Procedural, Bespoke and Improvised Content in Roleplaying Game Sessions

Roleplaying games are played in sessions where the participants bring the experience to life through improvisation. Often, but not always, these sessions are based on published roleplaying game books and use published and pre-designed game mechanics.

When you examine the actual play of the roleplaying game session, consider: Where does the content come from? What sorts of processes lead to it appearing in play? Although there are many different types of roleplay content from scenes, storylines, events and interactions to combat, they’re all generated by three basic processes.

These are bespoke design, procedural design and improvisation.

In bespoke design, someone handcrafts material that appears in the game piece by piece. The person doing this can be the designer of a published release, the Game Master or even a player. As an example, there are many published adventures for Dungeons & Dragons that feature preplanned maps, scenes, stories and characters. As the players explore the adventure, they encounter these individually created elements. In a Fastaval game, the bespoke design can manifest in a list of preplanned scenes.

In procedural design, the game has a system or other mechanisms which generate content procedurally. As long as the participants follow the mechanics, things keep happening in the game. Many Storygames work like this.

The division between bespoke and procedural design is common in videogames. An individual videogame level can be handcrafted or generated by an algorithm. Both choices have their own implications in terms of how it feels to the player and how the labor of creating the work is organized.

With improvisation, the content of the game doesn’t flow directly from the design of the materials used for play. The participants just make it up on the fly. Sometimes this happens because the design of a published game has blank sections that have been left there purposefully to be filled with player creativity. Other times the players just ignore the game they’re playing and go freestyle. Or there’s no game to begin with and everything is improvised.

There are many creative stakeholders in a roleplaying game session. If the participants are playing a published game or adventure, the designers of that publication influence the game. The GM (if there is one) contributes creatively, as do the players. The question of who among these stakeholders brings any particular piece of content into the session is an interesting one but not what we’re looking at here. This is about how, not who.

These categories are not pure in the sense that most actual game sessions contain a mix of different content types. For example, a combat encounter in the D&D adventure Lost Mine of Phandelver combines procedural and bespoke. The set up and narrative context of the fight is bespoke while the actual combat action flows from the procedural design of the combat mechanics. If the players like to veer from the rules into made-up combat maneuvers (“What do I roll to swing from the chandelier?”), a significant amount of improvised content comes into play as well.


The original example of procedural content generator is the random encounter tables found in the earliest published roleplaying games. Throw a die and the table tells you that three orcs attack. Although it wouldn’t be very exciting in the long term, theoretically you could run an entire session just by using such a table and stringing one fight after another, one for each room in a dungeon.

That’s the core idea of procedural design: If you follow the mechanics provided by the game, you’ll have content for your game sessions. It’s good to remember that many published roleplaying games don’t do this at all. Castle Falkenstein is a good example. It provides a world, character creation and task resolution. It doesn’t provide almost anything you could run straight out of the box, except for a rudimentary example scenario a few pages long.

The Dungeon Master’s Guide for 5th edition D&D has a system for randomly generating dungeons and adventures. That’s an example of procedural content generation, much more sophisticated than the random encounter tables of earlier D&D editions. Use the system and you’ll have the outline for an adventure for your game session.

Many OSR games offer anarchic examples of the same design ethos. The scifi game Death in Space has tables for everything from personal trinkets to space travel.

Combat in many roleplaying games combines bespoke elements with procedural mechanics. In D&D combat, the bespoke elements are things like Owlbears and Magic Missiles while the procedural elements are the system of rounds, to hit rolls and hit points organizing the flow of events.

The Finnish roleplaying game Lamentations of the Flame Princess uses the term “chaos bomb” in The Referee Guide of its Grindhouse Edition. A chaos bomb is a magic item or a spell with a worldshaking, random effect. For example, if a character uses the item, the player rolls on a table and as a result, volcanoes erupt all over the game world, destroying major cities. This is an example of a procedural mechanism injecting content into the game.

Fiasco is a great example of a game designed largely around procedural principles. It’s built from the ground up to generate content, making it possible to play with little preparation. Although there are pre-prepared playbooks providing the rough outline of a scenario, in practice almost everything in a session is created in the moment by the participants following the rules.

The distinction between procedural content and purely improvised content is in the system. Fiasco has mechanisms and structures that guide player creativity towards desired outcomes. Thus, you just have to follow the rules with some enthusiasm and content will happen.

The great advantage of procedural content for a game session is that nobody has to make it by hand. If the procedural system is good, content happens seemingly by magic. The system guides player creativity so that they come up with fun, engaging things.


Looking back at the history of published roleplaying games, the earliest examples of bespoke content design are D&D adventures like Palace of the Vampire Queen. It presents an environment the characters can explore and going through the material in the booklet is enough to sustain a game session.

It’s interesting to note that Gary Gygax originally thought that creating bespoke game content would be one of the chief pleasures of being a GM. Because of this, he assumed that readymade published adventures would not be successful. This resulted in a wealth of third-party D&D adventures in the early days of the game before TSR started publishing its own.

