North Dakota — The Booklet

Deer Bow

Sept. 4 (12 noon) - Jan. 3

open now ends 2027-01-03 · derived as of Sep 23

2026-09-04 → 2027-01-03 derived (year inferred)

areas_as_printed	Statewide
closes_as_printed	Jan. 3
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 100%, reading B 100% (longest run shared with A 165 characters, with B 165). VERBATIM IN BOTH: every character of this row's text appears verbatim in both independent blind visual readings of the signed scan. That is the strongest corroboration available for this state. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
duration_label_as_printed	(none printed)
licence_class	Residents
opens_as_printed	Sept. 4 (12 noon)
restrictions	Residents – Any deer is legal. No unit restrictions. Orange clothing required of all bow hunters and those accompanying apprentice license holders during Deer Gun season.
section	1. Season Dates, Times, Open Areas
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.

show me the page

No page is shown. no corpus PDF matches this record: no file hashes to bd09260e047d4e9f... and no .pdf.source sidecar names 'https://gf.nd.gov/regulations/deer'. The source was most likely not a PDF, or the PDF is not in the corpus.

source

Deer Bow

Statewide

Sept. 4 (12 noon)

Jan. 3

Residents – Any deer is legal. No unit restrictions. Orange clothing required of all bow hunters and those accompanying apprentice license holders during Deer Gun season.

2026-2027 North Dakota Deer Hunting Proclamation, as published in text by the North Dakota Game and Fish Department at gf.nd.gov/regulations/deer under the heading '2026-2027 Deer Hunting Proclamation' and the Department's own note 'The following is a copy of the text from the 2026-2027 Deer Hunting Proclamation'; signed instrument issued by Governor Kelly Armstrong, 5th day of May 2026, and published as a 22-page scan at gf.nd.gov/gnf/regulations/docs/deer/proc-deer-2026.pdf (sha256 e4e1ac7d…, ZERO text characters on all 22 pages)

areas_as_printed	All Units
closes_as_printed	Nov. 22
contradicted_by_signed_instrument	THE SIGNED SCAN PRINTS A DIFFERENT DEER GUN SEASON ON p.12. The heading over the licence-quota table on p.12 of proc-deer-2026.pdf reads '16 1/2 DAY SEASON / November 8 (12 Noon) - November 24', recorded independently and byte-identically by both blind readers. Counted on this run, the string 'November 8' occurs 0 times in this HTML page and 'November 24' occurs 0 times. p.1 of the same scan prints the dates published on this row, and p.13's 'the first 2 1/2 days (November 6-8, 2026)' -- carried verbatim in this HTML -- corroborates November 6. This row publishes what the Department's text publishes and does not resolve the conflict; see the 'SOURCE CONTRADICTION' scope row.
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 100%, reading B 100% (longest run shared with A 340 characters, with B 340). VERBATIM IN BOTH: every character of this row's text appears verbatim in both independent blind visual readings of the signed scan. That is the strongest corroboration available for this state. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
duration_label_as_printed	16 ½ Day Season
licence_class	all licence classes
opens_as_printed	Nov. 6 (12 noon)
restrictions	Restricted to type of deer and unit on license. Type of deer includes species and whether antlerless or antlered (at least one visible antler). See Section 16 for unit descriptions. Note exception for white-tailed deer hunter in 4B, 4C, 4D and 4E. Orange clothing required for gun hunters and those accompanying apprentice and youth license holders.
section	1. Season Dates, Times, Open Areas
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.

areas_as_printed	the seven Deer Management Zones described in Section 4 -- South Bismarck/Mandan, City of Fargo, City of Minot, City of Watford City, City of Portland, Grand Forks Air Force Base, and the USDA-ARS and NDDOCR properties within South Bismarck/Mandan. Each zone's extent is drawn on a map (Figures 1-7) and is a declared gap; no boundary is derived here from any zone's name.
closes_as_printed	30 minutes after sunset January 3 of the following year
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 100%, reading B 100% (longest run shared with A 399 characters, with B 399). VERBATIM IN BOTH: every character of this row's text appears verbatim in both independent blind visual readings of the signed scan. That is the strongest corroboration available for this state. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
duration_label_as_printed	(none printed)
licence_class	Special Deer-Bow Herd Reduction licence holders (concurrent with deer bow season)
not_in_the_season_table	This season is NOT listed in Section 1's season table. It is set out in prose in Section 4 and is emitted as a season row here so that a reader of season rows sees it.
opens_as_printed	12:00 (noon) Central Time, in conjunction with deer bow season
restrictions	A person may possess no more than two Deer-Bow Herd Reduction licenses per year. Further restrictions may apply for season dates, locations, number of licenses and other specifications in the permitted management zones according to the managing entity. Concurrent Deer-Bow Herd Reduction Seasons for any antlerless deer will be available for the areas listed below. These areas include public and privately owned land. All privately owned land requires access permission from the landowner. All lands within city limits or in specified management zones require an access permit in addition to the Special Deer-Bow Herd Reduction license unless otherwise noted. Access permits are limited and based on availability. The cost for each Deer-Bow Herd Reduction license is the same as their respective resident and non-resident big game fee. The bag limit shall be one any antlerless deer for each license.
second_opening_date_in_the_figure_7_legend	THE SIGNED SCAN CARRIES A SECOND OPENING DATE FOR PART OF THE GRAND FORKS AIR FORCE BASE ZONE, AND IT EXISTS ONLY IN A MAP LEGEND. Figure 7 on p.22 of proc-deer-2026.pdf distinguishes a DARK GREEN fill, 'Available All Bow Season', from a LIGHT GREEN fill, 'Opens November 1st'. The legend text is corroborated by BOTH independent blind visual readings, verified against their own JSON this run: reading A records 'Opens November 1st' and reading B records 'Opens November 1st (light green)'. WHICH GROUND EACH SHADE COVERS IS NOT KNOWN TO THIS RECORD AND IS NOT GUESSED: both readers reported in `uncertain` that the two greens are too close in hue on this scan to assign any numbered zone to either, and binding a zone to a colour would be reading colour as data. So this row's opening time is the only one the TEXT of the proclamation states, and it is NOT known to be the only opening date that applies inside the Grand Forks Air Force Base zone. The unread map is declared separately as a gap; this field is on the row because it changes what the row means.
section	4. Special Deer-Bow Herd Reduction Seasons
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.

