "Why is the GM usually the driving force in RPG?", We need to understand the role of the GM.

The GM or storyteller is everything in the story, from the plot, to the narrative, to the birds in the sky and the leaves on the trees, the storyteller is in charge of it all. If this was a book he would be the author, and if it were a movie he would be the director. But because it is a game, there are also player characters, they are characters that also exist in the same world, and follow the same rules, but their actions and thoughts and interactions with this world come from players and not from the storyteller.

The story created from a role playing game is then a bit more collaborative than other types of stories because there are multiple authors, in some role playing games the players have more control over the other parts of the story as well, and in others almost none. But the concept of what a role playing game is is defined by multiple people each taking a different role in a story and playing together.

So, I think that you are mistaken, it makes no difference if it is a role playing game or a book or a movie, the story has characters, it has a world, it has interactions, the only difference is who controls those interaction, or to put it in a writing analogy - the person who writes the chapter.


Disclaimer: I play systemless narrative games.

Most, if not all, the games I run follow the same structure: There is a cast of NPCs and factions that have certain goals. Clearly, this generates a conflict of some kind. As a GM, I know what the various NPCs can do, how they react, and how they behave. If there were no players, I could tell who would achieve their goal, how they would deal with the opposition, and how events would play out.

The PCs interact with some aspect of said conflict(s). From there on, there is no plan. Whatever the players do influences, changes, and/or morphs the NPCs/Factions. So, after each player action (technically, player character action), I (as the GM) have to work out how the world will react to their action.

In addition, I always heavily link the player characters to some factions/NPCs. This allow the player to know that their background was used to write the game. It tightly couples PCs to the (sub)plots. Thus, if a payer wants a certain story arc for their character, they will get this.

The difference between this approach and a book/script is simple: the players get to decide where the plot goes and how the conflict(s) get resolved. It is a collaborative effect where what happens is cast in stone and is canon. You cannot retroactively change it.

Could we all write a book/script instead of playing the game? The no retroactive change allowed make this difficult. Without it, it could lead to something like Good Omens... If the participants were Gaiman and Pratchett.


As I've said before in at least a couple of places, you can't really plot an RPG. Because the plot is the sequence of events that make up the story, and the players are going to make many of the most important decisions - namely, they will determine the actions of the main characters. Therefore, the plot is jointly in the hands of the PCs and the GM. What a GM can do is provide a situation.

And that's why in most games (excepting GM-less games such as Fiasco where the situation is determined collaboratively and games where the situation is dictated by the game itself like Lady Blackbird) the GM is a driving force.

Because the GM traditionally is responsible for two things:

  • The initial situation - the set of conditions that drives the PCs to action.
  • Feeding the results of those actions back into the situation to create another situation that demands action

These responsibilities apply across the vast majority of games, whether the game rules couch them in those terms or not.

In traditional dungeon-crawling games, the initial situation is largely set up in advance - there is a subterranean complex (possibly inhabited by a fire-breathing reptilian) and the confines of that complex are stocked with obstacles (monsters, traps, puzzles) and rewards (treasure, XP).

The GM in that instance is also responsible for providing a "hook" - a situation that would prompt a party of adventurers to test themselves against the dungeon. This is why old men die so often in taverns clutching scraps of parchment or whispering cryptic last words. The GM is then responsible for applying the results of the PC actions to the dungeon environment and modifying the situation to match what has happened in the story so far. For example - there was a room with a sleeping troll, and the GM expected his players to try to sneak past this overwhelming opponent because they had learned that there was an important key in the niche beyond him. But the players fled from a goblin ambush and made their stand outside the sleeping troll's door. Now the GM has to decide (or allow the fortune mechanic of the game to decide, but either way, the GM has made the decision) whether the troll is still asleep or not. And if not, the GM has to decide what the troll's next actions might be - does he go to his door to see who made such a racket? What will the PCs do when the troll emerges?

In some games, this cycle is implied, in some it is explicit, in some it is enshrined in the mechanics.

Take the games using the Apocalypse World engine, for example. This family of highly narrative games builds the feedback directly into the mechanics. When a player rolls dice, the action will either succeed, suceed at a cost, or cause trouble (which doesn't keep the character from succeeding, it just means trouble) depending on the result of the roll.

