Classic advice vs. the easier method
Among the genres that people often want to touch on in a TTRPG, mysteries may be the hardest to get right. Getting players to figure out the clues that you’ve laid for them can be frustrating, and inexperienced GMs can find their games in a kind of cul-de-sac, where the players have nowhere to go because they can’t find any way out.
Justin Alexander’s series of blogposts about the three-clue rule and node-based scenarios are where people are often sent when they want to run a mystery in an RPG. That can require a good deal of preparation and still may not have the results you want, if the players don't manage to pick up on your clues.
The other common option is to suggest that the person use a ruleset in which mysteries don't have a canonical solution, such as the Carved by Brindlewood system. That doesn't satisfy people who want a canonical solution that they are uncovering.
I think, though, that there is an easier way to run mysteries than relying on the three-clue rule while still having a canonical solution. A couple of authors have discussed the general concept before. Jesse Burneko wrote a zine about it, entitled Unchained Mysteries. Dwiz discusses it in a post entitled “Action Mysteries.”
Both are worthwhile reads, but I think that they may be confusing to some readers; they were a bit confusing to me. Personally, the idea boils down to two points:
1. Make it an ongoing situation.
In your classic Sherlock Holmes-style mystery, for example, there is a pre-existing situation that the detective must solve through his or her genius. By examining the clues of the case, the detective deduces what the solution is. The problem with using such a mystery in an RPG is well-known: it requires that your players be Sherlock Holmes-level geniuses, and you need to somehow feed them the clues as openly as possible and hope they will understand what to do with them. (Hence the three-clue rule or rulesets like Gumshoe.)
But if you make your mystery something that is ongoing—say, a serial killer is continuing to kill—your players won’t have to only deal with clues. They will have a chance to interact with the present facts of the case. They may have a chance to run into the criminal or be on the spot when a crime is occurring. They won’t have to rely on genius-level intelligence and will instead be able to use the usual tools players use to investigate in other kinds of RPGs: getting involved in the world and fooling around with it.
2. Put the mystery where the PCs are.
Ensure that all of the ongoing mystery stuff is happening in a place where your PCs are going to be interacting already. So not The ABC Murders, where people are being killed all around the greater London area, but something more like a village mystery (or apartment building mystery), where the PCs will learn about what's going on and might run into it as they go about their investigations and other events.
These two elements mean that the GM will have to do less prep in placing clues. Making the investigation involve an ongoing situation in the place where the PCs are located will naturally give the GM the structure needed to place things where they make sense. You just put stuff where it would make sense, and the PCs will likely stumble into it. And running the game is also easier. Rather than feeling like you need to push the PCs to clues, you simply keep the action going around them until they interact with it.
Mysteries are fun, and you should definitely run them for your group. If you follow the two points above in creating them, I think you’ll also find them easy to prep and run.
An Addendum
A friend of mine and I were recently discussing mystery shows that emphasize the emotional or psychological side of the detection, after he watched True Detective season 1. I mentioned at the time that a lot of the shows tend to focus on dour, downbeat, or obsessive detectives in a way that gets a bit overdone or oppressive. (See, for example, The Killing.)
I said that one I suggested trying out was Endeavour, the Inspector Morse prequel show. I just now finally saw the ninth and final season, and I'd just reiterate that I do think that it could work well as a template for a detection RPG campaign (whether you're using cops or less formal detectives) that focuses on personality over simple detection.
In Endeavour, the mysteries and their solutions are sometimes a bit silly or convoluted, but the personalities and emotions of the main characters are a throughline that make the series feel deep and interesting. There are mysteries that overlap. Sometimes there is an underlying mystery that lasts through a whole season, even as the detectives solve individual mysteries each episode. And there are some mysteries that have lasted for multiple seasons. For example, one from the second season returns in the ninth season.
That episodic format that still has overarching stories seems ideal for RPGs, where having recurring characters and storylines is good but dividing the campaign into episodes can be helpful.
I don't think that Endeavour is everyone's cup of tea. It's certainly not the same kind of hardboiled (and mystical) stuff that True Detective is (and has fewer gratuitously nude women than True Detective), but i think it might be even better fodder for RPG inspiration because of its mostly episodic nature that still maintains throughlines and emotional arcs.
I appreciate the intent here. If a character has a skill, don’t roll. Just give them the information. Now no player has to be Sherlock. And in many games, characters are a Sherlock. The level 10 rogue with +20 linguistics and perception? That’s a first-rate detective.
Good article and great ideas!