cross_check	NOT APPLICABLE -- this sentence belongs to the HTML publication and is not part of the signed instrument, so neither blind reading of the scan could contain it. That is the correct result, not a coverage gap.
value	NOTE: The following is a copy of the text from the 2026-2027 Deer Hunting Proclamation. Additional regulations can be found on the page sidebar (if on mobile, additional regulations are posted at the bottom of the page). A scanned copy of the signed proclamation is also available (pdf).
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
why_it_matters	This record is built from a COPY of the proclamation, published by the issuing agency, and not from the signed instrument. The signed instrument is the 22-page scan at https://gf.nd.gov/gnf/regulations/docs/deer/proc-deer-2026.pdf (sha256 e4e1ac7db059b769e9f85a031a66c955906292b03bb9393a5424b995e50354b3), which carries ZERO text characters on all 22 pages and therefore cannot be parsed at all. Where the two differ, this record says so on the row rather than choosing.

cross_check	NOT APPLICABLE -- the sidebar index is a web-page navigation element and is not part of the signed instrument, so neither blind reading could contain it.
value	5. Special Deer-Bow Herd Reduction Seasons \|\| 5. Aircraft, Motor-Driven Vehicles, and Lights
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
why_it_matters	The body of the proclamation puts Special Deer-Bow Herd Reduction Seasons INSIDE Section 4 (Section 6 cites it as '(Section 4)') and numbers Aircraft, Motor-Driven Vehicles, and Lights as Section 5. The index is therefore wrong and the body is self-consistent. Recorded rather than silently corrected: a reader following the index to 'Section 5' for herd reduction lands on the aircraft rules.

cross_check	NOT APPLICABLE -- these are hyperlinks on the web page, not text of the signed instrument.
value	Additional Regulations; Youth Specific Season Information; Season General Information; Eligibility; Firearms and Archery Equipment; Note: Regular season regulations also apply to the youth season.; Posting and Trespass; Hunting by Nontribal Members on North Dakota Indian Reservations; Road Rights of Way; Hunter Education Requirement; Minimum Age Requirements; Hunting Over Bait; Use of Animals and Tree Stands; Century and Administrative Code; Century Code; Administrative Code; General Information Links; Lead in Venison
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
why_it_matters	The proclamation is not the whole of North Dakota deer law and the page says so by linking out. Posting and trespass, hunting by nontribal members on North Dakota Indian reservations, road rights of way, the hunter education requirement, minimum age requirements, hunting over bait, use of animals and tree stands, N.D. Century Code Title 20.1 and N.D. Administrative Code Title 30 are each a separate document. None of them is fetched, parsed or published here. This row exists so the omission is on the record and not only in a known_gap.

cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 39%, reading B 10% (longest run shared with A 95 characters, with B 34). NO FULL VERBATIM MATCH IN EITHER READING. The two figures are published and no verdict is drawn from them. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
publisher_note	*Note: This was changed in the html version of the proclamation after signing. Please use the Watford City Police Department number for Watford City Deer Management Access Permits.
struck_through_in_the_html	City of Watford City (701-444-2280)
value	Portions of the City of Watford City and adjacent areas referred to as the Watford City Deer Management Zone (Figure 5). Hunters who desire to hunt within the described Watford City Deer Management Zone must first obtain a Watford City Deer Management Access Permit from the City of Watford City (701-444-2280) * City of Watford City Police Department (701-842-2280).
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
why_it_matters	The HTML strikes '701-444-2280' through with an <s> element and prints '701-842-2280' beside it, and the Department's own note says the change was made in the HTML version AFTER SIGNING. A hunter holding the signed PDF therefore sees only the struck number. Both numbers are published here. The blind readers read '701-444-2280' on p.7 and '701-444-2533' in the Figure 5 legend on p.20 of the scan; counted on this run, '701-444-2533' occurs 0 times in this HTML while '701-444-2280' occurs 1 and '701-842-2280' occurs 1. Reading B rated its own '2533' reading only moderately confident. All three numbers are on this record and none is chosen between.

cross_check	Blank/printed pattern confirmed cell by cell against both blind readings -- see the quota cross-check scope row for the totals.
hunting_unit	1
management_unit	Turtle Mtns
value	Type C, Type D, Type E, Type F
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
why_it_matters	These cells are EMPTY in the source, not '0'. The same table prints an explicit '0' in 20 other cells, so the document itself distinguishes blank from zero -- and it nowhere states what a blank means. It could mean no licences are offered of that type in that unit, or that the type is not available there at all. Both blind readers of the signed scan flagged this distinction as load-bearing and unexplained. NO quota row is emitted for these types in this unit: inventing 0 would publish a figure the proclamation does not state.

cross_check	Reading A independently reconciled the same six totals from the pixels and recorded that it did so; the grand total of 39,100 is printed in Section 11 prose and carried verbatim in this record.
value	ALL RECONCILE -- Type A: computed 21,250 vs printed 21,250 OK; Type B: computed 13,400 vs printed 13,400 OK; Type C: computed 2,000 vs printed 2,000 OK; Type D: computed 850 vs printed 850 OK; Type E: computed 1,200 vs printed 1,200 OK; Type F: computed 400 vs printed 400 OK; grand total computed 39,100 vs '39,100 licenses, including gratis licenses, will be issued' printed in Section 11 prose.
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
why_it_matters	ADDING-A-STATE.md §5: wherever a source prints a total, use it. Six column totals and one prose grand total are checks the Department computed on its own numbers, and reproducing them from the 112 cells this adapter parsed is stronger evidence about this table than anything the extractor can say about itself. Recomputed on every run rather than asserted once.

