A plot is a chain of events where each one makes the next one necessary. That is the whole definition, and it is stricter than it sounds.
If you can remove a chapter and the following chapter still works unchanged, that chapter was not part of the plot. It was something that happened.
Most sagging middles are made of things that happened.
What plotting actually has to produce
Not a document. A plan you can still use on day ninety, when the excitement is gone and you have to write anyway.
That is a different requirement from the one most plotting advice optimises for. Elaborate frameworks feel productive in week one and get abandoned by chapter four, because they were built to answer "what could happen" when the question you need answered later is "what do I write today."
A working plot plan tells you three things on any given morning. What scene comes next, what has to change in it, and why the reader cannot skip it.
If your plan cannot answer those, more detail will not save it. That is usually a sign the plan is describing events rather than consequences.
Five ways to plot, and what each is for
| Method | Time | Best when | Fails when |
|---|---|---|---|
| Beat sheet | A day | Genre plots with expected pacing | Character-led work it bullies into false beats |
| Chapter list | An afternoon | Most novels, the safe default | You need scene-level granularity |
| Snowflake | Two to three days | You think top-down and want completeness | It becomes a way to avoid drafting |
| Working backward | A day | You know the ending, not the route | You genuinely do not know the ending |
| Index cards | Ongoing | Structure keeps refusing to settle | You need a fixed plan to commit to |
The last column is the one to read. Every method works, and picking one whose failure mode matches your temperament is how people end up fighting their own plan.
If you improvise well, a chapter list is enough. If you stall without structure, a beat sheet earns its rigidity. How to outline a book goes deeper on the outlining side, and the 24 chapter outline assigns a specific job to each of twenty-four chapters if you would rather have the count decided for you.
Working backward is underrated
Start at the ending and ask what had to be true one step earlier. Then again. Then again.
This produces causally linked events by construction, which is the thing most plots are missing. Forward plotting tends to generate a sequence, and sequences are what leave you with a middle full of incident and no pressure.
It also surfaces your setup requirements for free. Every backward step tells you what has to be planted earlier for the step to land, so foreshadowing stops being a separate pass you do later and worry about.
The obvious limitation is that you need an ending. If you do not have one, a provisional one still works better than none, and discovering it was wrong is cheap at the outline stage.
The midpoint carries the book
One structural idea matters more than the rest, and it is the one most plans skip.
Around the middle, your protagonist's understanding of the problem has to change. Not the stakes and not the difficulty. What they think is actually going on.
Before it they are solving the wrong problem. After it they are solving the right one, and the first half retroactively becomes necessary error rather than throat-clearing.
Without that reversal the second half is the first half repeated louder, which is precisely what a sagging middle is. If you write one sentence for what your protagonist thinks the problem is at the start and one for the end, and those sentences are similar, the plan has no engine in it yet.
A worked example from one of my own worlds. Score runs on an opaque Contributor Score that decides whether you are allowed to exist in society, and the obvious version of the story is someone trying to raise a number.
That is the wrong problem, and the midpoint is where they find out. The high-paying Special Tasks that fix the number also take memory, and after the sixth they erase you entirely. Everything done in the first half to climb now reads as someone cheerfully digging.
The want stayed identical. What they understood it to cost did not, and that is the whole job of a midpoint.
A chapter template
CH [n]: [working title]
POV:
Opens with: (the situation as it stands)
What changes: (required, one sentence)
Cost: (what it takes from someone)
Ends on: (the state that makes the next chapter necessary)
Two fields carry the weight. "What changes" and "Ends on".
A chapter with no change is a scene collection. A chapter that does not end on something requiring the next one is where readers stop, and they will not write to tell you which chapter it was.
Fill this in for twenty-five chapters and pacing problems become visible on one screen. Six consecutive chapters whose "what changes" line reads vague is a diagnosis you can act on, and it is nearly invisible from inside the prose.
Keep the whole thing under about 5 percent of your target word count. For an 80,000-word novel that is 4,000 words of plan. Past that, extra planning starts reducing the odds of finishing rather than improving them.
Three ways plots break
Events without consequence. Things occur, the situation does not change. Test it by asking what is different afterward, and if the answer is "the reader knows more" it is exposition wearing a scene costume.
A protagonist who reacts and never chooses. If the plot happens to your character, tension leaks out. Readers engage with decisions, particularly costly ones, and a character who only responds is a passenger.
