Audit Record — The Label On The Back

The 2026 Official Playing Rules of the National Football League describe a helmet in one paragraph. It is seven sentences long. It fixes the colour of the chin strap, the diameter of the facemask to a sixteenth of an inch, where a manufacturer’s name may and may not appear, and where two labels are placed on the back of the shell.

It does not say what a helmet must withstand. Neither does any other passage in the document.

Nothing in this case was concealed and no rule was broken. The numbers that describe a helmet’s performance exist — they are simply written somewhere else, by bodies the playing rules never name, and they are different numbers produced by different machines. This record is the working behind the audit: what each document says, where each figure comes from, and — set out explicitly at the end — what this record does not claim.

The paragraph, in full

Rule 5, Section 4, Article 3, Item 1 of the 2026 Official Playing Rules:

Item 1. Helmet, Face Protectors. Helmet with all points of the chin strap (white only) fastened and facemask attached. Facemasks must not be more than ⅝-inch in diameter and must be made of rounded material; transparent materials are prohibited. Clear (transparent) plastic eye shields are optional. Tinted eye shields may be worn only after the League office is supplied with appropriate medical documentation and approval is subsequently granted. The League office has final approval. No visible identification of a manufacturer’s name or logo on the exterior of a helmet or on any attachment to a helmet is permitted unless provided for under a commercial arrangement between the League and manufacturer; in no event is identification of any helmet manufacturer permitted on the visible surface of a rear cervical pad. All helmets must carry a small NFL shield logo on the rear lower-left exterior, and an approved warning label on the rear lower-right exterior. Both labels will be supplied in quantity by the League office.

Read what the seven sentences are about. One mandates presence. One dimensions the facemask. Two govern eye shields. Two govern manufacturer branding. The last places two labels.

Six of the seven concern identification — whose name appears, where, and what is stuck to the shell. The most precise number in the paragraph, and the only dimension in it, describes the diameter of the tubing in front of the player’s face.

That a helmet must be worn at all is fixed separately, in the clause listing the equipment a player may not omit at his option.

What the rulebook does not contain

The rulebook was downloaded as a PDF and extracted to plain text with pdftotext -layout — 453,356 bytes, 452,278 characters. The following strings occur zero times in it:

The word helmet occurs 76 times, and the distribution is the finding:

The largest group by a distance is about hitting someone with a helmet. Not one of the 76 requires a helmet to meet a performance standard, hold a certification, or appear on an approved list.

Method note, because it matters: an earlier pass using a summarising fetch reported this same negative confidently — from a document it had silently truncated at the very paragraph in question. That confident negative was worthless. The claim above rests on the extracted text of the complete PDF, and the counts are reproducible from the archived file.

Where the numbers are, part one: permission to sell

NOCSAE ND002-25, Standard Performance Specification for Newly Manufactured Football Helmets, revision July 2025, sets what a helmet must survive to be sold:

The apparatus is a headform — drop impacts at 11.34, 13.89, 16.04 and 17.94 ft/s across front, side, front boss, rear boss, rear, top and random sites, and a pneumatic ram at 19.7 ft/s. Small, medium and large headforms, conditioned at 72°F ± 5°F for at least four hours, and at 100 ± 3°F for the high-temperature condition.

A headform. No neck, no body.

The label the rulebook places, and the statement the standard requires

These are two requirements from two documents, and this record does not merge them.

The playing rules require “an approved warning label on the rear lower-right exterior”, supplied in quantity by the League office. They do not state what it says.

Separately, ND002-25 §9.3 requires the manufacturer to affix to the exterior of the shell a clearly legible statement communicating the following information, “using the same or similar language”:

WARNING NO HELMET CAN PREVENT ALL HEAD OR ANY NECK INJURIES A PLAYER MIGHT RECEIVE WHILE PARTICIPATING IN FOOTBALL. HELMETS CANNOT PREVENT CONCUSSION/BRAIN INJURY. SEEK MEDICAL ADVICE BEFORE RETURNING TO PLAY IF YOU SUSPECT ANY INJURY.

§9.2.1 separately requires a label warning the user that no helmet can protect against all possible impacts.

Whether the League-supplied label carries the ND002 wording is not documented in either direction. No located League document specifies that label’s wording, appearance or origin. The two requirements are stated here separately because that is how the documents state them.

The warning text above is quoted as the standard’s own words. It is not this record’s conclusion about any helmet.

