Five types of RPG game
Recently I’ve talked to several friends who had frustrating experiences GMing an RGP. As I talked with them, I noticed that they seemed to be attempting the wrong type of game for either their experience level, personal disposition, or player group. But I couldn’t find a good description of the types of games I had in mind, so I decided to write this post describing them.
I’m going to assume my reader is familiar with basic RPG terminology and am restricting my discussion to the most popular family of RPGs, those that are designed to support multi-session adventure stories.
Within this genre, I think of RPG stories as fitting into one of five types:
An bounded, interesting environment full of monsters to defeat.
Easy to run and play, the core of the genre, no acting required.
A Dungeon Crawl that also has non-monster denizens to speak with.
Easy to play, require a little improvisation to run, some role-play required.
A story planned out by one person that all the players follow with some limited agency along the way. In practice resembles a series of small worlds selected by the story author.
Harder to run than they sound, some players like the satisfying story but others are annoyed by the limits on their agency.
A Small World that’s not small, allowing players to go anywhere and do anything with neither walls nor a planned story to limit them.
Requires extensive GM improvisation and ongoing preparation and players who are willing to write their own story.
A story being planned as it goes, with sandbox-like freedom and planned-story-like plot elements and eventual resolution.
Fun to watch (the most popular streamed RPGs are of this type) but hard to GM. Some players love the immersion in a vast story, but many find them less than relaxing to play.
Each of these types deserves additional details.
When Gygax and Arneson invented the first RPG, no one had experience GMing, so any design that relied on experienced GMs would have failed. Fortunately, they understood the essentials of the genre, essentials I had no trouble GMing as a 9-year-old: environments to explore and monsters to defeat. They even named their game after these two elements: Dungeons and Dragons.
A tip for making a dungeon crawl survivable: have some reason why the various monsters can’t all gang up on the PCs at once. Most RPGs are built to ensure PC success by letting them act as a group against a series of challenges that are individually weaker than the PC group. If the monsters can get together and attack the PCs as a team, they don’t stand a chance.
Beginners to dungeon crawls generally plan rooms with monsters and loot. More advanced dungeon crawls tend to add traps, mechanisms, puzzles, and echoes of the story of whomever created the environment. Variations on the theme replace the dungeon with a ship, train, pocket dimension, or other type of bounded scene.
The difference between a dungeon crawl and a small world is that PCs meet NPCs who they can engage in conversation. And what a difference that is! It means the PCs can ask questions and expect a reply, and that means the GM has to have the ability to reply to those questions. That said, its limited scope makes it work well as a game, even 11-year-old GMs like I was when I started running this type of game.
A tip for planning a small world: plan out what the NPCs would do if the PCs weren’t there, what an average day in the environment is like. Once you know that, answering most questions PCs might ask becomes much easier than if you tried to plan out answers to likely questions instead.
Beginner small worlds generally limit the NPCs to prisoners and hired hands who have limited knowledge of what’s going on. Intermediate small worlds have full ecosystems with each NPC having opinions about all of the others. Advanced small worlds have multiple layers with secrets, lies, and intrigue.
Published multi-session adventures are generally organized around a sequence of plot beats that the PCs are to pass through. Some are written to allow some beats to be skipped or taken out of order, but many are not. These plans give a GM something to shoot for and ideas for who the PCs might meet and so on, but also require considerable effort making PC actions (which never line up with what the published adventure anticipated) fit in with the planned story. I’ve known first-time GMs with a year of experience as players who successfully started GMing with planned stories, but I’ve also known some with five times as much experience playing who find GMing a planned story to be an exercise in frustration.
A tip for running a planned story: Five hints or clues beats one instruction or request. Players feel much happier when they feel like they’ve figured our what’s going on or come up with a good idea for the next step instead of having the GM tell them what it is. But hints and clues are much easier to drop then they are to pick up on, so drop a lot of them.
Some players are never happy with a planned story. Perhaps it breaks their sense of immersion and fun to feel compelled along a path that they believe their character would not choose. Perhaps they want to escape from reason and consequences and take out their frustration with the world by having their PCs do whatever pops into their heads, even if in any real world it would land them in prison or an executioner’s block. I have no tips for running a successful planned story with such players.
Beginner planned story GMing tends to rely on quest giver- or babysitter-NPCs to keep the PCs on the planned path. More advanced planned story GMs find ways to insert events or NPCs into the story that cause the PCs decisions to lead where the story was going anyway, requiring quite a bit of improvisation and creativity to achieve this end.
Sandbox worlds are like small worlds without walls, and hence not small anymore. The resulting feel is quite different than a small world, especially for the GM who cannot even hope to list, let along plan something about, all the places and people the PCs might interact with. No matter how well prepared the GM is, the PCs will always go outside that preparation and the GM will revert to improvisation to fill the gap. I have known experienced players to cut there GMing teeth on a sandbox world, but more often sandbox worlds come later after initial GMing experiences with small worlds and/or planned stories.
A tip for running a sandbox world: prepare relocatable adventure elements and personalities that can be adopted by whomever the PCs decide to speak with. For a world to feel like a real place you won’t always be able to use what you prepared—don’t plop your crotchety octogenarian personality into the child the PCs decide to speak to—but it can help a lot.
A tip for making a sandbox world feel both interesting and open: have two or three clear here be dragons
-type pointers, things that clearly say if you make this choice you’ll find this type of adventure.
To make the world feel alive, remove some not-taken pointers: the dragon they don’t investigate either wipes out the town that was asking for dragon slayers or is defeated by some NPCs or moves on.
A tip I learned late in my own running of sandbox worlds: as a GM I forget what I improvised quickly, often before the next session, unless I make extensive notes immediately after the session or record what I say and listen to the recording afterwards.
Some players grow dissatisfied with sandbox worlds over time. Because the GM is providing a world, not a story, players who want a story can feel like the game is drifting without a purpose or feel like they messed up in not making more story-like decisions.
Beginner sandbox worlds are only a little larger than a small world and keep the PCs in small regions with few NPCs for several sessions, opening up pathways to other regions only as the GM has time to prepare them. The more time a GM spends running sandbox worlds, the more components they have to build new regions out of and the more open and rapidly-accessible the world can become.
From the outside, an epic tale looks like the ideal variant of a planned story: there’s a sense of progress and resolution, passing through setup, conflict, failed attempt, discovery, determined effort, resolution, and epilogue. From the inside, it feels far more like a sandbox world, where the story goes wherever the PC choices and fortunes take it. Tying these two views together is a lot of labor, dedication, and acquired skill. I’ve heard rumors of first-time GMs successfully running epic tales, but I’ve never seen it done nor spoken with those who have.
RPG epics have a scale and scope that most other epics lack, and usually fill that size with smaller fully-fledged stories in their own right. For example, one 180-hour epic I GMed had a failure beat that was itself a full story that resolved with the successful rescue of refugees from a temple besieged by ghosts; that success of the inner story filled the failure beat of the larger story because they realized that a rescue was the limit of their ability and they had no immediate means of banishing the ghosts. Larger epics (500 play hours for one epic tale is not uncommon) may have a third tier of full stories for each of the beats of the smaller stories that make up the epic tale.
A tip for running and epic tale: keep track of where in the epic arc you are, and only put pointers into your sandbox world that will help provide that next story beat. Particularly important is the failure beat: an epic feels epic in part because the protagonists tried and failed to resolve the conflict and thus learned that it was of epic proportions. But they can’t fail unless they try, so put some pointers suggesting the problem is within their reach, just another random sandbox curio, and have its epic scale unfold gradually.
A tip for dealing with unexpected player choices: it’s OK to have one or maybe even two Jonah-and-the-whale moments where you return them to the plot, but more than that can ruin the story. Often, though, you can re-plan the remaining epic to fit their new direction, re-framing what you had planned to be the core of the tale as a supporting part and putting a new resolution in their path.
In my experience, most players want an epic tale, but don’t really. Playing an epic means investing in and relating to the struggles of the characters and that’s not most people’s idea of fun. Of the dozen groups of players I’ve played with for long enough that an epic could have been told, only two were made up entirely of players who where OK playing an epic tale. Epics are fun to watch, but not everyone has fun playing them.
Beginner epic tales tend to start out with epic tale aspirations but veer off into a sandbox world or a planned story or simply end unfinished. Expect at least two or three such incomplete epic attempts before you get one that works; I had five such warm-up attempts myself, and that was after more than a decade of running successful sandbox worlds.