Before you fix a single scene, fix the shape. A polished scene in the wrong place is still in the wrong place.
Phase 3 · Revision & Professionalization · Week 23 of 32Welcome to Phase 3. You have a complete first draft and a shape report. The temptation now — the urge that will ambush you every morning you sit down to revise — is to start at page one, read a sentence, fix it, read the next sentence, fix that, and crawl forward through the manuscript correcting individual problems until you reach FADE OUT. That approach will produce a slightly more polished version of the same draft. It will not produce a better screenplay. Revision is not polishing. Revision is re-seeing — re-examining the draft at the level of its largest structural decisions and adjusting those before touching anything smaller. You don't fix dialogue until you've fixed scenes. You don't fix scenes until you've fixed sequences. You don't fix sequences until you've fixed the arc. This week is the arc. The macro pass examines the draft's three largest structures — story arc, character arc, and thematic argument — and produces a revised outline and change list that will guide every subsequent pass. Fix the shape first. Everything else follows.
The three macro questions. The macro pass asks three questions about the draft. Each question examines the screenplay at its highest level of organization — not "does this scene work?" but "does the thing this scene is part of work?" Your shape report from Week 22 gave you the raw data. This week, you process that data into structural decisions.
Question 1: Does the story arc hold? The story arc is the causal chain from inciting incident to climax — the "therefore/but" sequence you designed in Week 15 and drafted across Weeks 16–21. In the cold read, you tracked momentum (where the pull strengthened and weakened) and clarity (where the reader might lose track of what's happening). Now you map the arc as it actually exists in the draft — not as you planned it, but as you wrote it. Take your step outline from Week 15 and compare it, beat by beat, to the draft. Which beats survived intact? Which ones changed during drafting? Which ones never made it to the page? Which new beats appeared that weren't in the outline? The gap between the planned arc and the actual arc tells you where the draft diverged — and whether the divergences were improvements or drift.
A specific diagnostic: check the causality lock in the draft. In the outline, every beat connected via "therefore" or "but." In the draft, scenes may have drifted into "and then" territory — scenes that follow each other chronologically but aren't caused by the preceding scene. Mark every causality break. These are the first things the macro pass must fix, because a broken causal chain creates dead zones where the audience's forward momentum stalls.
Question 2: Does the character arc track? The character arc is the protagonist's journey from Want through Wound-distorted strategy through Need recognition through Line decision. You designed this arc in Week 7 and embedded it in the outline. In the draft, the arc may have evolved — characters change during drafting, as you discovered during the midpoint rewrite. The macro question isn't whether the arc matches the dossier. It's whether the arc is visible in the draft — whether a reader who knows nothing about your Week 7 dossier can track the protagonist's internal journey through behavior, decisions, and consequences.
A specific diagnostic: identify the four landmarks of the character arc in the draft. At what page does the Want first become active? At what page does the Wound first distort the protagonist's strategy in a visible, consequential way? At what page does the Need become visible (even partially)? At what page is the Line tested? If any landmark is missing or unclear — if you can't point to a specific page where it occurs — the character arc has a gap that the macro pass must fill.
Question 3: Does the theme land? The thematic argument is the question the plot tests through consequence. You defined it in Week 6 and designed every structural element to test it. In the draft, the theme may have shifted — your Week 22 shape report asked you to state "what this screenplay is actually about." If the actual theme matches the intended theme, the thematic architecture is intact. If they diverge, you face a choice: revise the draft to serve the original theme, or revise the theme statement to match the draft. Neither option is automatically correct. The question is which version — the planned theme or the emergent theme — produces a more interesting, more contested, more dramatically productive argument. Sometimes the draft discovers a better question than the one you started with.
A specific diagnostic: find the three scenes where the thematic argument is under most pressure — where the competing answers to the thematic question collide most directly. These are usually the midpoint, the devil's bargain (Level 4), and the climax. In each scene, can you articulate the protagonist's position on the thematic question and the antagonist's position? Is the antagonist's position genuinely competitive — does the audience feel the pull of the opposing argument? If the theme is operating, these scenes should feel like a debate conducted through action. If the theme is inert, they'll feel like plot events with no philosophical undercurrent.
