Mystery | Dungeons & Dragons
The Art of Mystery in RPGs
Five principles for better investigation adventures
Dungeons & Dragons and all similar tabletop RPGs have three classic “pillars of play”: combat, social interaction, and exploration. That is, the game’s action consists of some combination of fighting, talking, and discovering things.
The third pillar — exploration — is all about discovering information. What’s over that ridge? What’s around that corner? What’s the significance of this item? What’s the evil cult plotting? All these questions involve exploring the story world and uncovering its secrets.
Great exploration-centered games catch players at the perfect point between knowledge and obliviousness. On the one hand, if players know everything, there’s nothing to discover. There’s no space for wonder or anticipation. On the other hand, if players are entirely oblivious, they lack the knowledge required even to ask compelling questions about the world. For example, the evil cult’s plan only becomes a point of interest once the players are introduced to the cult’s existence, and traps only become scary once the players discover them.
In the space between lies mystery: the thrill of an incomplete puzzle to which the players have some of the pieces and are driven to assemble the rest.
I defer a discussion of the general role of mystery and secrets in RPGs, and how game mechanics support these, to a future article. Here, let’s focus on mysteries proper: adventures that center on a core unknown (e.g., who committed a murder) and put players in the role of investigators.
Many others have laid out time-tested techniques for how GMs can run mysteries well. In what follows, I summarize these methods as five consensus principles that can work in (almost) any game system or setting.
Principle 1: Think in information flows, backwards
With most game styles, a GM who is comfortable improvising can wing it, making up characters and plot points on the fly, and end up with a satisfying result. But this approach seldom works for mysteries. The plot is too tight and requires too much internal consistency.
Moreover, much of the satisfaction of working through a mystery comes from knowing that there is a right answer from the outset and finding it out. If the players suspect that there is no such answer (e.g., the DM hasn’t decided who committed the murder), the mystery loses its stakes. (In other words, keep quantum ogres to a minimum when constructing mysteries.)
So, start from the conclusion and nail it down. Who committed the crime (for example)? Why? How? Where? Be as concrete as possible.
Then work your way backwards. Think: what information would reveal the answer? Then for each answer (clue) you give, ask the same question, creating clues to those clues. Together these linked clues form a network of information flows. Each node in the network is a clue. Each line connecting clues is either a lead to the next or a direct inference.
The network serves as a map of the possible routes the players could take through the mystery. In reality, the party won’t perfectly follow any of these routes. It may abandon leads or skip over clues through insights you didn’t foresee. Nonetheless, the map provides a useful visualization of the party’s information environment, just as a dungeon map visualizes a spatial environment without prescribing the stories that will unfold there.
The remaining four principles are all about massaging the shape of the clue network.
Principle 2: Give redundant clues
Perhaps the most common advice for running RPG mysteries is this:
For every piece of information, give multiple, independent clues.
This advice was originally proposed as the three-clue rule on The Alexandrian blog, and giving at least three clues per piece of information is standard.
Giving multiple clues works because it makes the clue network robust in two ways. First, the players can miss one or more clues but find the last one and still discover the key information. They might not search the environment where a clue is hidden. They might make nothing of a clue only mentioned in passing. Or they might fail a dice roll that would have revealed the clue (but see Principle 3). Each of these possibilities presents a point of failure in the clue network. To keep the information flowing, provide multiple clues.
Second, the players may find the clue and know it’s important but misinterpret it or fail to interpret it at all. What’s obvious to the person constructing the mystery often isn’t obvious to those trying to navigate it. Once again, additional clues can get the party back on track by eliminating bad interpretations and lending extra evidence to the right one.
Note that these clues should be fully redundant. If you have to have Clue A to make sense of Clue B, then there’s no redundancy, no added robustness. When you include these sorts of dependent clues, you should include multiple of each — backups for Clue A and for Clue B.
Principle 3: Give clues automatically
An early insight from Robin Laws’s Gumshoe System for mystery RPGs is that interpreting clues is engaging but finding clues is not. So, it’s best to make it easy or automatic — and never random — to find them.
Concretely, this means: do not make players roll to find clues. If the characters take actions that could reasonably turn up a key piece of information, they find it. If they have special skills or abilities, perhaps they find bonus clues — shortcuts through the clue network — but the essential information is available to everyone to interpret. Hence, the mystery is always solvable without dice.
The effect is that players don’t run into dead ends due to bad rolls. Once again, the whole clue network becomes more robust because no links are missing at random. In my experience, players also maintain a greater feeling of agency and more satisfaction in solving the mystery, since outcomes are less a matter of luck and more within their control.
Principle 4: Connect clues nonlinearly
Your job as a GM in a mystery adventure is to help players experience the satisfaction of solving it through their own choices. This means preparing situations — plural — that the players could choose to enter, each of which (within reason) could lead to the mystery’s solution, even if some routes are more direct than others. In terms of the clue network, this means creating a web of clues, not a linear, forward path.
