5 Things I’ve Learned About Writing Effective Mysteries
Since I love to read mysteries, it makes sense that I also love to write mysteries. When I write speculative fiction these days, my stories have an element of discovery and investigation. My characters uncover clues, wrestle with what they’ve learned to extract facts, and finally reach the conclusion around the climax of the plot.
You guys, mysteries are hard.
When I revise earlier drafts of mystery stories I’ve written, I realize how much I’ve learned since I started. Because there are so many different types of mystery, I don’t know if what I’ve learned about my process applies to anyone else’s. Still, for what they’re worth, here are my conclusions.
1. Be patient.
I’m the least patient person, like, ever. And when I’m writing, that lack of patience makes me so excited about the answers to questions I’ve raised for my characters that I resolve them right away. Who killed the victim? (two chapters later) It was character XYZ, and that raises this even bigger mystery!
Which is well and good (bigger mysteries are great, and more mysteries can be better than fewer). But I’ve started to learn that it’s okay to ask “Who killed the victim?” and… not say the answer until the end when things come together. I can still lead the protagonists to new clues that shed light on the current mystery and add a new one to the list–those bigger mysteries can keep coming. But I can keep teasing the reader with the original problem until the very last moment.
This helps me build a pattern of consistently rising tension instead of a stop-start plot. It lets me keep the reader curious and add new reasons to wonder instead of shutting down possibilities. Plus, very often, having the protagonists wrestle with most of the same clues gives the reader a chance to develop theories of what happened themselves. Which brings me to my next personal lesson…
2. It’s okay if the reader occasionally gets to be smarter than the characters.
If the reader “solves” the mystery before the characters but doesn’t get confirmation until the end of the book… that’s a good feeling (as long as the rest of the plot is still exciting–they need to stay actively engaged with new developments instead of waiting passively for the characters to catch up).
I should have understood this a long time ago, considering I had to learn it for my treasure hunts. When designing a puzzle, it’s easy to get sucked into the idea that “successful puzzle = a puzzle the audience can’t solve any way but the way you designed.”
And that’s true to some extent–you don’t want to set up a code your audience can read as easily as regular text or write a mystery where it’s obvious from the first page whodunnit every time.* But it’s also true that audiences don’t care about “the way it was designed.” They care about: am I engaged? Am I enjoying myself?
As my past self wrote, people like to feel smart more than they like feeling puzzled or unintelligent. It’s okay if, in an otherwise engaging story, the reader occasionally sees what’s coming before the characters do. It’s very much not okay to design the mystery with the sole purpose of tricking the reader, because any fool can design an unsolvable puzzle if they go trivial and boring enough.
3. The antagonist has to matter.
One of the ways I used to try to fool the reader was to make the “bad guy”–the murderer/kidnapper/secret evil wizard/etc.–a minor character from the first few pages of the story. Surprise! It was the janitor all along! What’s that? You met the janitor for approximately a page and a half when she yelled at the protagonist for littering? And then the protagonist didn’t think about her, and she only ever appeared in the background until the big reveal?
Those kinds of stories can certainly work. I’ve read plenty of published mysteries where they do. But, I’ve learned, as a writer, they’re not for me.
That’s in part based on what I like as a reader: when I read Sherlock Holmes vs. Jack the Ripper pastiches, for example, I could care less when Saucy Jack turns out to be an OC (Original Character). I’m much more interested when the world’s most famous serial murderer is someone I already care about.
Do I “choose” whodunnit by imagining whose betrayal would hurt the protagonist the most? Well… no. But I’ve found that if I make the antagonist someone who doesn’t matter personally to the main character, the plot doesn’t come together in the way I love.
4. Read/watch lots, but make sure to unlearn shitty patterns
To write an interesting mystery, you have to know what’s out there. I like mysteries enough that I’m pretty good at guessing the final reveal based on the tropes of the story. And that’s okay–really. I’m never going to write a mystery that nobody will ever guess the end of, because that’s impossible. The point is to make the story a good story even if you already know the ending. Still, subverting rather than relying on tropes can mean the difference between engaging and boring the reader.
Plus, some tropes are, well, toxic. Gender-based violence is unfortunately a thing in real life, so it’s not surprising to see fiction reflect reality, especially in the hands of female writers who identify with the victims and survivors of that violence.
Still, I can’t count the number of thrillers I’ve read/watched that start with young, white women’s dead bodies. I can’t count the number of chapters I’ve read from the psychopathic male killer’s twisted perspective, inserted between the sympathetic characters’ viewpoints to increase the tension. Or, for that matter, the number of male protagonists I’ve encountered who also consider these dead women to be motivation, not people.
Before I use tropes like those, I have to carefully consider what they’re saying, and whether I’m using them because “that’s how these stories go” or because I have a good reason to do so.
5. Characters reasoning isn’t boring, but learning to write it well can be tough.
When I first started writing mysteries, I leaned heavily on the TV-genius-detective process of having my protagonist come across a new, random idea as part of the story, repeat that idea while staring into the middle distance, and then Eureka!ing into a solution. You know the drill:
“But I just can’t put together how the killer could have poisoned her when she was having tea in the Jacuzzi.”
“I’ve never seen you this stumped, Holmes. This case is one tough cookie.”
“Cookie…”
“Holmes?”
“Cookie… cake… cake of soap… I’ve been so careless! Quick, if we don’t hurry, there will be a third murder!”
Etc.
Upon reviewing a draft full of scenes like these, my agent gently suggested that perhaps my intended-to-be-intelligent characters might feel more intelligent if, you know, they actually thought about stuff and showed the reader how they arrived at their conclusions. (My phrasing here, not hers.)
This is, of course, way tougher to write. The hard part of a mystery isn’t making your character sound believable as they put the clues together; it’s making it feel believable that they don’t do so until the right point in the story. Obviously, having the investigator require an arbitrary cue to put the pieces together makes that part easy.
But, slowly, I’ve learned how to write my protagonists’ deductive reasoning in an engaging, believable way. In general, I’m assisted by the fact that I get to control when they receive each clue. If they have a clue, I try to let them squeeze the answers out of it–no giving them all the information and forcing them to run in place until the plot is ready for movement. I let them work at the incomplete clues they already have and consider the questions that arise from those. And, occasionally, I let them explore incorrect conclusions that make logical sense with the information they have–but which then get upended when their assumptions lead them to a new piece of contradictory evidence.
* Okay, yes, you could make either of these interesting in a different context–a comedy mystery that’s more about character relationships; a puzzle where the nature of the code rather than the content of the message is what helps you get to the next step; etc., but I think you understand the more common, general cases I’m talking about here.