2020-21 Donruss Optic Premium Box Set

• G-001 SHARED BASE CHECKLIST / BASE-CARD IDENTITY COLLISION between TWO releases, per EXTRACTION_SPEC 5.2. This release (opticpremium-bkb-2020-21) and data/optic-bkb-2020-21 describe THE SAME 200 physical base card slots. Measured, not estimated: a mechanical positional row-by-row diff of this base_cards.csv against data/optic-bkb-2020-21/base_cards.csv matches on 200 of 200 card numbers, 200 of 200 player names and 200 of 200 teams - a complete overlap, zero differences on any field, and the rookie_flag split is the same 150 no / 50 yes. The Premium Box Set is therefore the FULL parent base run on premium stock, not a subset or a re-selection, and the 201 in "201-card factory set" is 200 base + 1 autograph, not 201 base slots. tcdb catalogues it as a PARALLEL of the Optic base rather than as a set of its own (subset sid 281858 under the 2020-21 Donruss Optic parent sid 264676, https://www.tcdb.com/ViewSet.cfm/sid/281858/2020-21-Donruss-Optic---Premium-Box-Set), and checklistcenter lists it under "Box Set Parallels" as the single line "Premium Set #/249" (https://www.checklistcenter.com/2020-21-donruss-optic-basketball-card-checklist/). This repeats the identical 200-of-200 finding recorded for the prior season in data/opticpremium-bkb-2019-20 gap G-001. Built here as instructed and NOT deduplicated and NOT double-counted; the duplication is deliberate and must be resolved at catalogue level, not by deleting either file.
• G-002 TWO DIFFERENT QUANTITIES ON THIS PRODUCT ARE BOTH 249 AND THEY ARE IN DIFFERENT UNITS (EXTRACTION_SPEC 5.4). (a) PER-CARD SERIAL RUN = 249: "200 cards. Serial numbered #/249. Full set per box." (https://www.cardboardconnection.com/2020-21-donruss-optic-premium-box-set-basketball-cards); "Each Card #'d /249!" (https://www.dacardworld.com/sports-cards/2020-21-panini-donruss-optic-basketball-hobby-premium-box-set); "Premium Set #/249" (https://www.checklistcenter.com/2020-21-donruss-optic-basketball-card-checklist/); tcdb flag "SN249" on every base row. This is what base_print_run=249 records. (b) NUMBER OF BOXES/SETS PRODUCED = 249: "Offered straight from the Panini website, the exclusive format is limited to just 249 boxes." (https://www.cardboardconnection.com/2020-21-donruss-optic-premium-box-set-basketball-cards) and "limited to only 249 boxes" (https://cardlines.com/2020-21-donruss-optic-premium-box-set-basketball-review/). Quantity (b) is NOT written in any CSV column - there is no column for it - and was NOT used to derive (a). Recorded because the coincidence makes the release row look self-confirming when it is not: a later reader must not take the box count as corroboration of the per-card run, nor the reverse. The prior season of this line (opticpremium-bkb-2019-20) also wrote base_print_run=249 and its gap G-002 recorded that the set count there was UNKNOWN; for 2020-21 the box count IS stated, at 249.
• G-003 release_date 2021-10-15 rests on two sources that may not be independent. cardboardconnection prints "Release Date: October 15, 2021" (https://www.cardboardconnection.com/2020-21-donruss-optic-premium-box-set-basketball-cards), labelled Release Date and not Estimated Release Date - which is a stronger label than the prior season carried. cardlines corroborates in prose: "The Optic Premium Box Set was sold on October 15 through Panini's website" (https://cardlines.com/2020-21-donruss-optic-premium-box-set-basketball-review/). tcdb carries no release date for the subset and Beckett was not fetched. Aggregator sites in this space are known to copy one another (EXTRACTION_SPEC 5.4), so these two may be one opinion.