cross_check	This row IS the cross-check.
value	228 cells compared (112 carrying a printed figure, 116 blank). Reading A agrees on 228; reading B agrees on 124; disagreements 0; cells neither blind reader recorded 0.
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
why_it_matters	This is what replaces core.library_divergence for North Dakota, and on this table it is a stronger check than library_divergence is anywhere else in this compilation: library_divergence asks whether two libraries read the same bytes the same way, whereas this asks whether a text artifact and two independent human readings of a scanned artifact of the SAME instrument agree, value by value, including the blank-versus-zero distinction. Computed on every run from the readings' own JSON, not asserted.

cross_check	This row IS the coverage census, computed on this run from the two readings' own JSON. It replaces a per-row 'NO pixel witness' sentence that claimed to mark the rows no reader saw and could mark only the 9 rows with zero string overlap.
the_paragraph_neither_reading_reproduces	Youth Deer Season (Residents Only) – Residents whose 11th, 12th or 13th birthday occurs in the same calendar year as a youth deer hunting season are eligible for an any antlerless deer license valid for Youth Deer season and Deer Gun season. Individuals age 14 and 15 who have either not previously applied for a regular Deer Gun license or who have applied but have never received a regular Deer Gun license are eligible to receive one deer license valid for the Youth Deer season and Deer Gun season, this includes youth who are age 13 providing they will be age 14 during the same year as the respective big game hunting season as well as those youth who apply at age 15 but will be turning 16 after they have applied for their license.
value	91 of 350 proclamation paragraphs are reproduced VERBATIM IN FULL by BOTH independent blind visual readings of the signed scan. By section -- section Preamble: 1 of 5 reproduced in full \|\| section 1: 15 of 37 reproduced in full \|\| section 2: 3 of 5 reproduced in full (NOT in full: node 206 "Licenses for Military Personnel – North Dakota residents who..." A 6% B 25%; node 207 "Youth Deer Season (Residents Only) – Residents whose 11th, 1..." A 5% B 3%) \|\| section 3: 3 of 10 reproduced in full \|\| section 4: 2 of 12 reproduced in full \|\| section 4 (Special Deer-Bow Herd Reduction Seasons): 3 of 28 reproduced in full \|\| section 5: 0 of 7 reproduced in full \|\| section 6: 8 of 8 reproduced in full \|\| section 7: 2 of 3 reproduced in full (NOT in full: node 283 "When any part of an animal is mounted, if the tag is removed..." A 4% B 0%) \|\| section 8: 0 of 5 reproduced in full \|\| section 9: 2 of 8 reproduced in full \|\| section 10: 1 of 1 reproduced in full \|\| section 11: 43 of 192 reproduced in full \|\| section 12: 1 of 1 reproduced in full \|\| section 13: 0 of 3 reproduced in full \|\| section 14: 0 of 1 reproduced in full \|\| section 15: 0 of 10 reproduced in full \|\| section 16: 7 of 7 reproduced in full \|\| section Figures: 0 of 7 reproduced in full
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
why_it_matters	TWO INDEPENDENT READERS SHARING A BLIND SPOT IS THE FAILURE THIS METHOD CANNOT DETECT ROW BY ROW, and it is not theoretical: it happened on this document. Section 2 has five paragraphs; both readings reproduce 'Who May Apply', 'Anyone applying for these licenses...' and the Quadriplegics paragraph verbatim IN FULL, and neither reproduces 'Youth Deer Season (Residents Only)' in full -- 5% and 3%, the figures published on that row -- which is the paragraph HANDOFF-CYCLE-43.md §4 names as the one substantive provision BOTH blind readers missed. That is re-derived here by measurement on this run and is not carried over from the handoff. The fifth, 'Licenses for Military Personnel', is at 6% and 25%, and HANDOFF-CYCLE-43.md §4 records it as the one entitlement that rested on a single reader. A LOW FIGURE IS NOT A FINDING THAT A PARAGRAPH WAS UNREAD: both readers transcribed selectively and summarised in places, and reading A's summaries of the seven Section 4 centerfire areas are the measured example. What this row establishes is the opposite direction -- where a section is otherwise transcribed in full, a paragraph that is not stands out, and the reader can see it rather than being told there are none.

cross_check	DIVERGENCE, and it is a real one about document content rather than a recording artefact: reading A and reading B both record the p.12 heading, and the text route does not contain it.
value	signed scan p.1 and this HTML: 'Nov. 6 (12 noon)' to 'Nov. 22', labelled '16 ½ Day Season' \|\| signed scan p.12, heading over the licence-quota table: '16 ½ DAY SEASON / November 8 (12 Noon) - November 24' \|\| this HTML page: 'November 8' occurs 0 times and 'November 24' occurs 0 times, counted on this run
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
why_it_matters	This is the state's main season and the two printings differ by two days at each end. Both blind readers recorded the p.12 heading independently and byte-identically, both re-rendered it at higher scale, and both flagged it first among their uncertainties. p.13 of the same scan prints 'the first 2 1/2 days (November 6-8, 2026)', which corroborates November 6 and which this HTML carries verbatim. A compiler that scrapes the p.12 table header emits the wrong season for the entire state; a compiler that silently prefers p.1 has made the same class of decision in the other direction and told nobody. This record publishes the dates the Department's own text publishes, and publishes the conflict beside them rather than resolving it.

