How to Write a Mystery in Any Genre

by James Scott Bell

There’s an old joke about a guy walking into a bar with a squirrel in a cage. The bartender says, “What’s that squirrel doing here?” And the guy says, “Thinking about his next mystery.” The bartender asks what he means, and the guy says, “My squirrel is a mystery writer.”

“Come on!” the bartender says. “How can a squirrel write mysteries?”

“Easy,” the guy says. “He comes up with the ending, then works his way back.”

That squirrel was onto one way to write a mystery: Know the ending, the who of the whodunit, before you start writing. Some of the most successful mystery writers of all time—e.g., Dame Agatha, Erle Stanley Gardner—did it this way.

Their formula was simple. You have a dead body and several possible suspects, each with a motive and an alibi. The killer, when revealed, is a surprise.

That’s always the fun of a Perry Mason. In the classic TV show, the redoubtable Raymond Burr would be grilling a witness on cross-examination so incisively that someone out in the gallery would be forced to stand up and say, “Yes! I killed her! But she was going to force me to give up everything I worked for!”

This trope was hilariously sent up in Woody Allen’s Bananas:

But this is not the only way to do it. Other writers “pants” their way forward, not knowing until the end who the killer is going to be. They extol this method by saying, “If I can’t guess the killer until the end, then readers surely won’t be able to!”

That’s a pretty good argument, though it may entail substantial rewriting and rewiring the plot.

A rejoinder of the plotters is this: If I work out the motive and method first, I can design a whole web of red herrings to throw readers off the scent.

James N. Frey, in How to Write a Damn Good Mystery, is of the latter type. He advises picking a killer, then writing a lengthy biography to explore and justify the murder.

Personally, I don’t like to do lengthy character bios. I find it closes me in before I really get into the story. I like a little living and breathing space for my cast.

I do, however, want to know a few key things about my main characters:

  • Looks
  • Dominant Impression (a Dwight Swain advisory, which means a noun of vocation and an adjective of manner)
  • Timeline of Key Events. I identify the year of birth and go forward to other important years: first day of school, first job, first love. I always like to ask what happened to this character at the age of 16, which seems a pivotal year in everyone’s life.

Another thing the timeline gives me is a basis for cultural markers. I like to know what music, movies, and TV shows were popular in a given year. A few will pop out that seem right for the character.

When it comes to the villain, I have to come up with the most important thing: the motive. I want to have a “hidden” motive that is revealed near the end.

That’s when I write “the speech.”

We’ve seen this in many classic mysteries. The sleuth gathers all the suspects together in a room and starts explaining the clues. Hammett does this in The Thin Man, and Gardner in his courtroom scenes.

It can also be done one-on-one, as in The Maltese Falcon.

In the speech, my hero explains the whole setup, the red herrings, the clues that lead him to identify the killer. (Note: this speech is not intended for the final product, though I may use some of it. It’s a brainstorming exercise above all else.) I work on the speech over several days, sleep on it, add layers to it. This enables me to set up the “game” from the start, to know the hidden moves made by the villain “off-stage.”

Now, I’m not a pure mystery writer. I walk down the thriller street. But I believe all good fiction has a mystery to it, a question in the readers’ minds: What is going on here? Why are these things happening to the character? Why is the character acting this way? This is essential for any genre, from romance to thrillers to literary.

Because the great driver of fiction is a reader turning the pages to find out what happens next, and why. Otherwise, the story becomes predictable. And predictability is boring.

What about you? Do you agree that a mystery element, as defined above, is essential to good fiction? And when it comes to mysteries and character secrets, are you like that squirrel who knows the ending up front? Or do you like to pants-and-wait?

26 thoughts on “How to Write a Mystery in Any Genre

  1. Though I never thought about it in those terms, I agree there’s mystery in every book regardless of genre. Otherwise the reader has no reason to pick it up and read it. I have a historical fiction series to be written that is not mystery genre, but definitely one of the key issues that chases the protag throughout those books is getting answers for a critical life event that occurred (i.e. justice). It’s what drives him.