• G-004 parallels.csv is deliberately EMPTY (header only, zero data rows), per EXTRACTION_SPEC 2.3. No parallel OF THE PREMIUM BOX SET BASE RUN was found on any source, and the absence is evidenced rather than unresolved. (a) On tcdb the Premium Box Set is a single subset, sid 281858, with no colour siblings: its immediate neighbours in the alphabetically-ordered parallel block are sid 281857 "2020-21 Donruss Optic - Pink Velocity" and sid 281859 "2020-21 Donruss Optic - Purple Stars 1st Off the Line", so no "Premium Box Set <colour>" subset exists between them; the sixteen unnumbered base-parallel subsets sids 281722-281737 were enumerated individually (Blue Pulsar, Blue Velocity, Checkerboard, Choice Dragon, Choice Red and Green, Fanatics, Fast Break Holo, Gold Wave, Holo, Hyper Pink, Photon, Purple, Purple Shock, Red Pulsar, Red Wave, Pulsar) and none carries the Premium name. (b) checklistcenter's Optic page lists the whole tier as one line, "Premium Set #/249", under "Box Set Parallels" (https://www.checklistcenter.com/2020-21-donruss-optic-basketball-card-checklist/). (c) cardboardconnection's product page carries no parallel list at all (https://www.cardboardconnection.com/2020-21-donruss-optic-premium-box-set-basketball-cards). The prior season of this line reached the same result (opticpremium-bkb-2019-20 gap G-004). CAVEAT ON COMPLETENESS: tcdb's own full parallel inventory at https://www.tcdb.com/Inserts.cfm/sid/264676/2020-21-Donruss-Optic could not be read end to end - the fetch layer truncates that page inside the Insert Sets section and never reaches the Parallel Sets heading - so (a) rests on individually enumerated sids rather than on one rendered list.
• G-005 The autograph set's card numbers duplicate base card numbers WITHIN this release. Rated Rookies Signatures Premium Box Set renders 25 rows at numbers 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 165, 166, 167, 168, 169, 170, 171, 172, 174, 175, 176 and 180 (https://www.tcdb.com/Checklist.cfm/sid/281712/2020-21-Donruss-Optic---Rated-Rookies-Signatures-Premium-Box-Set, header "Total Cards: 25", every row flagged "AU, SN10") - 164, 173, 177, 178 and 179 are not issued in the autograph set. Every one of those 25 numbers is also a base card number 1-200 in this release's base_cards.csv. Any card-number-keyed join across this release's sets will collide unless it is also keyed by set. Flagged now because the card-level autograph checklist belongs to a later pass (CHASE_CHECKLIST_SPEC).
• G-006 Structural divergence between our release-level model and every source, unchanged from the prior season. Neither tcdb nor checklistcenter treats this product as a release of its own: tcdb has NO parent sid for it - ViewAll.cfm/sp/Basketball/year/2020 (https://www.tcdb.com/ViewAll.cfm/sp/Basketball/year/2020) lists exactly two Optic entries for this season, "2020-21 Donruss Optic" sid 264676 and "2020-21 Panini Contenders Optic" sid 287620, and no Premium Box Set entry - and the product exists only as two subsets of the Optic parent (base sid 281858, autograph sid 281712). checklistcenter carries it as one line inside the Optic checklist page. Only cardboardconnection and cardlines give it a page of its own. Everything in this release therefore comes from sources that consider it part of another product; no independent product-level structural source exists.
• G-008 rookie_flag basis is tcdb's "RR" flag, not an "RC" designation. EXTRACTION_SPEC 2.4 says to write yes/no "only where the source explicitly shows an RC designation". tcdb prints "RR" on every row of the 151-200 block on this subset and prints nothing of the kind on rows 1-150 - read off https://www.tcdb.com/Checklist.cfm/sid/281858/2020-21-Donruss-Optic---Premium-Box-Set?PageIndex=2 for 151-175 and Gallery.cfm PageIndex 18, 19 and 20 for 171-200. rookie_flag=yes was written for 151-200 and no for 1-150 on that flag, which also matches what data/optic-bkb-2020-21/base_cards.csv carries for the identical checklist (150 no, 50 yes) and what data/opticpremium-bkb-2019-20 did for its own season (its gap G-010). If the spec means RC strictly and RR does not count, all 200 rows here should be UNKNOWN and at least four releases need the same correction.
• G-009 base_set_size is written as 200 while the most quotable headline figure on both dedicated product pages is 201: cardboardconnection "Product Configuration: 201-card factory set" (https://www.cardboardconnection.com/2020-21-donruss-optic-premium-box-set-basketball-cards) and DA Card World "contains 201 cards per box" (https://www.dacardworld.com/sports-cards/2020-21-panini-donruss-optic-basketball-hobby-premium-box-set). 200 was written because the same cardboardconnection page decomposes the 201 as "Full base set (200 cards)" plus "1 Rated Rookie Autograph", because the same DA Card World page decomposes it as "1 rated rookies signature premium box set auto (#'d/10) and 1 200 card complete set of Donruss Optic Premium Box Set (#'d/249) per box", because tcdb's subset header reads "Total Cards: 200" and its Gallery ends at card 200 on page 20 of 20, and because EXTRACTION_SPEC 5.3 requires the autograph block to be excluded. Flagged because any later source quoting "201 cards" will look like a coverage hole against this file and is not one.
