The short answer
Tag every chapter with one letter: M when the chapter serves the main plot, S when it serves the subplot you are measuring, and R when it is romance or rest. The checker turns those tags into four numbers: your main-plot share, your subplot share, your longest unbroken run of subplot chapters, and your main-plot share in each third of the book. A draft with roughly 55 to 65 percent main-plot chapters and no three-chapter subplot run is balanced. A draft under 45 percent main plot is main-starved, and any three-in-a-row subplot run makes a draft subplot-heavy even when the totals look acceptable.
Those bands are craft conventions, not rules of nature. They describe what working novelists usually find readable, and the tool applies them the same way every time so you can compare drafts fairly.
How the balance math works, with worked numbers
The tool performs five small calculations on your tag list. Understanding each one helps you trust the verdict and act on it.
First, it counts your chapters. The default tag list is M,M,S,M,R,M,S,M,M,R,S,M, which contains twelve tags, so the book in the example has twelve chapters. Anything that is not M, S, or R is ignored, which means stray commas and spaces never corrupt the count.
Second, it counts each kind of tag. In the default list, M appears seven times, S appears three times, and R appears two times. Seven plus three plus two equals twelve, so every chapter is accounted for exactly once.
Third, it converts each count into a share. Seven main-plot chapters out of twelve gives 7 divided by 12, which is 0.583, rounded to 58 percent. Three subplot chapters out of twelve gives 25 percent. Two rest chapters out of twelve gives 16.7 percent, rounded to 17 percent. Rounding is to the nearest whole number, so shares sometimes add to 99 or 101 instead of 100, and that small wobble is harmless.
Fourth, it finds the longest subplot run. It walks the list from chapter one to chapter twelve and tracks consecutive S tags. In the default list the S tags sit at chapters three, seven, and eleven, each isolated between other letters, so the longest run is one. A run resets whenever an M or an R appears, because either one returns the reader to the main story or gives a breather.
Fifth, it splits the book into thirds and measures each third separately. Twelve chapters divide evenly into three groups of four: chapters one through four form Act 1, chapters five through eight form Act 2, and chapters nine through twelve form Act 3. Act 1 contains M,M,S,M, which is three main chapters out of four, or 75 percent. Act 2 contains R,M,S,M, which is two main chapters out of four, or 50 percent. Act 3 contains M,R,S,M, which is also 50 percent. The whole book averages 58 percent main plot, but the table shows you that the opening carries more weight than the middle and the close.
The verdict follows from three thresholds. Main plot below 45 percent returns main-starved. Subplot at or above 35 percent, or any subplot run of three or more, returns subplot-heavy. Everything else returns balanced. The default list lands balanced: 58 percent main, 25 percent subplot, longest run one, weakest act 50 percent.
How to tag chapters so the numbers mean something
One tag per chapter is the whole method, but the tagging decision deserves care because every verdict downstream depends on it.
Ask which plot owns the chapter. A chapter belongs to the plot that sets its goal and pays off its closing beat. If the protagonist spends the chapter chasing the main quest and the chapter ends with the quest advanced, tag it M even when subplot characters appear. If the chapter exists to advance the subplot debt, romance, or mystery, and the main quest only waits in the background, tag it S.
Mixed chapters go to the plot that would be missed. Imagine deleting the subplot thread from the chapter entirely. If the chapter still stands because its main-plot work carries it, tag it M. If removing the subplot thread leaves an empty scene with no goal and no change, tag it S. Most chapters with a subplot flavor but a main-plot spine are M chapters, and tagging them honestly keeps the subplot share from inflating.
Keep the tags in reading order. The run detector and the act table both read left to right as the reader experiences the book, so chapter one comes first and the finale comes last. Reordering chapters later means re-entering the tags, which takes a minute and keeps every measurement honest.
Treat the R tag consistently. Decide once whether your romance thread is a measured subplot or a rest thread, then tag every romance chapter the same way for the whole book. The checker cannot tell the difference between a breather scene and a secondary plotline, so consistency is what makes the rest share readable.
Count short chapters the same as long ones. A two-page interlude and a twenty-page set piece each get one tag. Chapter-level balance measures attention and momentum, not word count, and a short subplot chapter still asks the reader to leave the main story.