But the GM can't be the driving force, because the players are in control of the actions of the main characters. So unless the GM overrides the spirit of (most) games and dictates player actions, the GM cannot possibly plot the game alone. But by creating the initial situation and evolving that situation in response to player actions, the GM has an extraordinary amount of power to shape and influence the plot - to be a driving force in the game.


Note that many books/movies are about the characters reacting to their environments.

If you are comparing traditional RPGs to certain genres including anything that might be called "Action" or following along with most of Horror, Sci-Fi, and Fantasy, I think the difference may be smaller than you are suggesting. In most of those situations, it really is about the protagonists reacting to the environment and the environment driving the story (so long as you include antogonists/NPCs as part of the environment.)

It is worth looking at an examples. In the first book of Harry Potter, Harry is not really driving the action. He is instead constantly reacting to the environment. He reacts to the sumons of the school, to the environment of the school, and eventually to Quirrel. He is constantly reacting much as the PCs do in a traditional RPG. The main difference is that the author controls both the PC and NPCs/environment rather than just one or the other.

Now that anology only fits for certain types of books. More literary books that focus more on internal conflict are a bit different. But that is where the narrative style RPGs come in and some RPGs which lean towards the "traditional" format still are able to have heavy internal conflict. Vampire: The Masquerade has a definite storyteller and isn't truly narrative, but it was designed to focus heavily on personal horror and internal conflict.

Traditional RPGs also tend to have more player driven variations

You also seem to focus on the character driving the story and trying to achieve something rather than reacting so much. But RPGs have an anolog for that too. They are called sandbox games. The character's decide what their goals are and the GM just provides a world where they try to achieve those goals.


RPGs originated in the early 70s in the miniature wargaming community of the Upper Midwest of the United States. It was an extension of the type of games that were already being run. A campaign of interlinked sessions was already known mostly through the playing of Diplomacy variants that combined the game with miniature wargaming. Miniature wargaming already made use of referees to handle double blind scenarios and rules adjudication. And the playing of individual characters was pioneered by David Weseley's Braustein games.

Dave Arneson was the first to combine these elements in his Blackmoor campaign. His campaign contained many wargame elements like a epic war between the forces of Law and Chaos. However one major difference was that Dave allowed his players to attempt anything their characters can do. This was the key innovation that turned a wargame variant into tabletop roleplaying and into a hugely addictive pastime.

A referee was mandated because the players, as their characters, had incomplete knowledge of the world in which they inhabited. Dave needed to handle the NPCs, monsters, and most important the description of the locales they explored. Often overlooked is the fact that not only players fought against the NPCs of Blackmoor, they often fought each other as well. Again underscoring the importance of having Dave as an impartial referee.

This role of the referee carried over when Gary Gygax started running his Greyhawk campaign Ultimately was published in the original 1974 rules of Dungeons & Dragons.

The referee is driving force in tabletop RPGs not because of story but because it what enables players to explore a setting as individual characters. The story comes about afterward as a description or an elaboration of what the players did as their characters.

If you want to pretend to be an individual character and you are limited to what only what that character can do then you need a referee to describe the locale you are in, act as the NPCs you interact with, and run the monsters that you fight. In MMORPGs and CRPGs, the referee is a piece of software, in tabletop roleplaying the referee is a human being.

In the past decade several games introduced narrative mechanics to make tabletop roleplaying more about collaborative storytelling. To do this they add metagame mechanics that allows the player, not the character, to determine the action. Either its outcome, whether it occurs at all, etc. The referee has a different role than that of traditional tabletop roleplaying. More like being the chairman of a committee working collaboratively together.

There is some animosity between the two styles mostly because traditional tabletop roleplayers view metagame mechanics negatively. It is viewed as a form of cheating as the one of the original focus of tabletop roleplaying was the fact you were limited to only what your character can do.

Fans of narrative roleplaying view the traditional setup as too restrictive. Sometimes boring because of how bad dice rolls can result in a uninteresting game, along with the negative impact of poor refereeing has on a campaign.