area_as_printed	That portion of unit 2L starting where ND Highway 19 meets the east shore of Six Mile Bay on Devils Lake, then south and east along the north shore of Devils Lake to the South Creel Township line, then north to ND Highway 19, then west on ND Highway 19 to the point of origin, except Camp Grafton where a limited number of special either sex permits will be issued to disabled veterans. Check with the Barnes County Veterans Service Office for details (701) 845-8511.
area_number	area 1 of 7 listed under 'Areas Closed to Hunting with Centerfire Rifles'
blind_reading_a_record	READING A's OWN RECORD OF THIS AREA, paired by measurement rather than by list position (this area overlaps reading A entry 0 at 15% and no other entry above 0%): 'A. That portion of unit 2L bounded by ND Highway 19 / east shore of Six Mile Bay on Devils Lake / north shore of Devils Lake / South Creel Township line, except Camp Grafton'. Reading A recorded all seven of these areas as SUMMARIES rather than transcriptions, which is why the verbatim overlap figures across the seven range from single digits to 100% while the reader read every one of them. That is the measured proof that a low verbatim figure is not a finding of non-coverage.
blind_reading_b_record	reading B recorded the lead-in sentence of this list verbatim and did not transcribe the seven areas beneath it.
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 54%, reading B 33% (longest run shared with A 231 characters, with B 231). NO FULL VERBATIM MATCH IN EITHER READING. The two figures are published and no verdict is drawn from them. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
joined_from	two adjacent nodes: the lead-in (index 225) states the restriction and ends in a colon, the area (index 226) states where it applies. Neither is a whole statement alone.
section	4. Firearms and Archery Equipment (Areas Closed to Hunting with Centerfire Rifles)
this_is_not_a_closure_of_the_area	THE AREA IS OPEN TO DEER HUNTING. Section 4's heading reads 'Areas Closed to Hunting with Centerfire Rifles' and closes a class of firearm, not the ground. These seven were published as `closure` rows until this run, so a consumer reading this state's closures got seven weapon rules and none of the real closures. The real closures are Section 9's, and they are emitted as `closure` rows.
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
weapon_restriction	During the open Deer Gun season, on this area, deer may be taken ONLY with handguns, shotguns firing rifled slugs, legal muzzleloading firearms or legal bow equipment.

area_as_printed	The areas north and south of Bismarck described as follows: In unit 3B3, starting on ND Highway 1804 at the north boundary of the Double Ditch Indian Village site, then south on ND Highway 1804 to the junction with the road known as River Road, then south on River Road to the Bismarck city limits, then west to the west bank of the Missouri River, then following the west bank of the Missouri River to a point directly west of the north boundary of the Double Ditch Indian Village site, to include river islands and sandbars, then east to the point of origin. In unit 3C, an area starting where the southwest boundary of the city limits of Bismarck join the east bank of the Missouri River, then following the city limits of Bismarck easterly to the point where it meets the west bank of Apple Creek in the northeast one-quarter of Section 26, Township 138 North, Range 80 West, then following the west bank of Apple Creek in a general southwest direction to its junction with the north boundary of Apple Creek Wildlife Management Area (WMA) and then west and south along the WMA boundary to the Missouri River, then following the east bank of the Missouri River to the point of origin.
area_number	area 2 of 7 listed under 'Areas Closed to Hunting with Centerfire Rifles'
blind_reading_a_record	READING A's OWN RECORD OF THIS AREA, paired by measurement rather than by list position (this area overlaps reading A entry 1 at 4% and no other entry above 0%): 'B. Areas north and south of Bismarck: described portion of unit 3B3, and described portion of unit 3C'. Reading A recorded all seven of these areas as SUMMARIES rather than transcriptions, which is why the verbatim overlap figures across the seven range from single digits to 100% while the reader read every one of them. That is the measured proof that a low verbatim figure is not a finding of non-coverage.
blind_reading_b_record	reading B recorded the lead-in sentence of this list verbatim and did not transcribe the seven areas beneath it.
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 20%, reading B 17% (longest run shared with A 231 characters, with B 231). NO FULL VERBATIM MATCH IN EITHER READING. The two figures are published and no verdict is drawn from them. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
joined_from	two adjacent nodes: the lead-in (index 225) states the restriction and ends in a colon, the area (index 227) states where it applies. Neither is a whole statement alone.
section	4. Firearms and Archery Equipment (Areas Closed to Hunting with Centerfire Rifles)
this_is_not_a_closure_of_the_area	THE AREA IS OPEN TO DEER HUNTING. Section 4's heading reads 'Areas Closed to Hunting with Centerfire Rifles' and closes a class of firearm, not the ground. These seven were published as `closure` rows until this run, so a consumer reading this state's closures got seven weapon rules and none of the real closures. The real closures are Section 9's, and they are emitted as `closure` rows.
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
weapon_restriction	During the open Deer Gun season, on this area, deer may be taken ONLY with handguns, shotguns firing rifled slugs, legal muzzleloading firearms or legal bow equipment.

area_as_printed	That portion of unit 3C starting on ND Highway 1806 from Mandan city limits south to a point directly west of the mouth of the Heart River. Then east to the west bank of the Missouri River, then northerly along the west bank of the Missouri River to Mandan city limits and then west along Mandan city limits to the point of origin.
area_number	area 3 of 7 listed under 'Areas Closed to Hunting with Centerfire Rifles'
blind_reading_a_record	READING A's OWN RECORD OF THIS AREA, paired by measurement rather than by list position (this area overlaps reading A entry 2 at 38% and no other entry above 5%): 'C. That portion of unit 3C from Mandan city limits south along ND Highway 1806 to a point directly west of the mouth of the Heart River'. Reading A recorded all seven of these areas as SUMMARIES rather than transcriptions, which is why the verbatim overlap figures across the seven range from single digits to 100% while the reader read every one of them. That is the measured proof that a low verbatim figure is not a finding of non-coverage.
blind_reading_b_record	reading B recorded the lead-in sentence of this list verbatim and did not transcribe the seven areas beneath it.
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 63%, reading B 42% (longest run shared with A 231 characters, with B 231). NO FULL VERBATIM MATCH IN EITHER READING. The two figures are published and no verdict is drawn from them. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
joined_from	two adjacent nodes: the lead-in (index 225) states the restriction and ends in a colon, the area (index 228) states where it applies. Neither is a whole statement alone.
section	4. Firearms and Archery Equipment (Areas Closed to Hunting with Centerfire Rifles)
this_is_not_a_closure_of_the_area	THE AREA IS OPEN TO DEER HUNTING. Section 4's heading reads 'Areas Closed to Hunting with Centerfire Rifles' and closes a class of firearm, not the ground. These seven were published as `closure` rows until this run, so a consumer reading this state's closures got seven weapon rules and none of the real closures. The real closures are Section 9's, and they are emitted as `closure` rows.
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
weapon_restriction	During the open Deer Gun season, on this area, deer may be taken ONLY with handguns, shotguns firing rifled slugs, legal muzzleloading firearms or legal bow equipment.