The problem solved by something the reader did not know about. New information at the climax reads as cheating even when it was set up badly rather than not at all. Everything needed to solve the book should be visible by the two-thirds mark.
That third one is really a plant-and-payoff problem, and working backward mostly prevents it.
The fantasy-specific traps
Speculative fiction adds failure modes that realist fiction does not have, and they are worth naming because they break plots quietly.
Magic that dissolves the problem. If the magic can solve the central conflict cheaply, the plot has to keep inventing reasons it does not, and readers feel the author's hand. The fix is upstream in the system rather than in the plot: cost, access limits, consequence. Magic systems covers building one that constrains a story rather than rescuing it.
World-ending stakes that flatten everything. Apocalypse is the easiest stake to reach for and the hardest to feel, because scale and weight move in opposite directions past a certain point. The fix is to make the large threat arrive through a small door: one relationship, one obligation, one thing the protagonist can still hold.
Prophecy that removes uncertainty. A prophecy the reader believes tells them the ending. It works when the prophecy is true in a way nobody predicted, or when the cost of fulfilling it is the real question. Crafting fantasy prophecies covers making them generate tension rather than spend it.
Worldbuilding that competes with the plot. A world can be interesting enough that the writer keeps touring it. Every excursion that does not change the situation is a chapter the reader skims.
Plot is not story
A perfectly plotted novel can still be inert, and this catches a lot of writers who did the structural work properly.
Plot is the sequence of causally linked events. Whether the reader cares comes from a character who wants something badly enough to be damaged by not getting it, which is a separate build. Creating believable characters covers that side.
What the book is arguing is a third thing again, and theme and message covers making that argument through choices rather than dialogue. Tension is a fourth, and it is more mechanical than most expect, which creating suspense in storytelling gets into.
Plot is the skeleton. It holds everything up and nobody falls in love with it.
Where plots survive and drafts do not
Here is the uncomfortable part. Plotting is the solvable problem, and solving it does not get the book written.
Drafts stall around 30,000 words with the outline intact, the midpoint planned, and the writer simply not opening the file. The plan was never the obstacle, which is why better plans do not fix it.
Go the Distance is built for that stretch specifically, taking a writer from finishing short pieces to a complete first draft at book length. It is organised around the two places drafts reliably die rather than around craft lessons.
Straight about the fit. If you have never finished a short story, start there instead. If you already finish novels consistently, the problem it solves is not one you have.
Start here
Write your ending in one sentence. Provisional is fine.
Work backward five steps, asking what had to be true immediately before each one. You now have the spine of a plot with causality built in rather than bolted on.
Then fill the chapter template for the first five chapters only. Not the whole book, because the plan will change and planning chapter twenty now is speculation.
Then write chapter one. The rest of the plan improves faster from drafting than from more planning.
Frequently Asked Questions
What is the best way to plot a novel?
Working backward from the ending produces the strongest plots, because asking what had to be true immediately before each event builds causal links by construction rather than leaving you with a sequence of things that merely happen. Pair it with a one-line-per-chapter list stating what changes in each chapter, and keep the whole plan under about 5 percent of your target word count.
What are the 5 steps of plot?
Exposition, rising action, climax, falling action and resolution. Rising action is by far the longest and the only genuinely difficult one, because it has to be escalating failed attempts that differ from each other rather than the same obstacle repeated with new scenery.
What are the 7 basic plot types?
Christopher Booker's seven are overcoming the monster, rags to riches, the quest, voyage and return, comedy, tragedy, and rebirth. They describe shapes stories tend to fall into rather than instructions for building one, so they are more useful for diagnosing what kind of story you are already writing than for generating a plot from scratch.
How detailed should a plot outline be?
Detailed enough to answer "what do I write next" on a morning when you have no ideas, and no more. One line per chapter stating what changes is enough for most novels. Once an outline exceeds roughly 5 percent of the target word count, added detail tends to reduce the odds of finishing rather than improve them.
Do I need to plot before writing?
No, and many finished novels were improvised. What you cannot skip is a destination, because the aimless middle is where unplanned books stall. A single page with an ending on it counts, so the real question is how little planning you can get away with rather than whether to plan.
Why does my plot feel boring in the middle?
Almost always a missing midpoint reversal. If your protagonist understands the problem the same way at the 60 percent mark as they did on page one, the second half is repeating the first half rather than escalating it. Decide what they get wrong early and what forces them to see it differently, and the middle usually fixes itself.