What each verdict means
Balanced means the main plot leads from start to finish. Your main share sits at or above 45 percent, your subplot sits below 35 percent, and no subplot stretch runs three chapters. Readers following the main story get fed regularly, and subplot chapters read as welcome depth rather than detours. When you see this verdict, protect the shape as you revise instead of rebuilding it.
Subplot-heavy means the subplot occupies stretches long enough to feel like the main event. Either its total share reaches 35 percent or it strings three or more chapters together without a main-plot chapter between them. The main plot may still lead on totals, which is why this verdict surprises writers. Totals can pass while the reading experience fails, because three consecutive subplot chapters ask readers to shelve the main story for a long sitting.
Main-starved means the main plot holds fewer than 45 percent of chapters. Readers were promised one story and are receiving another. This verdict calls for structural repair rather than line editing: chapters need to change ownership, merge, or move before sentences matter.
Too few chapters is not a verdict on quality. With fewer than three tags the tool reports that it needs more data, because one chapter would swing every share by thirty points or more. Tag at least a full act before drawing conclusions.
Why three subplot chapters in a row is the flag
One subplot chapter reads as a turn toward depth. Two in a row reads as a subplot sequence with its own arc, which most novels contain and most readers enjoy. Three in a row reads as a change of protagonist, goal, or book, because the reader has spent three sittings inside a secondary concern while the main tension cools.
Established series sometimes break this pattern deliberately near the end of a volume, when a subplot climax needs room. That choice works when the writer knows the cost: the main plot must then re-earn attention with a strong chapter immediately after. The checker still flags the run, and it is right to flag it, because flagged does not mean forbidden. It means expensive, and you should spend that expense on purpose.
Romance and rest chapters break runs. An R between two S chapters resets the counter, because the reader stepped out of the subplot into something else. Main-plot chapters break runs the same way. Only unbroken S sequences count, which is why honest tagging of breather chapters matters so much.
When the tool reports a longest run of four or five, look at the middle of the run first. The chapter at the center is usually the one that can move off-page, merge with a neighbor, or convert to a main-plot chapter carrying a subplot beat inside it.
Act splits catch what totals hide
A book can average 58 percent main plot while its middle collapses. Suppose Act 1 runs 75 percent main, Act 2 runs 25 percent, and Act 3 runs 75 percent. The average is 58, yet readers will report that the middle dragged, because they experienced Act 2 as a 25-percent stretch, not as part of an average.
The thirds comparison exists to catch exactly that. Read the three act shares side by side and look for the weakest one. A gap of twenty points or more between your strongest and weakest act deserves a revision pass even under a balanced verdict. Put your next new main-plot scene in the weakest act rather than where writing feels easiest.
Endings deserve a special note. Act 3 carries the climax, and climaxes belong to the main plot in nearly every genre, so an Act 3 below 40 percent main plot usually means the subplot is resolving on the pages where the main story should peak. Openings mirror this: an Act 1 below 40 percent main plot teaches readers the wrong lesson about what the book is.
The rest share and how much breathing room you can afford
Rest chapters include romance beats treated as breathers, quiet interludes, travel chapters, and comic relief that serves neither plot directly. These chapters have real value: they vary pace, deepen character, and let tension reset before the next push. A book with zero rest often feels relentless in a way that exhausts rather than excites.
The working budget is roughly one rest chapter in four. Up to about 25 percent rest leaves the main plot intact in most drafts. Past 30 percent, rest chapters start competing with the main plot for the reader's sense of what the book is.
Illustrative hypothetical example one: the fantasy debt that ate the middle
Consider an illustrative hypothetical novel, invented purely to show the math. A twelve-chapter fantasy draft follows a courier carrying war orders, which is the main plot, plus a smuggler-debt subplot the writer loves. The writer tags the draft M,M,S,S,S,M,M,R,M,S,M,R.
Counting gives six main chapters out of ten? No, count carefully: the list holds twelve tags with M at positions one, two, six, seven, nine, and eleven, which is six main chapters, or 50 percent. S appears at positions three, four, five, and ten, which is four subplot chapters, or 33 percent. R appears twice, or 17 percent. The longest subplot run is three, at chapters three through five.
