How to Run a Mystery in TTRPGs

When I have my druthers, my long-running campaigns begin to resemble a long-running sitcom. To wit, bit-players from the early seasons may return to great fanfare, or the characters have genre-infused episodes (unfortunately I have yet to have a Musical Episode - if you have tips on transforming a session of Gamma World into a Musical, I am all ears). As a result, I’ve run a handful and a half of sessions that were one-off murder mysteries, all of which have been a blast.

I’ve finally cracked the case of how to run a murder mystery with minimal prep. I abhor prep in general. I’m a busy dude (and when not busy, I am a lazy dude) and too much prep is not great vis-a-vis bangs for bucks and tends to stifle emergent gameplay. My early attempts at running murder mysteries involved the classic elements of Clues, Red Herrings, Suspects and The Murderer. I will give an example of a little adventure I think is a good encapsulation of the classic way to run a murder: “Guilty!” a mini-adventure in Weird Discoveries: Ten Instant Adventures(TM) for Numenera(TM). Forewarning, I am going to talk about this adventure and the classic method before I give up the secret on how to effortlessly run mysteries. Skip the next couple paragraphs if you are pressed for time or don’t want to spoil Guilty! for yourself.

The classic focus on Clues, Red Herrings, Suspects and The Murderer is a perfectly cromulent method for running mysteries. In Guilty!, you have a city segregated between humans and varjellen (aliens) as well as an understated class divide. There have been 4 connected murders in the past two months, all varjellen. Everyone suspects a “self-proclaimed varjellen hater” in town (our Suspect), but the real killer is some sentient mist under the bridge dividing the city (our The Murderer). There are a few Red Herrings suggesting guilt for the Suspect (his self-proclaimed hatred, a witness who saw him at the bridge on the night of the third murder, the Suspect’s flight when confronted). There are also Clues that point to the Suspect’s innocence (an ironclad alibi for the first murder, a arthritic arm unable to strangle the victims, a witness that admits he was lying if grilled) and the identity of The Murderer (the 4th victim described the mist to their housemate, a witness that saw the mist on the night of the 5th murder [more on that later]). Honestly, The Murderer is the weakest link of this adventure, and I tend to replace it wholesale in favor of a murderer who isn’t sentient mist living under a bridge.

Too often, the characters come in after-the-fact and the world reacts to them. What Guilty! really does well is not presenting a static mystery. Guilty! has two moments that force characters to react to the world. The first is a vengeful varjellen who finds the PCs midway through the investigation and demands they reveal the murderer’s identity so it can extract vengeance. The second and more major development is that a 5th murder takes place while the characters are investigating. The victim for this murder is a human, throwing a wrench in the efforts of the local constable and others to pin the murder on the Suspect. Overall, the mystery is very easy but that is by design. In a TTRPG, it is gospel that not all clues will be found and solving even a simple mystery is a rewarding experience for a team of murder-hobos.

The problem with the classic method for running mysteries is its relationship with the truth. There are Clues, and there are Red Herrings. Clues lead to the truth, and Red Herrings elide the truth. There are Suspects, and there is The Murderer. Only The Murderer represents the truth, and catching them is the goal of well-meaning adventurers. But what if all Clues are potentially Red Herrings? What if all Suspects are potentially The Murderer? All gamemasters have likely had the experience of listening to their players talk, strategize and theorize amongst themselves (always a welcome reprieve) and thought “oh that is a cool idea, I will do that instead!” The players are none-the-wiser that you changed “the truth” based on their conversation and, instead, feel like Sherlock Holmes for having predicted it. I prefer this collective method to truth-making. If the players make deductions that make sense based on the evidence they have, it is fun to go along with it. It is un-fun to pooh-pooh their deduction based on the Clues they failed to uncover. The truth about truth in TTRPGs is that nothing is True unless it has been said out-loud at the table (or virtual table). If I never tell the players that the Suspect has an arthritic arm and they never discover this for themselves, then the Suspect doesn’t have an arthritic arm. By speaking a fact at the table, you reify that as a truth. All clues unuttered are merely potential truths, waiting their turn.