The two main modes of bespoke content design are published adventures and the auteur model of GM. A published adventure presents a story, locations and supporting characters for the session to go through. Right now, publishing readymade adventures is very much in vogue, with the industry leader D&D mainly supported by smaller online scenarios and big campaign books such as Curse of Strahd and Out of the Abyss. Old classic adventures are being reprinted as new expanded editions such as Into the Borderlands by Goodman Games and the Masks of Nyarlathotep revival for Call of Cthulhu by Chaosium.

The auteur model of GM means that the GM makes up the content of the game, sometimes using a published game as basis. This is the model Gygax was thinking of when he predicted published adventures would not be successful.

When I was a child playing Red Box D&D, I spent a lot of time drawing dungeons of my own. In this sense, I started as a little Gygaxian content creator.

The Fastaval tradition of scene based scenarios where players go through preplanned situations at a rapid pace is a great modern example of bespoke session content. In these games, the script tells you where to go next after the current scene has been exhausted. Play the scenes in order and you have played the game.

There are some definite advantages to bespoke design. It’s great for creating coherent, deep environments. You can have meaningful secrets and the believable illusion of a world bigger than the characters. Done well, it makes the game feel real.

Another benefit is the possibility of introducing new elements and ideas that are not obvious. Although procedural and improvisational content creation produces many surprises, they also rely on what the participants already know. Improvisation flows from whatever you have in your head. For non-obvious, new and strange game elements, you have to go bespoke.

The downside of bespoke content is that it’s a lot of work. If the GM creates the content, it can take hours of preparation for each game. It also sometimes lacks the energy and the dazzle of more improvised content types.


This is a tricky category because it doesn’t seem like design at all. Instead of having a designed source for the content of the game, the participants just wing it. Sometimes instead of relying on anything prepared, the participants really do just make it up.

Theoretically it’s entirely possible to improvise from a blank slate but I suspect that the most common scenario is in the context of an existing roleplaying campaign. Thus, there is history, existing world and characters already known to the players and the GM. They are not improvising based on nothing.

Rather, the situation is that every participant, including the GM, has arrived to the session unprepared. Nobody has anything when the game starts.

The first thing the GM says is: “What do you do?”

At this point, a player can go: “Well, in the last session, we were talking about founding a bar, so I think we could make that the goal for this session.”

From this point onward, the players improvise what their characters do to fulfil this goal. The GM improvises the necessary world elements (for example, different candidates for existing bars the characters could buy). Everyone improvises relevant social scenes, like a disagreement among characters about the bar’s decor. Perhaps the GM can improvise a municipal health inspector once the bar is up and running.

This example shows that improvisation doesn’t have to be based on nothing. It can build on existing material in the fiction and use the structures present in the game. Still, if we evaluate the content of the session honestly, it’s apparent that it was mostly just made up on the spot.

Note that the game can have systems that guide and control the process of improvisation towards a pre-designed goal. That would be procedural content generation.

A huge advantage with improvisation is that when energy is flowing and everyone is elbow-deep in the action, it’s just great fun. You feel that these ideas, scenes and situations come out of nowhere, coalescing out of the collective creativity in the room.

The downside to improvising is that it’s unreliable and hard. If you have participants who are used to it and good at coming up with new stuff it works well but otherwise there’s a big danger of the game just fizzling out. With bespoke and procedural content you have something to lean on. With improvisation, you have nothing but a wing and a prayer.

Some games feature settings that make improvisation easy. One example is Mike Pohjola’s cyberpunk game Tähti in which the characters are teenage girls in a band. Experience with the game demonstrates that first time players quickly grasp the social dynamics involved and are capable of improvising entire scenes about issues such as creative differences about the band’s direction without any further support. Playing teenagers makes social drama easy to come up with and the band gives it context.

Another game that has a good basis for improvisation is Vampire: the Masquerade. The basic idea that the player characters are vampires living in a modern city suggests obvious actions such as going to a bar to hunt for blood. These are relatively easy to play out improvisationally.

If you’re organizing a roleplaying game session, one thing you can do to foster improvisation is to invite participants you know to be good at it. Player selection can also be design.

Pure Forms

It’s a rare game where the content of the session appears purely as a result of one of these three types. It’s much more common that the session is a mix of all three, perhaps tilted in the direction of one of them.

I’ve run many sessions where I’ve prepared materials beforehand with the assumption that the majority of content will be bespoke. Then, once the session has been underway, the players have started improvising with such creativity that I’ve discarded my prepared scenes and gone with their ideas instead. That’s an example of a game shifting from one type of content source to another.

The role of procedural mechanics can be pretty much the whole of the game, as in Fiasco. Or it can be just a bit of extra, as happens in games where mechanics are mainly used for combat and other types of activities are played out in a freeform, improvisatory manner.

Looking closer at what happens in Fiasco, it becomes increasingly difficult to separate procedural generation from improvisation. If the mechanics haven’t been referred to in the last half an hour, has the action shifted to spontaneous, unsupported improvisation or are we still being propelled my mechanical inspiration?

One variable is player and GM skill. Some players are more used to pure improvisation while others are good at engaging with systems. Still others respond well to bespoke concepts and ideas. In these cases, different content strategies can play out very differently depending on the individuals playing.

Related Post