The verdict is subplot-heavy because of the run, even though 50 percent main plot clears the starvation line and 33 percent subplot sits just under the share line. The act table sharpens the picture: Act 1 holds chapters one through four at M,M,S,S, which is 50 percent main; Act 2 holds S,M,M,R at 50 percent main; Act 3 holds M,S,M,R at 50 percent main. Every act ties at half, and the opening four chapters contain a three-run split across the act boundary region that dominates the reader's first impression.
The repair is concrete. The writer converts chapter four, the middle of the run, into a main-plot chapter in which the courier carries the subplot creditor's message inside a main-quest journey. New tags read M,M,S,M,S,M,M,R,M,S,M,R: seven main out of twelve, or 58 percent, with the longest run cut to one. Nothing else in the draft changes, yet the verdict flips to balanced and the opening now alternates instead of stalling.
Illustrative hypothetical example two: the mystery cold case that behaved
Consider a second illustrative hypothetical novel, also invented to demonstrate healthy numbers. A ten-chapter mystery follows a detective on a current murder case, with a decades-old cold case as the subplot. The writer tags it M,S,M,M,R,M,S,M,M,R.
Count the tags from left to right: M appears at positions one, three, four, six, eight, and nine, which is six main chapters out of ten, or 60 percent. S appears twice, or 20 percent, and R appears twice, or 20 percent, and sixty plus twenty plus twenty equals one hundred.
The longest run is one, since the subplot chapters at positions two and seven stand alone. Thirds of ten chapters split four, three, and three: Act 1 covers chapters one through four at M,S,M,M, which is 75 percent main; Act 2 covers chapters five through seven at R,M,S, which is 33 percent main; Act 3 covers chapters eight through ten at M,M,R, which is 67 percent main. The verdict reads balanced on totals, but the act table waves a flag at Act 2, the weakest third.
This is the tool working as designed. The writer keeps the balanced verdict in perspective and drafts the next new scene, a witness interview advancing the current murder case, into Act 2. If that scene converts chapter five from rest to main, Act 2 rises to 67 percent and the middle stops sagging. The subplot chapters stay exactly where they are, because isolated subplot chapters surrounded by main plot are precisely what good balance looks like.
Edge cases and failure modes
Very short drafts produce jumpy numbers. In a five-chapter outline, moving one chapter shifts every share by twenty points. Use the checker on outlines to spot runs and obvious starvation, but wait until you hold at least ten chapters before treating small share differences as real problems.
Very long drafts hide local problems inside healthy totals. A forty-chapter novel can average 60 percent main plot while containing a six-chapter subplot bender in the middle. The longest-run detector catches the bender, but you should also run the checker on individual sections: tag chapters eleven through twenty separately and read that verdict on its own.
Two subplots sharing one S tag inflate the subplot share. If your book carries both a debt subplot and a sibling-rivalry subplot, tagging both as S measures their combined weight, which can read subplot-heavy when each thread alone is modest. Measure the heavier thread first, then measure the second thread separately, and treat the combined reading as a ceiling rather than a precise diagnosis.
Tagging by feeling instead of by chapter goal corrupts every number. Writers who love a subplot tend to tag ambiguous chapters for the main plot, hiding the imbalance they came to find. Apply the deletion test described above to each chapter and accept the result even when it stings. The tool is only as honest as its tags.
Prologues and epilogues need a decision before tagging. Either include them as chapters one and last, or exclude them entirely. Including a subplot prologue drops Act 1 artificially, while excluding it hides a real first impression. Pick the option that matches how readers will experience the book and stay with it across revisions.
Dual points of view do not change the method. Tag by plot ownership, not by narrator. A subplot viewpoint chapter is still S, and a main-plot chapter stays M regardless of whose head it occupies.
Genre-specific guidance
Mystery and crime drafts usually run the highest main-plot shares, because the investigation is the engine and readers track clues chapter by chapter. Aim for 60 to 70 percent main plot, keep subplot runs to one, and be strict about Act 3: the solution chapters must be M. A cold-case or personal subplot at 20 to 25 percent adds texture without obscuring the clue trail.
Thrillers follow the mystery pattern with even less patience for rest. Pace pressure means rest shares above 20 percent often read as stalling. Keep breathers short or folded into main-plot chapters, and treat any subplot run of two in the final third as a problem even though the formal flag sits at three.
