In Three Clue Rule the Alexandrian rightly criticizes roleplaying scenarios based on a trail of clues, and he proposes tripling the number of clues, but he is still putting cart before horse. Deliberately placing clues is still trying to cause a specific turn of events (“I want the players to go to the restaurant so I’ll place the matchbook in the guy’s pocket”). You are trying to predict exactly the players’ thought process. That is bound to be an exercise in frustration.
Designing a mystery doesn’t have to be about designing clues. Just design the actual mystery, the actual case in all its absurdities and horrors, and commit to that case. So let’s say you have the case of a tall man climbing a window and stealing a painting. Maybe the players somehow get the idea they can measure the stride, as Holmes does in Alexandrian’s example, and if they do, you don’t need to have thought of that clue in advance. You just need to know the actual true case: you committed to the man being tall so you can say that the stride is long. More likely, they won’t think of that at all but they’ll think of something else: measuring the height of the writing, as Holmes also does, or the depth of the foot prints in the flowerbed, or extrapolate from shoe size, or… you don’t have to predict.
You absolutely know the true case, what happened and when, the perp—you know he went into the window. Maybe you didn’t think “footprints in the flowerbed” but you know for sure he crossed the back lawn and went into the window so when they start looking for tracks, you ask them what they’re looking for and if they say front yard you can go “nah” but if they go back lawn, and or window, then you can think “yes, they are looking in the right place” and say that there are tracks there.
When you are placing specific clues, you are not only expecting them to fit into a rigid, pre-written Sherlock script. You are also expecting yourself to be a meta-genius, a hypersherlock that can predict all other sherlocks. That’s a tall order. Instead, just create the case and leave the sleuthing to the players. They might surprise you with what insights and conclusions they can make, and in what ways they “test” their case against your environment by asking specific questions about spatter patterns, prints, witnesses etc when you let them be their own sherlocks and draw their own conclusions instead of trying to match yours.
Do not be illusionist or quantum with the truth in the case. That is not what I’m saying here. There is a one true solution: but not just one true way to that solution.
If you can find the board game “Zendo”, it’s a great lesson in creating and running mysteries like I’m proposing here.
It’s possible to accidentally create un-find-outable truths and that isn’t the goal here. In a sandbox game the way I run it, it’s fine if some questions aren’t answered, but in a mystery game you want to make sure that the perp isn’t too much of a super ninja. Just like how in Zendo, there are puzzles that are impossible to anyone who hasn’t memorized the OEIS, you probably could cook up some scry-and-fry crimes that are unsolvable. Don’t do that. It’s possible to not do that without having to predict all possible clues and theories. Just don’t deliberately try to wreck the game by making it too hard, and then if you accidentally make an unsolvable case, come clean to your players.
I first learned this way of running mysteries by reading this thread.
As a simple example from one of my sessions, a player asked if a particular room was the residence of a magician. I asked “How would you even find out whether that was the case?” I genuinely had no idea how he could know that. He said “Those boots you mentioned. Were they magician’s boots?”
I had to say yes.
My notes said in hard writing that this was indeed the magician’s room. I had mentioned the boots there, and I had in mind that they belonged to the person whose room it was. But in my mind, “boots are boots” and I didn’t think of them as a clue at all, but it has in previous sessions been well established that clothing from the Windsea school is distinctive. That wasn’t something I thought of or remembered, but the player did. That’s the amazingness of this style: your players can out-sherlock even you, the DM!
I think the biggest difference in my style vs the style as proposed in that thread is that I don’t do the “blocks”, I just run a normal sandboxy game that happen to have some mysteries in it. I also run with NPC witnesses that can lie, if they have good reason to (maybe they are afraid, or they want money). To me as DM, when the player characters interrogate the witnesses, that’s like the best part. Having them pick up on subtle things the NPCs say is fun.