  • 1
    \$\begingroup\$ I don't think this is an issue of "narrative mechanics." Isn't playing the protagonist sufficient to drive the story? \$\endgroup\$
    – Alex P
    Commented Oct 27, 2013 at 23:30
  • 3
    \$\begingroup\$ When you are playing your character "as if they are really there" the game is an experience not a story. Like hiking Yosemite park is an experience. The story is about the experience. Sometime the experience makes for an interesting story and sometimes it doesn't. What makes for an interesting experience doesn't always make for an interesting story. Because this the setup for one approach doesn't automatically work for the other approach. As for the question the referees in the original tabletop RPGs were managing an experience not managing a collaborative story. \$\endgroup\$
    – RS Conley
    Commented Oct 28, 2013 at 14:01
  • 2
    \$\begingroup\$ This answer is as good as, or better than, the accepted answer. \$\endgroup\$ Commented Apr 19, 2017 at 0:18

There's an ambiguity in your last paragraph, about what you mean by 'drives the plot'. The GM sets up the plot, but the players (and the dice) decide how it is resolved. So the GM decides there is a dragon menacing the city, and the characters are sent off to deal with it. A good adventure will have a series of encounters, rising in intensity and each providing more information or assistance (designed by the GM, but who knows whether the party will gather what they should, or even work out something the GM hasn't thought of?). But when the denouement comes, the party don't have to use what they have found in a climactic battle, as the GM intended; perhaps they'll offer the dragon a bribe to go away, or even join it in looting the city.
A good game plot is a collaboration between the players and the GM: the players decide how the characters act (a major influence), but the GM decides (or at least knows) how everyone and everything else reacts.

  • 1
    \$\begingroup\$ "but the GM decides (or at least knows) how everyone and everything else reacts." but why. Couldn't the knight in the group need, for narrativistic reasons, need to kill a dragon, so the player proclaims that a dragon is menacing the city that the group is about to visit? Apparently this is possible, but why is it not common? \$\endgroup\$
    – Vorac
    Commented Oct 23, 2013 at 11:49
  • 2
    \$\begingroup\$ @Vorac It is fairly common. Narrative games are exactly that way. If I play a knight in a non-narrative RPG then I can't just proclaim that there is a dragon, but I can come close. I can tell the GM that I want a dragon and most GMs will accommodate. Some may tell me that I'm asking for a plot-coupon and then agree on the condition that I play along with something else to help the GM's plot later on. \$\endgroup\$ Commented Oct 23, 2013 at 15:38
  • 4
    \$\begingroup\$ @Vorac The indie-game community has this thing called the Czege Principle. Essentially: "when one person is the author of both the character's adversity and its resolution, play isn't fun." Which isn't to say that you can't have some part in declaring that there's a dragon now for you to fight, but the most successful story games generally have all kinds of structure that encourages an interplay between player and GM in establishing situation. Moreover, you don't just get to kill the dragon because you're "supposed to." \$\endgroup\$
    – Alex P
    Commented Oct 23, 2013 at 16:49

The Game Master's role is nuanced and complex. Depending on the game, table, and decisions they make they they can wear many hats:

  1. Rules arbitrator
  2. Storyteller
  3. Setting developer
  4. Mechanics designer

Some of these roles are not universal; someone using primarily prewritten adventures may serve almost exclusively as a rules arbitrator. In order to really understand the GM's role, however, it's important to consider each of the roles in context.


A Game Master has a major part of their traditional role defined by the fact that as they sit outside the game, they can make rules decisions about the play experience. This centralized system has a number of benefits: speed, simplicity, finality, and perhaps above all flexibility. Having a GM means that in cases where there is a rule to govern something you have an expert arbitrator who can cite or refer to a specific rule within the book or within the precedent of play. It also allows for new rules to be designed on the fly, without having to really come to a group consensus that may mean longer discussion or a lengthy and complicated rule where a simpler one is better.


Most tabletop games are driven by some sort of probabilistic method. Probability is a cruel mistress. Probability doesn't care about awesomeness. Probability does not create engaging characters. It allows these things to happen, on occasion, but it's not generally a good idea to leave everything not directed by a player up to the random number gods. Successful GM's are actually often highly skilled storytellers; when I interview my players who have become GM's about their style, I've noticed that those who have a literary focus or background are much more successful at keeping their groups happy, and those that do not often wind up with a discontented player or two.

Setting Developer

I always waver between the term "designer" and "developer", but the truth is that the latter is probably more accurate to describe what a GM does. A well-designed setting provides opportunities for exploration while also providing a consistent framework for events to occur; this isn't to say that mish-mashing together elements can't form an enjoyable game, but a developed setting can be a lot easier to keep track of and form a developed canon of references, elements, and characters for. In this case, the GM provides a setting to explore and be discovered, providing an experience that an open-design setting may not.

