It Is Impossible To Write A Mystery

November 11, 2026

A mystery is maybe the hardest type of fiction to write.

Here's the thing: good advice for other types of writing is often terrible advice for writing a mystery. It kind of turns the whole process on its head, because you aren't writing a story- you're writing a game. These mediums ask different things of the writer.

One good piece of writing advice is to avoid trying to write for every audience. You do want to make sure your ideas are coming across, you do want clarity and approachability, but there's a point where you say "okay, this is good enough- if the reader at this point doesn't get it, it's their fault, and the worst that can happen is they don't fully appreciate the piece". You don't need to dumb down your ideas for your most thoughtless reader. (And, other side of the coin, you don't need to inflate your ideas to seem impressive to the most erudite reader.) You put your ideas on the page, apply some spit and polish, and then the writing is out there. Maybe a reader doesn't absorb every single bit of what you put on the page, but it's not your job to guarantee they do that.

The reader is not your problem. How they consume the text and where they end up after reading it don't really need to be any of your business.

Games can't do this. A game needs to be challenging, but beatable. If your player can't get past level 2, they are not reaching level 3. They can't just shrug and turn the page if they don't get something. If level 4 is too hard, levels 5-10 are deleted from the player's experience.

And a fair-play mystery is a game between the writer and the reader. The game comes with an implicit contract: "I am going to give you the tools you need to win, and if you fail, it will be your fault." The reader is trying to rise to the challenge presented to them by the author, and figure out what's happening before the story tells it to them.

Three things can happen at the end of a mystery:

One is something that can happen with any piece of writing- the reader is bored. You did not make a difficult game. You presented them with a tic-tac-toe board with the center open. They wanted to be challenged, they wanted to uncover a story from beneath the facade of another story, and you just left the story lying out in the open where they didn't have to do anything.
Another is the outcome you're hoping for: they got it! They had a good time uncovering the clues and connecting the dots and solved the thing just in the nick of time. Maybe they didn't put everything together- maybe, in fact, they were on the wrong track the entire time, and had it all wrong- but when the reveal hit them, they went "Ohhhh! I messed up! You got me with this that and the other trick, but if I'd been thinking, I could've noticed this stuff beforehand!" They can delight either in their own victory, or in their opponent's cleverness.
But the third outcome is a failure mode that's both kind of unique to mystery writing, and the most common way for a mystery to fail. Your mystery was not just too hard, but not solvable. Until the final reveal, you didn't actually provide enough information to distinguish the truth from the many, many other possibilities. "How was I supposed to know that? You cheated!" The player was out of the game, and they had no idea until it was too late.

And this third outcome is what makes mystery writing such a tightrope walk! You fall off to the left, and it's into the pit of boredom. You fall off to the right, and it's into the abyss of incomprehension. You need to calibrate the information you present such that it's neither too obvious nor too obtuse.

Which! Is!! Literally impossible!!!

The reader isn't one specific guy! The reader is any number of people, as many as your writing can reach! And some of those people are eagle-eyed overanalyzers with insane attention to detail, and some of those people are... kinda dumb! What might be way too obvious a hint to one reader might be the only thing that gets the point through another reader's thick skull! What might be a clever puzzle to one reader might be some impenetrable Sherlock Holmes horseshit to another! There is no objectively correct level of difficulty you can try to hew to!

Nonetheless, you can't just write off the reader entirely. You could write for just the one specific reader that's completely on your wavelength, but then most people who read your story are going to think it sucks complete horse puckus. That's no good. One mystery won't do the trick. It is impossible to write a mystery. No, you need to write your story like... four different times.

Four? Yeah four seems about right. What are the four stories?

The thing that happened. The explanation for the mystery. You always have to write this one, unless you're J.J. Abrams and want your story to be some nonsense that everyone hates. This is a story that needs to exist, in full self-consistent detail- and it's the story that you will not be presenting to the reader on the page. It can live in pure outline, really. Communicating the events to this story to the reader in a way that asks them to participate in uncovering it is the job of the next three.
The story of your detective figuring out the thing that happened. This, too, is going to be kind of not shown to the reader. You do need to have this story exist- the story of someone who arrives to the crime scene, learns the facts, and puts those facts together. It's important that you write this story so that at the end, when the character in question recounts the story, the reader can believe that it happened. If you skip this step, you can end up with a detective who "solves" everything by just telling the first story as they read it off the author's notes. This feels to the reader like cheating- the character in the world didn't have access to this information, not really! How'd they pick this explanation out of all the other ones? If you can't answer that question, the reader will dismiss your game as rigged.
The story of your reader figuring out the thing that happened. This is the thing a lot of new mystery writers totally leave out- yes, okay, maybe your detective character was able to put all these really subtle clues together, but the reader was left out of that process until the end, with nothing to signpost what details were important and which questions they needed to be asking. You want your reader to get it- defeating them in the game is not the objective, here. The most important part of mystery writing is getting your reader to ask the right questions, so that when you leave them clues, they have something to do with them besides passively catalogue them. Write out what you want the reader to be asking and when, so you can highlight those questions in the story and time the delivery of clues to answer them.
The story of your reader figuring out the thing that happened, even though they're really really really dumb, like so dumb you can't believe it. No, seriously. Whatever you wrote in step 3 will not be good enough. You will think you are setting up obvious hints no one could possibly miss, and you will be wrong. No matter how many clues you leave for your detective and how simple the deductions seem to you, I guarantee the reader is going to be exploring more of the possibility space than you did. They will make assumptions you never considered assuming, misunderstand logistics, project tropes onto characters you never meant to use, imagine the story in a totally different genre... it won't work like how you thought. Your story should provide ways for even the wildly lost reader to get back on track, paring down the possibility space over time so they can realize their mistakes before the moment of truth.

These things can be very difficult to balance. Umineko no Naku Koro ni, for instance, suffered somewhat in later installments for having to reiterate and re-explain its core concepts to readers who somehow failed to grok them in earlier episodes, in ways that slowed down the pacing for the reader who's been keeping up. Your keener readers, who've figured out the obvious stuff and have more specific questions now, are going to be a little impatient if you keep belaboring the really obvious points for your absolute dumbest reader who refuses to get it.

Still, you can't leave any of these groups hanging for too long- it's helpful to track the questions people might be asking and when, and seed hints and answers in increasing order of obviousness. Having a few small nonessential sub-mysteries around can help in the latter half, when the clever reader has already figured out all the important stuff and needs something to chew on while you're helping the average reader through it.

Sounds... tough! There's a lot to keep track of, and a lot at stake if you screw it up. You're trying to run a game for every possible player simultaneously, weaving multiple reader-stories into the same text that also needs to support an invisible story that's had all its details hammered out from the jump. Winging it doesn't work.

So you can either do all this work... or you can cheat and make a videogame. More on that later.