area_as_printed	Approximately 171 acres so posted in unit 3B1 within Sections 2, 10, and 11, Township 151 North, Range 104 West in McKenzie County, North Dakota.
area_number	area 4 of 7 listed under 'Areas Closed to Hunting with Centerfire Rifles'
blind_reading_a_record	READING A's OWN RECORD OF THIS AREA, paired by measurement rather than by list position (this area overlaps reading A entry 3 at 91% and no other entry above 0%): 'D. Approximately 171 acres so posted in unit 3B1 within Sections 2, 10, and 11, Township 151 North, Range 104 West in McKenzie County'. Reading A recorded all seven of these areas as SUMMARIES rather than transcriptions, which is why the verbatim overlap figures across the seven range from single digits to 100% while the reader read every one of them. That is the measured proof that a low verbatim figure is not a finding of non-coverage.
blind_reading_b_record	reading B recorded the lead-in sentence of this list verbatim and did not transcribe the seven areas beneath it.
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 97%, reading B 65% (longest run shared with A 231 characters, with B 231). NO FULL VERBATIM MATCH IN EITHER READING. The two figures are published and no verdict is drawn from them. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
joined_from	two adjacent nodes: the lead-in (index 225) states the restriction and ends in a colon, the area (index 229) states where it applies. Neither is a whole statement alone.
section	4. Firearms and Archery Equipment (Areas Closed to Hunting with Centerfire Rifles)
this_is_not_a_closure_of_the_area	THE AREA IS OPEN TO DEER HUNTING. Section 4's heading reads 'Areas Closed to Hunting with Centerfire Rifles' and closes a class of firearm, not the ground. These seven were published as `closure` rows until this run, so a consumer reading this state's closures got seven weapon rules and none of the real closures. The real closures are Section 9's, and they are emitted as `closure` rows.
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
weapon_restriction	During the open Deer Gun season, on this area, deer may be taken ONLY with handguns, shotguns firing rifled slugs, legal muzzleloading firearms or legal bow equipment.

area_as_printed	Those wildlife management areas posted with firearms restrictions at major access points.
area_number	area 5 of 7 listed under 'Areas Closed to Hunting with Centerfire Rifles'
blind_reading_a_record	READING A's OWN RECORD OF THIS AREA, paired by measurement rather than by list position (this area overlaps reading A entry 4 at 100% and no other entry above 0%): 'E. Those wildlife management areas posted with firearms restrictions at major access points'. Reading A recorded all seven of these areas as SUMMARIES rather than transcriptions, which is why the verbatim overlap figures across the seven range from single digits to 100% while the reader read every one of them. That is the measured proof that a low verbatim figure is not a finding of non-coverage.
blind_reading_b_record	reading B recorded the lead-in sentence of this list verbatim and did not transcribe the seven areas beneath it.
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 100%, reading B 72% (longest run shared with A 231 characters, with B 231). VERBATIM IN ONE: blind reading A reproduces every character of this row's text; the other reading does not reproduce it in full, which is not a statement about whether that reader saw it. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
joined_from	two adjacent nodes: the lead-in (index 225) states the restriction and ends in a colon, the area (index 230) states where it applies. Neither is a whole statement alone.
section	4. Firearms and Archery Equipment (Areas Closed to Hunting with Centerfire Rifles)
this_is_not_a_closure_of_the_area	THE AREA IS OPEN TO DEER HUNTING. Section 4's heading reads 'Areas Closed to Hunting with Centerfire Rifles' and closes a class of firearm, not the ground. These seven were published as `closure` rows until this run, so a consumer reading this state's closures got seven weapon rules and none of the real closures. The real closures are Section 9's, and they are emitted as `closure` rows.
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
weapon_restriction	During the open Deer Gun season, on this area, deer may be taken ONLY with handguns, shotguns firing rifled slugs, legal muzzleloading firearms or legal bow equipment.

area_as_printed	Approximately 90 acres so posted in unit 3B3 within Section 18, Township 147 North, Range 83 West in the Wolf Creek Campground area in McLean County, North Dakota. The lakeside boundary is the water’s edge.
area_number	area 6 of 7 listed under 'Areas Closed to Hunting with Centerfire Rifles'
blind_reading_a_record	READING A's OWN RECORD OF THIS AREA, paired by measurement rather than by list position (this area overlaps reading A entry 5 at 73% and no other entry above 0%): 'F. Approximately 90 acres so posted in unit 3B3 within Section 18, Township 147 North, Range 83 West in the Wolf Creek Campground area in McLean County'. Reading A recorded all seven of these areas as SUMMARIES rather than transcriptions, which is why the verbatim overlap figures across the seven range from single digits to 100% while the reader read every one of them. That is the measured proof that a low verbatim figure is not a finding of non-coverage.
blind_reading_b_record	reading B recorded the lead-in sentence of this list verbatim and did not transcribe the seven areas beneath it.
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 88%, reading B 55% (longest run shared with A 231 characters, with B 231). NO FULL VERBATIM MATCH IN EITHER READING. The two figures are published and no verdict is drawn from them. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
joined_from	two adjacent nodes: the lead-in (index 225) states the restriction and ends in a colon, the area (index 231) states where it applies. Neither is a whole statement alone.
section	4. Firearms and Archery Equipment (Areas Closed to Hunting with Centerfire Rifles)
this_is_not_a_closure_of_the_area	THE AREA IS OPEN TO DEER HUNTING. Section 4's heading reads 'Areas Closed to Hunting with Centerfire Rifles' and closes a class of firearm, not the ground. These seven were published as `closure` rows until this run, so a consumer reading this state's closures got seven weapon rules and none of the real closures. The real closures are Section 9's, and they are emitted as `closure` rows.
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
weapon_restriction	During the open Deer Gun season, on this area, deer may be taken ONLY with handguns, shotguns firing rifled slugs, legal muzzleloading firearms or legal bow equipment.

area_as_printed	Approximately 60 acres so posted in unit 2K1 within sections 21 and 28, Township 148 North, Range 83 West in the East Totten Trail Campground area in McLean County, North Dakota. The lakeside boundary is the water’s edge.
area_number	area 7 of 7 listed under 'Areas Closed to Hunting with Centerfire Rifles'
blind_reading_a_record	READING A's OWN RECORD OF THIS AREA, paired by measurement rather than by list position (this area overlaps reading A entry 6 at 75% and no other entry above 0%): 'G. Approximately 60 acres so posted in unit 2K1 within sections 21 and 28, Township 148 North, Range 83 West in the East Totten Trail Campground area in McLean County'. Reading A recorded all seven of these areas as SUMMARIES rather than transcriptions, which is why the verbatim overlap figures across the seven range from single digits to 100% while the reader read every one of them. That is the measured proof that a low verbatim figure is not a finding of non-coverage.
blind_reading_b_record	reading B recorded the lead-in sentence of this list verbatim and did not transcribe the seven areas beneath it.
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 88%, reading B 55% (longest run shared with A 231 characters, with B 231). NO FULL VERBATIM MATCH IN EITHER READING. The two figures are published and no verdict is drawn from them. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
joined_from	two adjacent nodes: the lead-in (index 225) states the restriction and ends in a colon, the area (index 232) states where it applies. Neither is a whole statement alone.
section	4. Firearms and Archery Equipment (Areas Closed to Hunting with Centerfire Rifles)
this_is_not_a_closure_of_the_area	THE AREA IS OPEN TO DEER HUNTING. Section 4's heading reads 'Areas Closed to Hunting with Centerfire Rifles' and closes a class of firearm, not the ground. These seven were published as `closure` rows until this run, so a consumer reading this state's closures got seven weapon rules and none of the real closures. The real closures are Section 9's, and they are emitted as `closure` rows.
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
weapon_restriction	During the open Deer Gun season, on this area, deer may be taken ONLY with handguns, shotguns firing rifled slugs, legal muzzleloading firearms or legal bow equipment.

boundary_rule_preamble	ALL UNITS ARE BOUNDED BY INTERSTATE HIGHWAYS, US HIGHWAYS OR NORTH DAKOTA STATE HIGHWAYS AS NOTED ON THE MAP, WITH THE FOLLOWING EXCEPTIONS:
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 100%, reading B 100% (longest run shared with A 170 characters, with B 170). VERBATIM IN BOTH: every character of this row's text appears verbatim in both independent blind visual readings of the signed scan. That is the strongest corroboration available for this state. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
section	16. Deer Gun Hunting Units
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.

cross_check	NOT APPLICABLE -- the Department added this note to the HTML version AFTER the proclamation was signed, and says so in the note itself. It is not part of the signed instrument, so neither blind reading of the scan could contain it. That is the correct result, not a gap.
section	4. Special Deer-Bow Herd Reduction Seasons -- City of Watford City
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.

Check this page against the source

season 7

Deer Bow

Deer Bow

Deer Bow

Deer Gun

Muzzeloader

Youth Deer

Special Deer-Bow Herd Reduction Seasons

scope 51

WHAT THIS SOURCE IS, in the Department's own words

SOURCE DEFECT: the page's index numbers two different sections '5.'

ADDITIONAL REGULATION THIS RECORD DOES NOT CONTAIN, named by the source page itself

SUPERSEDED CONTACT PRESERVED: Watford City Deer Management Access Permit

unit 1 -- deer types with NO printed figure

unit 2A -- deer types with NO printed figure

unit 2B -- deer types with NO printed figure

unit 2C -- deer types with NO printed figure

unit 2D -- deer types with NO printed figure

unit 2F1 -- deer types with NO printed figure

unit 2F2 -- deer types with NO printed figure

unit 2G -- deer types with NO printed figure

unit 2G1 -- deer types with NO printed figure

unit 2G2 -- deer types with NO printed figure

unit 2E -- deer types with NO printed figure

unit 2H -- deer types with NO printed figure

unit 2I -- deer types with NO printed figure

unit 2J1 -- deer types with NO printed figure

unit 2J2 -- deer types with NO printed figure

unit 2K1 -- deer types with NO printed figure

unit 2K2 -- deer types with NO printed figure

unit 3A1 -- deer types with NO printed figure

unit 3A3 -- deer types with NO printed figure

unit 2L -- deer types with NO printed figure

unit 3A2 -- deer types with NO printed figure

unit 3A4 -- deer types with NO printed figure

unit 3B3 -- deer types with NO printed figure

unit 3C -- deer types with NO printed figure

unit 3D1 -- deer types with NO printed figure

unit 3D2 -- deer types with NO printed figure

unit 3E1 -- deer types with NO printed figure

unit 3E2 -- deer types with NO printed figure

unit 3F1 -- deer types with NO printed figure

unit 3F2 -- deer types with NO printed figure

unit 4A -- deer types with NO printed figure

unit 4B -- deer types with NO printed figure

unit 4C -- deer types with NO printed figure

unit 4D -- deer types with NO printed figure

unit 4E -- deer types with NO printed figure

unit 4F -- deer types with NO printed figure

PUBLISHER CHECKSUM, recomputed this run

QUOTA TABLE CROSS-CHECK: text route against two blind pixel routes, cell by cell

BLIND-READING VERBATIM COVERAGE, per section of the proclamation -- which paragraphs BOTH readings reproduce in full, and which they do not

SOURCE CONTRADICTION: the signed scan gives the Deer Gun season two different sets of dates, and this HTML reproduces only one of them

Figure 1 (caption only)

Figure 2 (caption only)

Figure 3 (caption only)

Figure 4 (caption only)

Figure 5 (caption only)

Figure 6 (caption only)

Figure 7 (caption only)

rule 81

HOURS OF HUNTING are one-half (½) hour before sunrise to one-half (½) hour after sunset except when seasons open as noted above. All season openers are Central Time.

Western half of boundary between units 2F1 and 2L – Eddy County Highway 1 heading east from its junction with U.S. Highway 281 at Sheyenne to its junction with ND Highway 20.

Boundary between units 3A3 and 3B2 – North shore of Lake Sakakawea. Except for Mathews (de Trobriand) Island and Mallard Island; all islands are part of unit 3B2. Mathews (de Trobriand) Island is part of unit 3A3. Mallard Island is part of unit 3B3.

Boundary between units 3B2 and 3B3 – South shore of Lake Sakakawea heading west from U.S. Highway 83 to a point on the shore directly north of where ND Highway 200 turns southward (west of Pick City), then south to ND Highway 200. Mallard Island is part of unit 3B3.

Boundary between units 4B and 4C – The Little Missouri River.

Boundary between units 4D and 4E – The northern Slope County line heading west from its junction with U.S. Highway 85 to the Little Missouri River, then south on the Little Missouri River to U.S. Highway 12.

Quadriplegics who apply for and receive a Deer Gun season license through the lottery will be eligible to hunt during the Youth Deer season. Quadriplegics who hunt during the Youth Deer season must adhere to the unit and deer-type restrictions on their license.

Deer Gun, Youth Deer, Muzzleloader and Deer Bow hunters will have their license number printed on their certificate. It is illegal to purchase more than one regular Deer Bow license. Licenses are not transferable. No person may give, barter, or sell an issued license to another person.

Lottery application deadline will be determined by the Director and specified at the time of the lottery application period opening.

Nonresident Deer Bow – General nonresident bow licenses for white-tailed deer must be purchased online, either through the Department website, at Department District Offices or authorized license vendors. Once purchased, bow licenses are printed and mailed from the Department’s Bismarck office.

Muzzleloader – Licenses are issued by a weighted lottery drawing among eligible applicants.

Fees – Applicants must submit a payment type with their Deer Gun, Youth Deer, and or Muzzleloader applications. License fees will only be collected if the applicant is successful in the lottery.

Gratis and Nonresident Landowner Preferential Licenses – Gratis and nonresident preferential licenses are issued only through the Department's Bismarck office and the number of gratis licenses issued will be subtracted from the maximum number of Deer Gun licenses available for each unit.

A limited number of access permits will be issued for hunting in the Minot Deer Management Zone. After obtaining a permit, hunters must purchase a Special Deer-Bow Herd Reduction license online at the Game and Fish Department’s website at gf.nd.gov.

Restrictions including quantity of access permits, will be determined by the City of Watford City. After obtaining a permit, hunters must purchase Special Deer-Bow Herd Reduction licenses online at the Game and Fish Department’s website at gf.nd.gov.

*Note: This was changed in the html version of the proclamation after signing. Please use the Watford City Police Department number for Watford City Deer Management Access Permits.

Portions of the City of Portland and adjacent areas referred to as the Portland Deer Management Zone (Figure 6). Hunters who desire to hunt within the described Portland Deer Management Zone must first obtain a Portland Deer Management Access Permit from the City of Portland (701-788-2463).

Other Firearm and Archery Equipment Restrictions – It is illegal to go afield with a firearm or bow and arrow while intoxicated.

cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 93%, reading B 3% (longest run shared with A 47 characters, with B 4). NO FULL VERBATIM MATCH IN EITHER READING. The two figures are published and no verdict is drawn from them. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
penalty_preamble	The following violations have the listed penalties:
section	15. Violations and Penalties
value	$100. In excess of 30 minutes–Class B misdemeanor.
verbatim	Hunting before or after legal hours (Section 1) up to 30 minutes--penalty--$100. In excess of 30 minutes–Class B misdemeanor.
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.

cross_check	VALUE-LEVEL, against both blind visual readings of the signed scan: reading A agrees (4A = 15, recorded under its own key 'Youth Deer "I Tag" - Ages 14 and 15, any deer in specific units, by lottery (p.11)'); reading B agrees (4A = 15, recorded under its own key "Youth Deer 'I Tag' Licenses Available (Ages 14 and 15, any deer, specific units)"). This row is NOT cross-checked by string overlap. The Department's text prints these six units in the order 4A, 4D, 4B, 4E, 4C, 4F and both readers recorded them in the order 4A, 4B, 4C, 4D, 4E, 4F, so a verbatim substring measure scored these twelve rows at 28% and 17% while both readers had in fact recorded every figure exactly. Re-ordering defeats string overlap and does not defeat a value comparison, so the value comparison is what is published here.
hunting_unit	4A
licence_class	Youth Deer “I Tag”
section	11. Licenses Available by Season
value	15
verbatim	“I Tag” Licenses Available 4A – 15, 4D – 25, 4B – 30, 4E – 15, 4C – 20, 4F – 15
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.

cross_check	CELL-LEVEL, against both blind visual readings of the signed scan: reading A agrees; reading B agrees.
deer_type	Type A
deer_type_description	Any Antlered Deer - either species
hunting_unit	1
management_unit	Turtle Mtns
section	11. Licenses Available by Season (Deer Gun table)
value	300
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.

cross_check	Recomputed from the 112 printed cells this run: Type A: computed 21,250 vs printed 21,250 OK. Reading A recorded the TOTALS row as 'A 21,250; B 13,400; C 2,000; D 850; E 1,200; F 400' and reading B recorded each total at the foot of its own type list.
deer_type	Type A
hunting_unit	TOTALS (all units)
section	11. Licenses Available by Season (Deer Gun table)
value	21,250
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.

attached_to	Deer Bow (Nonresidents) -- the asterisk printed in the Opens cell, 'Sept. 4* (12 noon)'
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 100%, reading B 100% (longest run shared with A 272 characters, with B 272). VERBATIM IN BOTH: every character of this row's text appears verbatim in both independent blind visual readings of the signed scan. That is the strongest corroboration available for this state. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
restrictions_it_qualifies	Nonresidents – Restricted to type of deer described on license. Orange clothing required of all bow hunters and those accompanying apprentice license holders during Deer Gun season.
section	1. Season Dates, Times, Open Areas
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.

closure_applies_to	all deer hunting, on federal or state properties such as refuges, sanctuaries, military installations, parks or historic sites that are posted no trespassing or no hunting, unless specifically authorized. Section 9's five following paragraphs state named exceptions that OPEN portions of Grahams Island, Fort Stevenson, Fort Ransom and Lake Sakakawea State Parks and the Head of the Mountain Nature Preserve; they are published as rule rows and this closure must be read with them.
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 100%, reading B 100% (longest run shared with A 203 characters, with B 203). VERBATIM IN BOTH: every character of this row's text appears verbatim in both independent blind visual readings of the signed scan. That is the strongest corroboration available for this state. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
section	9. Areas Closed to Hunting
value	Unless specifically authorized, federal or state properties such as refuges, sanctuaries, military installations, parks, or historic sites posted no trespassing or no hunting are closed to the hunting of deer.
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
why_this_is_a_closure_and_section_4_is_not	This provision stops hunting -- of everyone, or of a named class of hunter, on named ground. Section 4's seven 'Areas Closed to Hunting with Centerfire Rifles' do not: those areas stay open and only a class of firearm is barred, so they are published as rule rows carrying the weapon restriction. Until this run the classification was the other way round and this state's closure rows contained no closures.

closure_applies_to	the Deer Gun and Muzzleloader seasons, in the described portion of unit 2L. The boundary is given in the paragraph itself as metes and bounds and is published verbatim; no boundary is derived here.
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 35%, reading B 88% (longest run shared with A 89 characters, with B 590). NO FULL VERBATIM MATCH IN EITHER READING. The two figures are published and no verdict is drawn from them. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
section	9. Areas Closed to Hunting
value	Also closed for the Deer Gun and Muzzleloader seasons is that portion of unit 2L starting at the junction of U.S. Highway 2 and ND Highway 20, then south along ND Highway 20 and ND Highway 57 to the north shore of Devils Lake, then northwest along the north shore of Devils Lake to the west edge of South Creel Township, then north along the township line to ND Highway 19, then east along ND Highway 19 to its junction with U.S. Highway 2, then southeast along U.S. Highway 2 to the point of origin, except Camp Grafton where a limited number of special permits will be issued to disabled veterans. Check with North Dakota National Guard headquarters at Camp Grafton for details (701-662-0200).
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
why_this_is_a_closure_and_section_4_is_not	This provision stops hunting -- of everyone, or of a named class of hunter, on named ground. Section 4's seven 'Areas Closed to Hunting with Centerfire Rifles' do not: those areas stay open and only a class of firearm is barred, so they are published as rule rows carrying the weapon restriction. Until this run the classification was the other way round and this state's closure rows contained no closures.

closure_applies_to	NONRESIDENTS ONLY, and only from October 10, 2026 through October 16, 2026. Residents are not affected and the areas are otherwise open. The paragraph states its own exception: nonresidents hunting on PLOTS land they own.
cross_check	DISTINCTIVE VERBATIM OVERLAP WITH THE TWO BLIND READINGS OF THE SIGNED SCAN: reading A 100%, reading B 100% (longest run shared with A 297 characters, with B 297). VERBATIM IN BOTH: every character of this row's text appears verbatim in both independent blind visual readings of the signed scan. That is the strongest corroboration available for this state. WHAT IS MEASURED: the fraction of this row's folded characters lying in a verbatim run that also appears in that blind reading's transcript, counting only runs that occur exactly once in the whole proclamation text so that shared boilerplate ('4A', 'Deer Bow', a section number) cannot inflate it. IT IS VERBATIM STRING OVERLAP AND NOTHING ELSE. It is a lower bound on what a reader saw: a reader who paraphrased, summarised or re-ordered a provision read it and still scores low. The seven Section 4 centerfire-rifle rows are the measured demonstration -- reading A recorded all seven as summaries rather than transcriptions, their figures spread across most of the range, and each of those rows carries reading A's own record of the area beside its figure. A figure below 100% is NOT a finding that the reader did not see this provision and this field makes no such claim. Where an exact comparison exists it is used instead of this one: the 112 quota cells, the six column totals and the 12 Youth Deer tag quotas are compared value by value. Which paragraphs neither reading reproduces is answered per SECTION, not per row -- see the 'BLIND-READING VERBATIM COVERAGE' scope row.
section	9. Areas Closed to Hunting
value	In accordance with N.D.C.C. 20.1-08-04.9, nonresidents shall be prohibited from hunting from October 10, 2026 through October 16, 2026 on Game and Fish Department wildlife management areas or Conservation PLOTS (Private Land Open To Sportsmen) areas. Exception: Nonresidents hunting on PLOTS land they own.
verification	TEXT ROUTE: the North Dakota Game and Fish Department's own HTML text of the proclamation. The SIGNED instrument is a 22-page scan with ZERO text characters, so core.library_divergence -- the second-PDF-library check every other state in this compilation passes -- CANNOT RUN for North Dakota. In its place this row was compared against the signed scan as read by TWO INDEPENDENT BLIND VISUAL READERS (sb/north-dakota-deer-2026-reading-a.json, -reading-b.json), and this row's OWN result is in cross_check -- read it, because it is not the same check on every row. Quota cells, the column totals and the Youth Deer tag quotas are compared VALUE BY VALUE, which is exact. Prose rows carry a VERBATIM STRING OVERLAP figure against each reading, which is a lower bound on what that reader saw and not a measure of whether the reader saw it; the cross_check field on each such row states what its figure does and does not establish. Verified differently from the other 49 states.
why_this_is_a_closure_and_section_4_is_not	This provision stops hunting -- of everyone, or of a named class of hunter, on named ground. Section 4's seven 'Areas Closed to Hunting with Centerfire Rifles' do not: those areas stay open and only a class of firearm is barred, so they are published as rule rows carrying the weapon restriction. Until this run the classification was the other way round and this state's closure rows contained no closures.