Planning multiple paths has two benefits. The first is that by preparing a web of clues, you force yourself to think broadly and openly about alternate directions the story could take and so reduce the chances of you getting attached to a single desired outcome. You’re less likely to railroad your players if your own preparation looks less like a story with a beginning, middle, and end.
Second, a web is more robust than a chain. Having multiple sequences of clues is like having redundant clues but at the scale of the whole network: if the players miss one direction, there are still other leads to follow to the same conclusions.
(For more advice on how to avoid railroading in mysteries, check out this excellent discussion on Dungeon Dudes.)
Principle 5: Prepare floating clues
No matter how robust the web of clues you planned seemed to be, sometimes the players get way offtrack. They miss or misinterpret every clue or decide to take a very different route.
One possibility is that the players are in fact more interested in something else in the story than the mystery you’ve planned. In that case, the best course may be to follow their interest rather than railroading through a plot of your own interest. But other times, the players remain invested in the mystery but take a drastically different tack. They’re “off the map” of the clues you’ve devised, and you have to improvise.
When this happens, it’s useful to have prepared floating clues, new clues to introduce as needed. For example, you might have the villains slip up somehow, revealing clues of their latest crime. Or you might prepare a list of observations by random witnesses on the street that the players might encounter. (Seth Skorkowsky gives great advice on floating clues here.)
Floating clues work because they balance your two main goals as a GM running a mystery: helping your players experience the satisfaction of solving a mystery and presenting them with real challenges that allow them to achieve it on their own.
On the one hand, floating clues help get the players back on track. On the other hand, if done well, they don’t give away the game — the players must still interpret them in the context of the other clues. They retain agency and involvement in a way they wouldn’t if you simply told them where to head next.
In terms of the clue network, floating clues are clues that no other clue leads to — they are rather introduced at the GM’s discretion, for free — but that lead back to the rest of the network. Defined like this, every mystery involves floating clues: the inciting incident that first reveals a crime is given for “free” — nothing else leads to it. And even in mystery novels, where, unlike in RPGs, the author plans every inference the detective makes, the protagonist often catches some lucky breaks at critical moments.
Floating clues make the network yet more robust by allowing its structure to adapt flexibly to the players’ actions. They’re not a substitute for improv, but they are a useful supplement.
Other advice
The five principles above summarize most of the advice I’ve encountered in my research on RPG mysteries. But a few more common techniques are worth mentioning:
Twists are good, red herrings are not
False clues that lead to dead ends can add intrigue to mystery novels because the detectives only make the mistakes the author chooses for them to and usually interpret clues correctly in the end. But in the RPGs, players will get themselves off track without assistance. Adding purposefully misleading clues usually just creates confusion and frustration.
What’s better is creating clues that provide unexpected answers — they seem to say one thing but are revealed to say something else. Such twists add interest while still helping the players reach the correct conclusion.
Anticipate and incorporate the player characters’ abilities
Fantasy RPGs like Dungeons & Dragons feature divination spells and other abilities (e.g., detect thoughts, divine sense, scrying, and zone of truth) that allow characters to gain information in supernatural ways. These abilities can easily render a mystery trivial if you don’t plan for them ahead of time.
There are two good solutions. The first is to ban certain abilities when players create their characters. The second is to plan the mystery with the magical abilities in mind. For instance, you can craft clues designed for the spells you know your characters have.
What you should not do is allow your characters to gain abilities and then thwart those abilities entirely, for example, by deciding that the villain has taken all the right magical precautions. Players choose character options because they want to use them, and you should design the adventure to enable that.
Consider the accusation and revelation scene
The climax of a mystery story is the scene where the protagonists solve the mystery. They accuse the villain and, usually, there is a revelation that they were correct. Players will expect such a dramatic scene at the end of the adventure because they will be unconsciously familiar with the genre conventions.
So think about what a satisfying resolution could look like. If you’re playing a combat-focused game like D&D, perhaps there will be a final battle between the player characters and the villain. Or if the mystery is just one arc in the campaign, perhaps you should reveal that the mystery is part of a larger conspiracy.
In any case, come up with multiple conclusions. Think broadly. And don’t railroad. It’s helpful to have creative ideas in your back pocket. But what will ultimately be most satisfying to your players is the conclusion that empowers their way of solving the mystery, which may not be one you anticipate.
Further reading and watching
The tips above are just a few key highlights of the wisdom the RPG community has accumulated for running mysteries. I’ll leave you with some sources, mostly video essays, that I found insightful on the subject:
- The Alexandrian in his original blog posts, here and here, and in a video essay, here.
- Pointy Hat, here.
- Dungeon Dudes, here.
- Dungeon Masterpiece, here.
- Enter the Dungeon, here.
- Master the Dungeon, here.
- Seth Skorkowsky, here.
May your adventures be mysterious and your investigations incisive.