The change list. After running the three macro diagnostics, you produce a change list — a document that categorizes every structural element of the draft into four columns: STAYS (working as written, protect during revision), MOVES (working as written but in the wrong structural position — needs relocation), CUTS (not working and not essential — remove to tighten), and MISSING (needed but absent — must be written or inserted). The change list is not a to-do list for this week. It's the master document for all of Phase 3. Weeks 24–28 will execute the changes. This week, you identify them.
The revised outline. Using the change list, produce a revised outline — an updated version of your Week 15 step outline that reflects the draft as it currently exists, plus the changes you've identified. The revised outline is the new map. Every revision decision in Weeks 24–28 will be checked against it. Scenes that survive the cut belong in the revised outline. Scenes that need relocation appear in their new positions. Scenes that need writing appear as [TO WRITE] markers. This is the document you'll work from for the rest of Phase 3.
Professional screenwriters develop personal systems for annotating their drafts during revision. The goal is to mark problems without fixing them — to separate diagnosis from surgery so that you can see the full pattern of issues before you start cutting. Here are the most common annotation methods, adapted for screenplay revision.
Color coding. If working on paper or in a PDF annotator, use four colors for four types of notes: red for structural problems (scene in wrong position, causality break, missing beat), blue for character problems (voice inconsistency, Wound not visible, motivation unclear), green for scenes that are working (protect during revision), and yellow for questions (uncertain whether something is a problem or a feature — needs further evaluation). A draft covered in one color has a single dominant problem type. A draft covered in all four colors has a distributed problem set that requires the pass-by-pass approach.
Margin codes. When working digitally in screenwriting software, use a shorthand code system in margin comments or inline notes. Professional shorthand varies, but a useful starter set for this curriculum:
The annotation system's value is in making the revision visible before you begin. A draft annotated with margin codes gives you a literal map of the work ahead — you can see at a glance which sections are heavily marked (need the most work) and which are clean (protect them). The macro pass this week should produce a densely annotated draft — not because the draft is terrible, but because the macro lens catches issues that line-level reading misses.
10-minute drill: Open your draft PDF (or print five pages from the middle section). Read the five pages with a pen in four colors (or four annotation types). Mark at least one note per page using the margin codes above. Focus on macro issues — [ARC], [CHAR], [THEME], [CUT], [MOVE] — not line-level problems. After ten minutes, review your annotations. What pattern emerges? Is one code type dominant? That's your draft's primary structural issue. Time yourself.
Assignment: Read 2 produced feature screenplays, chosen for structural elegance at the macro level.
Script Pair Brief: Select one screenplay with a story arc that feels seamlessly constructed — where every scene flows from the previous one, the character's journey is visible and earned, and the thematic argument is woven through the structure without ever being stated directly. Then select one screenplay that achieves structural complexity — multiple storylines, a non-linear timeline, or a protagonist whose arc is unconventional (circular, downward, or deliberately incomplete). The first teaches you what structural elegance looks like at the macro level — the invisible architecture that makes a story feel inevitable. The second teaches you that elegance doesn't require simplicity — complex structures can still feel unified if the macro shape is intentional.
Where to find scripts legally: IMSDb, The Script Lab, SimplyScripts, studio FYC releases, and public library systems.
Why this pairing: You're about to restructure your own draft. Reading two screenplays specifically for their macro architecture — not their scene craft or dialogue, but the shape of the whole — calibrates your sense of what a well-shaped screenplay feels like. The seamless screenplay shows you the standard. The complex screenplay shows you that the standard can be broken productively, as long as the macro shape serves the story's thematic argument.
Reading Lens (track these while reading):
1. Can you trace the story arc — the causal chain from inciting incident to climax — without gaps? Where is the chain tightest? Where does it loosen? 2. Map the character arc landmarks: Want activation, Wound distortion, Need visibility, Line test. At what pages do they occur? Are all four present and visible? 3. Identify the thematic argument. Where are the three scenes of maximum thematic pressure — the moments where the competing positions collide most directly? 4. If you had to write a change list for either screenplay — STAYS / MOVES / CUTS / MISSING — what would be on it? Or is the screenplay so structurally sound that the change list would be empty? 5. How does the screenplay's macro structure create the feeling of inevitability — the sense that the ending was the only possible ending? What's the structural mechanism that produces that feeling?