    As to writing a mystery, I can’t wait to complete my current project (my first mystery) so I know what my method is. LOL! Technically I know who the killer is, & how they killed the victim, and one of the reasons they killed the victim, but I need to dig deeper into causal reasons & develop those other suspects.

    Writing this first mystery is going to be really messy, but hopefully the lessons learned when I’m finished will make the next one a bit easier.

  2. Is a mystery element essential to good fiction?

    I think so. It’s almost automatic that a story will create unresolved issues during the writing, that can be wrapped up in a summation.

    Are you one who knows the ending? Or do you pants-and-wait?

    It varies from book to book. Generally, I pants:
    Mouth of the Lion is a WWII thriller with just one totally evil suspect (guess who), so the mystery was obviously never whodunit, but whydunit and howdunit, with elements like: Who was Der Failure’s mystery grandfather? Why the Holocaust? Why attack Russia? I had no idea what the answers were until I spent months researching the available information and analyzed AH.
    ❦ Writing Sail Away on My Silver Dream led me towards only one ending. The mystery element is why Sharon believes she once had a little sister.
    Tenirax I rejected a reader suggestion in favor of one wherein Bishop Filippo gets his revenge for Tenirax’s heretical notions and misdeeds. Will the Inquisition get Tenirax? That is left open.

  3. I plan all my key milestones in advance, including the ending. By having a roadmap to refer to, I can let my characters run free and still keep them on track and moving forward. It also prevents writing myself into a corner. If my characters do something unexpected, it’s a lot easier to figure out why with the skeleton of the story in place.

    • Your philosophy tracks with mine, Sue. I love doing my “signpost scenes” then being creative in between. I do scene cards for a lot of these transitions, knowing I can be creative with them when I get there.

  4. I think good writing involves both your rational and emotional selves. If you work out the ending first, you know the direction your story will take. Then your emotional, creative self is free to run wild around the foundation you’ve built, producing a livelier, more involving adventure.

    • That’s always the discussion, isn’t it, Mike? Left brain/Right brain. How much Left before you let the Right go? Vs. let the Right go wild before you have the Left figure out what it means. (“Jump off a cliff and grow wings on the way down.” Bradbury.)

      Personally, I let the Right run free as I build my foundation in the pre-writing stages. There’s a lot of rubble, but the stones I select are fitted to my purpose, so when I actually start writing I’ve got a clear view.

  5. Interesting post, JIm. I agree that all fiction – for it to be “good” (keep readers turning pages) – needs suspense, surprises, or mystery.

    And, I’m a squirrel on a squirrel cage in more ways than one. I do want to know the ending before I start writing, although it may change during the writing. I use a detailed outline where I can play with changes (set the boundaries) before I let the characters out to play.

  6. Terrific rundown of the elements of a mystery and how to create them. I second Steve in all fiction needing an element of mystery in it. All of my novels have had it, several also including a crime element, though my library mystery is my first *actual* mystery novel, in the full on genre sense. I aim to be like the squirrel that begins at the end.

    However, with that first novel, I did end up pantsing more than I would have liked, simply because my outline, despite a lot of prep, didn’t really contain a well realized mystery. In the course of rewriting and recasting the mystery, I’ve developed several principles of my own, one of which is having a chain of logical actions that both the murderer and the sleuth follow, as well as any character who hides a secret. “If this then that, followed by that, etc.” Logic internal to the character’s own motivations and subsequent actions.

    The challenge I encountered in the process of this is the tendency to over complicate the mystery, generating too many elements. In one case, I added too many things to a sub-mystery that ties into the main one. Simplify, even when I’m weaving a twisty web of clues, seems the best approach.

    Thanks for today’s post. It was right up my alley 🙂 Have a great Sunday.

    • I’ve developed several principles of my own, one of which is having a chain of logical actions that both the murderer and the sleuth follow, as well as any character who hides a secret.

      Dale, Erle Stanley Gardner had the exact same idea. He called it “the murder’s ladder.” I explain it in How to Write Pulp Fiction.

    • A recent mystery I read was ridiculously complicated. Very old school with multiple people being arrested then cleared of the murder but not of obstructing justice and messing with the crime scene. It was exhausting, and I stopped caring. Too many clever twists for the sake of being clever twists no longer works.

  7. Fascinating. I agree that all fiction needs some mysterious element to make it work.

    I’ve taken a different approach in each of my three novels.

    In the first book, The Watch on the Fencepost, I had a group of possibilities for the killer, and I decided on one about halfway through writing it. I didn’t know how the climactic scene would work out until I was writing it, and something happened in it that surprised even me.

    In the second book, Dead Man’s Watch, I knew who the killer was from the start, but didn’t know how to work out the climax until close to the end.

    In the third book, Time After Tyme, I knew how I wanted the climactic scene to unfold, but didn’t know the identify of the killer until more than halfway through.

    I think that makes me a hybrid plantser. But you’re right about having to rewire and rewrite. It takes me a long time to write a book, and maybe I can move closer to plotting the whole thing up front as I learn more. But in my case, I think leaving things open-ended gives me more opportunity to explore as I go from one scene to the next.

    • Thanks for the explication of your varied approaches, Kay. That’s really how it’s been for me over the whole of my career. I’ve tried things every which way, found ways I liked, then tweaked them. I’ve always learned something new with each book, and that’s kind of exciting. I love the sense of growth in the craft.

  8. Wonderful breakdown of various methods, Jim.

    I totally agree that every genre (including nonfiction) requires an element of mystery or at least uncertainty. That uncertainty is the secret sauce that keeps the reader’s interest.

    Like J and Kay, each book is different. In INSTRUMENT OF THE DEVIL and STALKING MIDAS, the villain is known from page 1. In EYES IN THE SKY, there’s an apparent villain masking the real villain lurking in the shadows. In FLIGHT TO FOREVER, a character is driven to illegal actions by circumstances outside his control so he’s not a typical bad guy.

    Sometimes I know the ending; other times, I make it up as I go along. As a pantser, from time to time, I have to back out of a dead end alley but, even then, there are usually elements that stay in the final draft.

  9. I’ve gone from total pantster to more of a planster, but I’m too lazy to work more than 3 chapters/scenes ahead. Given genre fiction has its reader expectations, I know a lot of the scenes (or kind of scenes) that I’ll have to hit.
    My second Pine Hills Police book, Hidden Fire came about simply because I wanted to know more about what had gone on with Randy and Sarah. Nobody told me in a romance 'series', the same h/h weren't allowed to be h/h in a second book, so I wrote it.
    All I knew when I started was that Randy had come back from a 3 week task force assignment, and Sarah was having second thoughts about their relationship. The book opens with them at a restaurant (neutral territory) and Randy has one thing on his mind. About the only thing that would pull him away from dinner would be a call out for a homicide. I didn't know who was dead, who killed him, or why at this point. But it got me writing.

  10. I always consider the killer/bad guy monologue reveal to be a bad one if the reader thinks in their head, “And I’d have gotten away with it, too, if not for you meddling kids/cops/PI/old biddy.” The SCOOBY DOO cliche curse is real.

    And, yes, even romances are mysteries. Each scene must move some element of the mystery forward and set up the next events for the reader, or the book is a bore.

  11. Pants and wait for me…

    When I started one WIP, I didn’t know why Annie is afraid every day of her life. A soccer Mom with 4 teenagers and a happy marriage?

    But, as I walked in her footsteps, she told me.

    However, I did have in mind the ending I wanted BEFORE she spilled the beans. And the cool thing is, I didn’t have to change it much. 🙂

Comments are closed.