Romance places the relationship at the center, which changes what counts as main. If the love story is the main plot, tag relationship chapters M and tag external threads like careers or family business as S. The same numeric bands then apply: roughly 55 to 65 percent of chapters advancing the relationship, no long external-plot runs. If instead the romance is the subplot inside an adventure or mystery, tag it S or R consistently and hold it under 35 percent.
Fantasy and science fiction face the largest subplot pressure, because invented worlds generate fascinating secondary threads: debts, guilds, histories, and side quests. The genre tolerates subplot shares near 30 percent when the threads feed the world the main quest traverses. It does not tolerate long runs, because readers of long books are quick to sense which thread the writer prefers. Keep runs to two at most, and verify that Act 2, where quest middles famously sag, holds at least half main plot.
Horror runs lean. Dread depends on sustained attention to the central threat, so main shares below 55 percent often dilute fear. Subplots about relationships or past trauma work best folded inside main-plot chapters rather than granted standalone chapters. A standalone subplot run of three in a horror draft almost always reads as a different story interrupting the frightening one.
Literary fiction gets the most flexibility, because some literary novels braid two equal threads on purpose. If your design intends a double narrative, the 45 percent starvation line may not apply to you. Still run the checker: knowing that your draft reads 50-50 with alternating chapters tells you the braid is executing as planned.
Repairing a subplot-heavy draft in order
Fix runs before shares. Splitting a three-chapter run changes the reading experience immediately, while trimming total share without breaking the run leaves the worst stretch intact. Find the longest run, pick its middle chapter, and convert, move, or merge that chapter first.
Then repair the weakest act. Open the act table, identify the third with the lowest main share, and place your next main-plot chapter there. One well-placed chapter can lift a weak act by twenty points or more in a twelve-chapter book.
Then trim totals if needed. Merge two subplot chapters into one by combining their revelations into a single stronger scene. Move a subplot beat off-page by reporting it in dialogue inside a main-plot chapter. Fold a rest chapter into its neighbor by attaching the breather scene to the end of a main-plot chapter. Each of these moves shifts one tag and moves every share two to eight points depending on book length.
Re-tag after every structural move. Revision changes chapter ownership constantly: a scene added here, a goal clarified there.
Questions writers ask once they see their numbers
Can the middle of the book carry more subplot than the opening? Yes, within limits. Readers grant the middle some wandering once the main story has earned their trust. An Act 2 near 50 percent main plot is normal and healthy. An Act 2 near 30 percent asks for repair even when the book-level verdict passes, because trust earned early is spent fast.
How many subplots can a novel carry? Most novels carry one major subplot well, two with discipline, and three only in long books where each is small. The checker measures one thread at a time, so run it once per subplot and add the shares mentally. Two threads at 20 percent each occupy 40 percent combined, which behaves like one heavy subplot and needs the same treatment.
Should tagging happen before drafting or after? After, for diagnosis, and before, for planning. Tag the finished draft to find the imbalance, then tag the planned revision outline to confirm the fix works on paper before rewriting scenes.
What if beta readers love the subplot more than the main story? That affection is data, not a verdict. Sometimes it means the subplot deserves promotion into a sequel or companion. Sometimes it means the main plot needs strengthening rather than the subplot needing cuts. And sometimes, rarely, it means the book you actually wrote has a different main plot than the one you planned, in which case re-tag with the threads swapped and see whether the numbers describe a better book.
Where to draft the chapters this tool prescribes
Once the checker points at the weakest act or the middle of a long run, the next step is writing the replacement chapter itself. Co-Writer at the /write page is the natural place for that drafting work, since it is built for producing and revising manuscript prose chapter by chapter. Bring the checker's prescription with you, such as converting chapter four into a main-plot journey that carries one subplot message inside it, and draft that chapter fresh rather than patching sentences. Then re-tag the revised list and confirm the new verdict before moving on.
A tagging session you can finish today
Set aside thirty minutes with your chapter list. Write one letter beside each chapter number, working in order and applying the deletion test wherever you hesitate. Enter the letters and your subplot name into the tool, read the verdict, and write down the longest run and the weakest act on a card. Fix the run first, then the act, then re-tag. Writers who repeat that loop once per revision pass keep their subplots seasoning the story instead of flooding it.