Journal Prompts:
1. For the seamless screenplay: reverse-engineer the macro pass. If this writer did what you're doing this week — evaluated story arc, character arc, and thematic argument — what would they find? Write a hypothetical shape report for the finished screenplay. Where would "momentum strong" and "momentum weak" fall? 2. For the complex screenplay: how does the unconventional structure serve the thematic argument? Is the complexity the point (the form mirrors the content), or is it a stylistic choice that could be simplified without thematic loss? 3. Compare the character arcs. Which protagonist undergoes a more visible transformation? Which arc is more surprising — and is surprise correlated with visibility or inversely related? 4. Find the thematic scene in each screenplay — the single scene where the theme is under maximum pressure. How does the scene deliver the pressure: through confrontation, through consequence, through the protagonist's internal processing, or through juxtaposition with another scene? 5. After reading these two screenplays, look at your own change list. Do either screenplay's structural strategies suggest a solution to one of your draft's macro problems? Name the specific technique and how you'd apply it.
Deliverable: Revised outline + change list.
Constraints: Produce two artifacts:
(a) Change list (2–3 typed pages). Organize every structural element of your draft into four columns: STAYS (what's working and must be protected), MOVES (what's working but needs relocation), CUTS (what should be removed and why), and MISSING (what needs to be written or inserted). Each entry should name the element, its current page number in the draft, and a one-sentence justification for the decision. The change list must address at least one finding from each of the three macro diagnostics: story arc (a causality fix), character arc (a landmark reinforcement), and thematic argument (a scene where the theme needs strengthening).
(b) Revised outline (4–6 typed pages). An updated step outline reflecting the draft as it currently exists, plus all changes from the change list. Scenes that STAY appear in their current positions. Scenes that MOVE appear in their new positions. Scenes that are CUT are removed. Scenes that are MISSING appear as [TO WRITE] markers with 2–3 sentences describing what they need to accomplish. The revised outline should include the four character-arc landmarks noted with their page positions. The causality lock should be re-verified: every beat connected via "therefore" or "but."
Quality bar: The change list must contain at least three CUTS (you cannot revise a first draft without cutting something — the first draft always has scenes that don't earn their place). It must contain at least two MISSING entries (a first draft always has gaps that the cold read revealed). The STAYS column should include the scenes identified in your shape report's emotional map as the draft's strongest moments — the surprises, the earned low points, the moments where the writing was most alive. Protecting these during revision is as important as fixing the weaknesses. The revised outline must be a complete, continuous arc from FADE IN to FADE OUT — no unresolved gaps between beats.
Estimated time: 8–12 hours (macro diagnostics: 3–4 hours; change list: 2–3 hours; revised outline: 3–4 hours; reading: remaining time).
Human Draft Reminder: You make the decisions. AI helps you diagnose, but every choice on the change list — what stays, what moves, what goes, what's missing — is yours. The revision reflects your judgment, not an algorithm's.
Welcome to Revision Specialist mode. From this week through Week 28, the Two Readers adopt narrowed roles specific to each revision pass. This week's macro pass uses these specific roles: Reader A asks "Does the structure hold?" — evaluating the story arc, the character-arc landmarks, and the causal chain. Reader B asks "Did I care at the end?" — evaluating whether the draft's emotional journey produces a felt experience worth having, regardless of structural soundness. The prompts are pass-specific. Don't reuse them for later passes — each week in Phase 3 gets its own narrowed prompts.
Add this week's entry to your Disagreement Log. The macro-pass divergence is the most consequential of Phase 3: Reader A evaluates the structural integrity of your revision plan, and Reader B evaluates its emotional cost. When they agree that a cut is justified, cut confidently. When Reader A says "cut it" and Reader B says "it's doing emotional work" — that's where your judgment is most needed. The question to ask: is the emotional work this scene does replaceable — could another scene, in a better structural position, accomplish the same emotional task? If yes, cut the scene and assign its emotional function to the replacement. If no — if the emotional work is specific to this scene and can't be transplanted — keep it and find a different structural solution. Log the disagreement. Note your reasoning. Execute the decision in subsequent passes.