Where the numbers are, part two: the ranking

The Helmet Test Protocol published by Biomechanics Consulting and Research (Biocore), dated 11 September 2019, does something else entirely:

• A linear ram of total mass 15.6 kg ± 3%, its face extruded nylon 6/6 in the shape of a spherical cap — 140 mm radius sphere, 203 mm diameter cap.
• The head and neck from a certified Hybrid III 50th-percentile male anthropomorphic test device.
• Six locations — Side Upper, Oblique Front, Oblique Rear, Side, Facemask Side, Facemask Central Oblique — at three speeds: 5.5, 7.4 and 9.3 m/s, tolerance ± 2%. Eighteen impacts.
• Weighted into a single figure: HARM = 0.0148·HIC + 15.6·DAMAGE, summed across locations with weights reflecting on-field concussion incidence.

Certification drops a headform. The ranking rams a head and a neck. Different apparatus, different metric, different threshold — and only the first decides whether a helmet may be sold at all.

Three groups, one gate

The NFL/NFLPA Helmet Laboratory Testing Performance Results sort helmet models into three groups. From the published results page:

Effective as of the 2026 season, all helmets outside of the top-performing group … will be marked as “Not Recommended.”

Top-performing is a recommendation. “Not Recommended” is a label. Prohibited removes a helmet from play.

Only the last of the three does anything a player must obey. A model can rank below every other permitted helmet and be worn on Sunday with no rule broken.

The instrument that gives the prohibition its force was not located by this audit. The playing rules do not carry it — that is proven above, not inferred. The 2020 collective bargaining agreement and the joint testing announcements through 14 April 2026 were also checked. Press releases state that the League and the players’ association jointly prohibit models; no located primary document names the instrument or carries the clause. A search that returns nothing is a fact about the search, not proof that nothing exists.

The chain, in one pass

1. The rule requires the object — present, fastened, a facemask under ⅝ of an inch, two labels on the back.
2. One standard, which the rule never names, decides whether it may be sold.
3. A different protocol, on different apparatus, decides where it ranks.
4. Of the three groups that ranking produces, only one removes a helmet from play — and the document carrying that authority was not located.

Four documents. The rulebook is the only one a player is bound by, and it is the one that states no performance requirement at all.

Sources

• 2026 Official Playing Rules of the National Football League — Rule 5, Section 4, Article 3, Item 1, quoted in full above; the string counts taken from the extracted text of the complete document.
• NOCSAE ND002-25, Standard Performance Specification for Newly Manufactured Football Helmets, revision July 2025, effective August 2026 — §6.1–6.4 thresholds, §9.2.1 and §9.3 labelling.
• NOCSAE ND 001-17m17b, Standard Test Method and Equipment Used in Evaluating the Performance Characteristics of Headgear/Equipment, January 2017 with December 2017 modification — the test apparatus ND002-25 refers to.
• Helmet Test Protocol, Biomechanics Consulting and Research (Biocore), 11 September 2019 — the ram, the Hybrid III head and neck, the locations and speeds, the HARM equation.
• NFL/NFLPA Helmet Laboratory Testing Performance Results — the three tiers, quoted from the published results page.

Every source above was fetched and read as a primary document. Each link was confirmed to resolve and to contain the claim attributed to it.

Open items

• The 2026 poster and the prohibited-models list are not held by this audit. These documents demonstrably exist; the results page announces and links them. We do not hold the files, so the tier language above is quoted from the results page and attributed to it, and nothing here is attributed to the poster. This is a gap in our archive, not an absence in the world.
• The governing revision of the Biocore protocol is not established. The copy held and quoted is the revision published in September 2019. A later revision may govern the current testing cycle; if so, the mass, speeds and coefficients above would need re-quoting. Every apparatus figure here is attributed to the 2019 edition.
• Currency is not independently corroborated. Each document is primary from its own publisher, which is authority but not corroboration that it is the edition in force.
• Neither the League, the players’ association, nor NOCSAE has been asked to comment on the reading set out here.

What this record does not claim

It does not claim that any helmet is unsafe, or that any helmet causes or prevents an injury. The record reports what each document measures and what each label says. The NOCSAE warning text is quoted as that standard’s own words.

It does not claim that the laboratory ranking is invalid because its apparatus is a head and neck without a body. The apparatus is stated and left to stand.

It does not claim that the prohibition has no legal instrument — only that this audit could not locate it in the playing rules, the collective bargaining agreement, or the joint announcements.

It does not claim that the League-supplied warning label and the NOCSAE statement are the same label, or that they are different. That relationship is undocumented in both directions, and inventing a link between them would be the thematic manufacture this series exists to avoid.

It does not claim that a rule was broken, by anyone. No player is accused of anything. No manufacturer is named. The finding is that the only document binding on a player specifies identification and says nothing about protection — narrower than a violation, and harder to correct than one.