Mechanics Designer

I mentioned this loosely before, but the GM will often wind up having to create rules and guidelines to govern game events. However, they're also responsible for creating a reasonable gameplay experience within the guidelines of the established rules. About half the games I see die die because the GM loses sense of scale and creates a monstrous abomination in which the players are never sure if it's time to fill in a new character sheet or if they'll handily crush everything. Obviously, the original game designers usually give this more thought than a GM will, but a good GM should be able to provide appropriate challenges and opportunities for players that don't leave them feeling like the point of a game was to kill as many of their characters as possible.


Wait, NO. The GM drives the plot! So what it main about the main characters?

The DM sets the boundaries and focuses the improvisation.

A lot of people have a hard time improvising (role playing) without boundaries. It's much easier when focused on a clear problem. The DM provides this focus.

That's also why railroading is still a topic. Without any railroading, it's hard to get a game off the ground; and a bit of railroading can often help a game that's going slow. On the other hand, once the player's are engaged and their creativity is sparked, it's important to allow them off the rails.

That also goes for writing. Sometimes you are stuck and just can't find a good way to continue - maybe you think "oh, it should be like this, somehow, but how...?" and you see lots of options and they all suck. Add in a new constraint (best one that you, the author, don't really like), and often this inspires you again. Human problem solving in action.

So in essence, I think the GM drives the plot to help the players be creative, and focuses their creativity by providing constraints.


This is a very interesting question/topic and there are a few games where they have tried to move away from this, with limited success. (examples like Mage the Ascension and Amber), there are also some OGL games that try to use other ways of determining chances and such so DM/GM/ST is less needed.

I think the main reason is people like someone to be in control and it makes it much less complex when a story is a linear progression from point A to C thru B. The DM is the one that controls this.

I have run a few StoryTeller games from White Wolf where I have let my players be the drive behind the story and I then let the story/environment respond to their choices. But this is really a hard way to dm/gm as the game is now open to all kinds of inputs and you have to think on the fly. But I think it makes for a much more fun game if you let your PC's drive the game and not the GM.

So let me say this start a game session where you have a bunch of small adventure hooks and a bunch of npc's that you can grab and run with depending on what your players do and say in game. Let the world respond to their actions and have no set plan on where they have to go let them decide and if they start an adventure and half way go chasing after another then let them that is how life works :-) But maybe they made some unknown enemies on the half completed adventure that now come at them without them knowing why :-)

Look at Mythic Role-playing System(PDF store), it is completely open and you can theoretically be without a DM if you take it to the extreme. Let the game drive itself.

Most Role-Playing Games operate under the principle that there are players and there is a Game Master. The GM prepares all the details of an adventure, and then "runs" the players through that adventure. This usually requires a great deal of preparation on the part of the GM and the handling of many details.

Mythic is different in that it requires no preparation from the GM. Mythic adventures are meant to be played off the cuff, with perhaps a few minutes of brainstorming to come up with the initial setup.

Mythic can also be played entirely without a GM. The same mechanics in Mythic that allow a GM to run an adventure without preparation also allows a group of players to do without the GM.

  • \$\begingroup\$ I am thinking actually even more extremely - the players create the NPC that they need to talk to, they create the quest hook, that they need. \$\endgroup\$
    – Vorac
    Commented Oct 23, 2013 at 9:28
  • 3
    \$\begingroup\$ @Vorac Some games do that, as AquaAlex points out. But some people don't want to play that type of game. They want to overcome challenges and that means they don't want too much control over the environment to handwave the challenge. They may also want to be able to discover rather than create. Both types of games exist with many in between. \$\endgroup\$ Commented Oct 23, 2013 at 15:40
  • \$\begingroup\$ @TimothyAWiseman This is my experience running Burning Wheel. Lots of power for the players to create (or rather, try to via certain rules) the NPCs or things they need, but the players I had weren't interested in that ability and the system simply didn't work well with them. \$\endgroup\$ Commented Oct 24, 2013 at 1:04

After watching on YouTube Viva La Dirt League D&D*'s GM** I conclude these are some of the reasons it happens:

  • Players might be new, while the GM is the one who is expected to care the most, know the rules, be passionate.

  • Metagaming is important. Before starting the campaign do sit together and discuss where on the Story Mode***-Open World spectrum do your friends want to be. All is fun, except for mismatched expectations.