Budget tier this week: Your chosen tier, now applied to revision decisions.
Top 3 production benefits of the macro pass: 1. Cuts reduce the production footprint — every scene you cut is a location, a setup, and a cast requirement removed from the shooting schedule. The macro pass is the most cost-effective revision exercise, because structural cuts eliminate entire line items rather than shaving minutes off existing ones. When you cut the newspaper subplot (three scenes, one recurring character, two unique locations), you've removed an actor's contract and two location fees. 2. Moves can consolidate locations — if you're moving David's introduction from page 52 to page 33, check whether the new scene can take place in a location already established by page 33. The move might eliminate a location rather than adding one. 3. [TO WRITE] scenes should be designed for existing locations — when you're creating new scenes to fill MISSING gaps, default to locations already in the screenplay. Every new scene written into an existing space costs nothing in production terms beyond the shooting time.
Cheaper equivalent: A [TO WRITE] marker that says "a scene demonstrating Nora's teaching competence — new location: elementary school where she guest-lectures" adds a new location to the budget. The same scene set in Nora's own classroom — already established, already dressed — costs nothing additional and is arguably stronger, because the audience sees her competence in the space she'll later lose.
Worth-it spend: The scenes in the STAYS column. Every scene you're protecting has survived the harshest diagnostic you'll perform on your screenplay. These are the scenes that work — the ones the shape report identified as emotionally alive, the surprises that the draft produced, the moments the audience will remember. They deserve the best production treatment: the right location, the right casting, the right shooting time. The rest of the screenplay supports them. They define the film.
Professional screenwriters know something amateurs don't: the hardest part of revision is cutting good material. Every screenplay has scenes that are well-written, well-performed on the page, emotionally effective — and structurally redundant. The scene works in isolation. But the screenplay doesn't need it, because another scene does the same job, or because the information the scene delivers is already conveyed elsewhere, or because the scene's position in the arc creates a momentum problem. Cutting a bad scene is easy. Cutting a good scene that's in the wrong place or doing redundant work is one of the most painful decisions a writer makes. The ability to make it — to kill a scene you love because the screenplay doesn't need it — is the dividing line between a writer who finishes good screenplays and a writer who finishes bloated ones.
Look at the CUTS column of your change list. Write about the scene you most regret cutting — the one you're most attached to, the one you're most tempted to save. Why are you attached to it? Is the attachment about the scene's quality (it's genuinely well-written and you're proud of it), its emotional content (it touches something personal), or its effort (you worked hard on it and cutting it feels like wasted work)? Each type of attachment creates a different revision temptation. Quality attachment is the most legitimate — but even a well-written scene must serve the arc. Effort attachment is the most dangerous — the sunk cost fallacy applied to screenwriting. Name the attachment. Then make the cut anyway. The scene can live in your notes. It doesn't have to live in the screenplay.
By the end of this week you should have:
• Read 2 produced screenplays with a focus on macro-level structural elegance
• Run the three macro diagnostics: story arc (causality check), character arc (four landmarks), and thematic argument (three pressure scenes)
• Produced a change list with four columns: STAYS, MOVES, CUTS, MISSING
• Produced a revised outline reflecting the draft plus all macro changes
• Annotated your draft with revision margin codes
• Completed the annotation drill (5 pages marked with macro codes, 10 minutes)
• Run the change list and revised outline through both Revision Specialist prompts
• Updated your Disagreement Log with this week's entry
Week 24: Rewrite Scene Pass (Pace + Redundancy). The macro shape is set. The change list is made. Next week, you drop one level — from arc to scene — and execute the first wave of revisions. The scene pass evaluates every scene in the draft for two criteria: does it earn its pace (is its length proportional to its dramatic importance?), and is it redundant (does another scene do the same work?). You'll produce a scene list with purpose tags, cut or merge at least five scenes, and begin implementing the CUTS and MOVES from your change list. The AI prompts narrow further: Reader A asks "Which scenes could be cut without losing story?" and Reader B asks "Where did I start skimming?"