To prep a murder mystery with the better approach to truth, you should know the crime scene(s), have some Suspects in mind (each a potential Murderer), and sprinkle ample Clues (each a potential Red Herring) pointing all which ways. No Clues should be mutually exclusive or contradictory. If the players uncover every single Clue, it should help, not hinder, their ability to fully deduce what happened. If all of this is too philosophical, I will show these principles in action in a short, surprise UVG play-report.

Art: Luka Rejec - wizardthieffighter.com

That lovable scamp, Jonky Bonko, hired the Company to clear out the Glass House of a Dead Prince, some recently acquired real estate, of potential invisible wolverines. This location is already a murder mystery in the Ultraviolet Grasslands, but I took a few liberties with it. First, I increased the body count, an always advisable practice in a murder mystery. The previous owner and murder victim is now a Porcelain Prince polybody and there are two dead bodies in the house instead of one, but pictures and furniture in the house suggest that Satrasco 3-Body had an additional body in tow. When the party discovered the first body, the gun was still held tightly in the polybodies’ hand, but our intrepid investigators never deduced that was suspicious given the gun’s recoil. The second body was drowned in a bathtub, an apparent suicide. The party learned that the invisible wolverines were part of a security system that had been tripped. That along with a love letter in the first corpse’s desk addressed from a J.B. immediately led the party to suspect their employer of not being entirely on the up-and-up. Curious about the potentially missing 3rd body, Sulmar asked whether polybodies offered spare-body life insurance. I answered that, given the cost of the bodies and the fiscal conservative attitude of polybodies, such financial instruments were available but expensive.

Thus enters Lorenzo Maltablano, legendary P.I. and far-traveling orange-lander. Inspired by Knives Out (what a wonderful movie), Lorenzo is solving both the murder mystery and the mystery of who hired him. He received an envelope of cash addressed from Satrasco, stating that Satrasco feared she would soon be murdered. But Lorenzo was suspicious from the start. Lorenzo was a random encounter. Another random encounter that never materialized was some debt squeezers from the Yellow Lily Finance Cooperative come to shake down Satrasco for her massive debt. But because it never materialized, it simply wasn’t true in my game. Although they agreed to cooperate with Lorenzo, they did not particularly like him. Their cooperation was also suspect, as they took the murder weapon as loot but convinced Lorenzo that the gun was already missing when they arrived. The suspicion swiftly became mutual. Rian cast a spell, Shadow of the Past, which causes any shadows that were cast over an area in the recent past (how recent is based on the effort put into the spell) to reappear. The shadows revealed that Lorenzo had arrived before the party, that morning, instead of in the late evening, which is what they told them.

When they caught Lorenzo in a lie, the tension escalated to an armed standoff in the twee gardens, involving a sentient lawn gnome Rian had secretly befriended. This all played out over the course of two-and-a-half sessions and many shenanigans are omitted for narrative semi-clarity. In the end, Hakon said he thinks he cracked the case. Although many party members (and Lorenzo) suspected Jonky Bonko and a few suspected Lorenzo, Hakon said that the missing body gnawed at him. He agreed that Jonky was greedy and his love letter disingenuous at best (not to mention riddled with spelling errors), but Jonky was being set up. Satrasco knew that Jonky was an unfaithful so-and-so who just wanted the house. Satrasco staged the two suicides and took off, hired Lorenzo and went to collect the polybody insurance payouts. The players all excited agreed that this must be the Truth. I thought it was a good deduction and decided that I would go along with it, although as a matter of principle, I would never say “You solved it!” That just isn’t how I run things. However, I had Lorenzo say that that explanation made sense to him. Had the players made a good, logical case for either Jonky’s or Lorenzo’s guilt, I could have believed that too. But my job was just to present clues and let the players deduce their own truth that logically followed those clues. The players felt smart for solving a tangled web of mystery and I felt like all of my minimal prep served me well. This may not work for you, but if you have the same quantum approach to truth and gamemastering, I would give it a try the next time the lights flicker in your game and the butler drops dead.