  • A good GM tailors the world to the characters (subtly). To their backstory, to their abilities. An even better GM asks the players to do it!

    • Describe your melee attack.
    • What is important to your character right now?
    • How does your character look, how do they talk, how do they walk?
  • Reward assertiveness (the guy in the video does it extremely well). You want to try something unorthodox? It will probably work!

* - I wanted to advertise the yt channel as little as possible but a more experienced user decided more important is naming it
** who is thousands of times better than me; however the group still exhibits DM-first dynamics
*** - a la Call of Duty — a movie with some combat


In stories, and in life, you always have the same elements:

  • A start state of the world. In RPGs this is established typically by either the DM or the DM and Players through session 0 and player backstories.
  • A Conflict. The conflict has to come from the GM, but can -- and should -- be informed/inspired by player back story. It can't come from the Players or there is no game. No obstacles to over come. When I was young and new to being a GM, I didn't fully get this. Either I would create the goals that I thought were interesting and players wouldn't take the plot hook (because I ignored PCs desires), or else I'd grant the players big rewards for little/no work.
  • Resolution. Overcoming the obstacles is the work of the Players. Choosing to overcome them, and choosing how to overcome them is the meat of the game. "Railroading" is the term for when a GM doesn't allow players the ability to choose to ignore an obstacle or choose their own way to overcome an obstacle. I have no problem with the fact that a small encounter I thought would be a combat to protect a father and child the party met on the road was turned into my players spending a session hiding children in an interdimensional daycare hidden in a magic shop. It wasn't the resolution to that obstacle that I thought the players would take, but we went from family being chased by something to commissioning the building of cribs and buying of toys. The party loved it, and kept playing.
  • Reward. In RPGs the system provides some of the rewards (XP, levels, abilities, etc.) and the DM provides some rewards (gold, titles, magic items etc.) but the biggest reward that keeps the players moving story wise is story and character development. If the players don't feel like their character mattered to story, they won't have fun. A lot of that feeling has to do with the player finding a way to be invested, and the DM has to give the player threads the player can choose to grab on to (or not grab on to).

In some games, it feels like the world exists for the PCs. In others it feels like a bigger world that the PCs happen to exist in. I personally try for the latter because it feels better to me as a DM when the players don't expect I'll bail them out, and they get to discover that things change when their characters aren't around. If they don't take a quest, someone else might step up or the problem will get worse or some other world effect from the choice will occur. If they succeed or fail on a quest it will also have implications, and returning to that town later will show some changes because "life went on" there without the heroes. That is stuff the players can't really do, unless you as the DM find creative ways to do that... Which is giving me an interesting idea for a game mechanic I might throw into my current campaign.


Nature of Heroism

I'll admit that it's been a while since I played D&D, but a common line that gets repeated in the Mutants & Masterminds campaigns is "Villains act, Heroes react" because, in a heroic scenario, the line between "We storm the evil stronghold!" and "We begin the massacre of the orcs!" can be very fine. If the heroes are attacking the orc stronghold from whence are coming raiders, or where the high evil necromancer is preparing his Doom Ritual, then there's still a very worthy cause, but again, you basically need that instigating incident (although it often doesn't take much to get the heroes to ride, and you can give them a lot of possible hooks to work off of).

Prep Work

On the GM's shoulder is having an adventure planned complete with hooks, swerves, and NPCs in a way that is satisfying. That is much easier when you can plan ahead. This doesn't mean railroading, necessarily, since players do get to decide their actions, and it's more fun if players feel like they're driving the story, but it's good to have that framework in place (sometimes multiple frameworks... my trick in most longer campaigns is to have 3-4 adventures open at a time, and mix-and-match hooks around so that my players have choices, and are never 100% certain that they picked the adventure I originally intended, or that they haven't branched over to another scenario entirely) alongside the usual grab-bag of NPCs and generic hooks for when your players decide that the village you just described must be at threat from the Orc bandits, and needs a Seven Samurai training scenario.

  • 1
    \$\begingroup\$ This answer ignores the fact that in many games (including older editions of D&D and its retroclones), player characters do build their own strongholds and instigate their own massacres, and NPCs react to them. It's pretty common in hexcrawl-style campaigns. \$\endgroup\$
    – GMJoe
    Commented May 7, 2022 at 23:18

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .