The Basics of Roleplaying, Part Four: Approaches to Gamemastering

In the last three parts of this series, I’ve covered what roleplaying is, why we enjoy roleplaying games, and how to be a good player. Now we turn to part four: how to GM.

There are so many aspects to gamemastering that no single blog post could cover them all. This is simply an introduction to the basic approaches, and I encourage you to explore the many other resources available.

When it comes to gamemastering, in my opinion there are basically three approaches: (1) prepare a railroad for the players; (2) prepare a situation; or (3) improvise on the spot.

Railroading

The term “railroad” is often seen as pejorative, but I don’t mean it that way. Similar terms include “linear adventure.” What I mean by railroad is that you, as the GM, have prepared specific encounters in particular locations that will occur in sequence, from the opening scene to the final scene of a session (or even a campaign).

Many people react negatively to this approach. A common criticism is that it reduces player agency—the idea that player choices should meaningfully impact the game world. If you have player agency, that means had you made different choices, the shared fiction would have developed differently.

Although there’s some truth to this, I’ve noticed that some people have a limited notion of player agency. For example, some argue that a dungeon with only one entrance is railroading because players must use that entrance, while multiple entrances provide meaningful choice. This view is quite narrow. What significant difference does it make if there are multiple entrances when players have no choice but to enter the dungeon? And what if, regardless of which entrance they choose, they’ll encounter the same group of goblins because that’s what the GM planned as their first encounter?

While railroading isn’t my preferred approach, it can work well for some groups. Players still have choices in how they deal with each encounter, so they still have some agency – their choices matter in terms of what happens in that particular scene. Whether this level of choice is satisfying enough depends on player preferences. So I wouldn’t declare railroading inherently flawed—it’s simply a matter of preference.

If railroading appeals to you, plenty of resources are available online. Most mainstream GM advice, particularly for Dungeons & Dragons, focuses on this approach. Check popular YouTube channels about being a good Dungeon Master, and you’ll find they’re primarily discussing railroading techniques.

But a word of caution: ask yourself: do you enjoy GMing this way?

Here’s where railroading often encounters problems: you’ve planned a series of encounters based on your own logic, but your players may think differently. Typically, players draw unexpected conclusions from encounters and head in directions you haven’t prepared for. This creates pressure to get them “back on track,” forcing them toward encounters you’ve prepared rather than where they want to go. This can frustrate players and diminish their enjoyment.

This is why I personally don’t recommend railroading.

Preparing a Situation

Another approach is what I call “preparing a situation.” This means creating a fictional scenario with inherent, dynamic tension that engages players with something meaningful to their characters. Essential elements of a situation usually include several NPCs with conflicting motivations and goals, that create friction with each other and with the player characters. It’s also a good idea to include a few interesting locations, items or objects, innocents who need saving, or other people or things the players are likely to care about.

With this style of gamemastering, the next essential element is to prepare an opening scene that informs players about the situation—just enough to get them started and see what’s at stake for them. After establishing this opening, you simply have your NPCs and the world respond naturally to player actions. Players will react to these responses, and the NPCs and other elements will respond to this reaction, and so on, creating a cycle of action and reaction. There’s no preplanned climax or encounters beyond the opening scene; outcomes are determined organically through play. Eventually a climax emerges naturally, rather than being planned in advance.

In this style, you focus on playing your NPCs and the setting rather than trying to manage the actions of players. I’ve written in detail about this approach (including with fleshed-out examples) in a document you can download for free:

https://drive.google.com/file/d/1Lf–HTDeQ7o2SgNuiSeQazuH-eRKd-Bq/view?usp=drivesdk

Improvisation

Finally, there’s on-the-spot improvisation, where you prepare nothing and make everything up as you go. This might seem entirely different from the other approaches, but I’m not entirely convinced. Even when improvising, our ideas come from somewhere—setting concepts, NPCs, etc.

We may think we’re creating everything spontaneously, but if we have a specific endpoint in mind—like a climactic confrontation with a villain at a particular location—then it’s still a kind of railroad, just without predetermined steps. Conversely, if we’re simply playing NPCs responding to events and player actions without a planned endpoint, that’s essentially the “situation” approach described earlier. Is pure improvisation truly its own method?

Regardless, if you do want to GM improvisationally without preparation, having some resources for inspiration helps tremendously. Random tables, computer-generated settings, NPCs with motivations, random taverns, towns, dungeons, treasures, and monsters can all provide elements to riff off of.

Over to You

These are my brief thoughts on the basic approaches to gamemastering. There’s much more to say about running specific encounters, making combat exciting, and other topics—much of which depends on your specific game system. Many people have already written extensively on these subjects.

So, what do you think? Are there other basic approaches to gamemastering I’ve missed? What approach do you use? What tools or resources have you found helpful? Are there other topics worth covering in this basic series? I welcome your comments.​​​​​​​​​​​​​​​​