How I Plan a Mystery Without Spoiling It for Myself
There is a delicious irony at the heart of writing mystery fiction: the author must know exactly whodunnit, why, when, and with what blunt instrument — yet somehow contrive to feel the thrill of not knowing while the story unfolds. Get the balance wrong and the prose goes flat, like a detective who has already glanced at the back of the box before starting the puzzle. Get it right, and you find yourself genuinely surprised by your own characters, typing at speed because you simply cannot wait to see what they do next.
I have tried various planning methods on the spectrum. At one extreme, the full architect's blueprint: a spreadsheet with every clue, every red herring, every alibi cross-referenced to the chapter in which it appears. Thorough, logical, and — for me, at least — utterly soul-destroying to write from. By chapter three I felt like a postman delivering letters I had already read. At the other extreme, the pure seat-of-the-pants approach: start with a body and a vague sense of grievance, then follow the characters wherever they wander. Thrilling for about forty pages, after which I found myself with six suspects, no coherent motive, and a subplot involving a missing cat that had somehow become central to the plot. Neither approach, on its own, serves the story.
What I have landed on — through a fair amount of trial, error, and the occasional dramatic deletion of an entire draft — is something I think of as planning the frame without painting the picture.
The first thing I settle is the ending. Not the solution in its full forensic detail, but the emotional truth of it: who is guilty, what drove them to it, and what it costs everyone — including the murderer — when the truth finally surfaces. In a cosy mystery, the reveal is never really about the mechanics of the crime. It is about human nature, about the cracks that form in communities and families when greed or fear or wounded pride is left to fester long enough. Once I know that emotional core, the ending has weight before I have written a single word of chapter one. Everything else I plan can shift and stretch; that does not.
Next, I sketch my suspects — not as names on an alibi chart, but as people with histories that existed long before the victim arrived in their lives. I ask each of them, in what I can only describe as a slightly odd internal conversation, what they most want to hide and why. Nine times out of ten, the answer surprises me, and that surprise is precious. If a character's secret catches me off guard in the planning stage, it will catch the reader off guard on the page, which is precisely where I want it. Paula Langford's cases work, I think, because the people she encounters in Brightcombe have whole lives that ripple out beyond the pages they appear on — and those lives only became visible to me when I stopped trying to control them and simply listened.
The middle of the book I plan only in loose chapters, more like a series of questions than a sequence of events. Who does my sleuth speak to next, and what will that conversation accidentally reveal? Where does the investigation need to go wrong before it can go right? Which piece of information is hiding in plain sight, waiting for someone — reader and detective alike — to notice it at exactly the right moment? I write these questions down rather than the answers, because the moment I answer them too precisely I have, in effect, already written the chapter in my head, and the actual writing becomes transcription rather than discovery.
Red herrings deserve particular care under this system. I plant them not as deliberate deceptions but as truths about the wrong person — genuine secrets that happen to look, from a certain angle, like guilt. When I write a scene in which a suspect behaves suspiciously, I have to believe in their reasons for behaving that way even though I know they did not commit the murder. If I do not believe it, the reader will not believe it, and the whole careful architecture of misdirection collapses. The trick is to give every suspect a reason to lie that has nothing to do with the crime at the centre of the story. People in fiction, as in life, are remarkably good at being accidentally incriminating.
The element of planning I have learned to leave almost entirely to instinct is the clue that solves everything. I know it exists — I know roughly what shape it takes — but I resist pinning it to the page until the story has grown around it naturally. In Murder Most Dramatic, one small, almost throwaway detail became the key to the whole puzzle; but it only became the key because of conversations I did not anticipate writing and a character who turned out to be far more observant than I had originally credited her. Had I locked that detail in place at the planning stage, I might have missed the far neater version the story eventually offered me.
The method is imperfect, as all writing methods are. There are drafts in my folders that confidently began with a plan and then quietly became something else entirely somewhere around chapter six, as if the story itself had grown bored of my intentions. But there is a particular kind of pleasure — quiet, a little smug, entirely private — in finishing a chapter and thinking: I did not know that was going to happen. That moment of surprise is, I suspect, the closest a mystery writer gets to the experience of the reader, and I guard it jealously. The frame holds the picture; the picture still gets to paint itself.
That, in the end, is the only plan worth keeping.