• G-010 NAME DRIFT: the drift to "Factory Set" wording BEGAN EARLIER THAN THE SCOPE MEMO STATES, though not on the sources that decide the product name. For 2020-21 both structural sources still say Premium Box Set - tcdb "2020-21 Donruss Optic - Premium Box Set" (https://www.tcdb.com/ViewSet.cfm/sid/281858/2020-21-Donruss-Optic---Premium-Box-Set) and cardboardconnection H1 "2020-21 Donruss Optic Premium Box Set Basketball Cards Checklist" (https://www.cardboardconnection.com/2020-21-donruss-optic-premium-box-set-basketball-cards) - and product_name was written accordingly. BUT a major retailer already titled this same 2020-21 product "2020/21 Panini Donruss Optic Basketball Factory Set Hobby Premium (Box) /249" (https://www.dacardworld.com/sports-cards/2020-21-panini-donruss-optic-basketball-hobby-premium-box-set) while its own body text calls the contents "Donruss Optic Premium Box Set". Separately, the later-season rename is real but is not universal: checklistinsider titles the 2023-24 and 2024-25 editions "Donruss Optic Premium Basketball Factory Set" (https://www.checklistinsider.com/2023-24-donruss-optic-premium-basketball-factory-set, https://www.checklistinsider.com/2024-25-donruss-optic-premium-basketball-factory-set, the former fetched and confirmed, "251-card factory set", "/249", release "July 15, 2024"), while Beckett still titles the 2024-25 edition "2024-25 Donruss Optic Basketball Premium Box Set" (https://www.beckett.com/news/2024-25-donruss-optic-basketball-premium-box-set/, not fetched - search-result title only). So from 2023-24 the two names run CONCURRENTLY across sites rather than one replacing the other, and a future season's product_name decision must not be taken from a single site. Raised, not decided - EXTRACTION_SPEC 5.1 says report, do not choose.
• G-011 LINE SPAN CORRECTION. The task scope states this line ran through 2024-25; the evidence supports that end date but shows the line starts EARLIER than the catalogue's earliest opticpremium release (2019-20). tcdb carries "2017-18 Donruss Optic - Premium Box Set" at sid 171625 (https://www.tcdb.com/Checklist.cfm/sid/171625/2017-18-Donruss-Optic---Premium-Box-Set, surfaced as an indexed page title; not fetched in this pass). data/ contains no opticpremium release before 2019-20. Recorded so a backfill pass knows the line predates the current earliest release; the 2017-18 and 2018-19 editions were NOT investigated here and nothing about them was written.
• G-012 CONTRADICTION AGAINST EXISTING DATA, across the two seasons of the sibling Optic line. data/optic-bkb-2019-20/parallels.csv carries a row for the Premium Box Set as a base parallel of the parent product (par_01M19N6THK40T9E6FQ52MB2D5R, "Premium Box Set Base", print_run 249, with its own gap G-011), but data/optic-bkb-2020-21/parallels.csv carries NO equivalent row for the 2020-21 Premium Box Set even though tcdb and checklistcenter both catalogue it exactly the same way in both seasons (tcdb subset under the Optic parent; checklistcenter under "Box Set Parallels"). So the identical real-world object is a parallel row in one season of optic-bkb and absent in the next. Nothing outside this release directory was touched. Recorded for a repair pass to decide one convention; note that whichever way it goes it interacts with G-001 above, since the object is simultaneously this release's whole base set.
• G-013 license_status was NOT specified in the task prompt, which EXTRACTION_SPEC 2.1 says is where it comes from. "licensed (NBA/NBPA)" was written to match this line's own prior season (data/opticpremium-bkb-2019-20). That DISAGREES with the same-season sibling data/optic-bkb-2020-21, which carries "licensed (NBA)" for what is physically the same base checklist. Both are live values per the spec; no source was fetched that states either wording, so this is a catalogue-convention choice, not an evidenced one, and it is flagged rather than silently picked. licensed=yes and league=NBA are unaffected and were given in the task prompt.
• G-014 Single-source and unresolved items, listed so they are not mistaken for settled values. (a) The base checklist's only card-level structural source is tcdb sid 281858; no aggregator publishes a card-by-card Premium Box Set checklist, so the roster is SINGLE-SOURCE at card level even though the set size and the print run each have three or four sources. Positional corroboration exists only through the identity with data/optic-bkb-2020-21 (G-001), which is not an independent source. (b) Marketplace corroboration was obtained for two individual cards only, as search-result titles: "2020-21 Panini Donruss Optic Premium Box Set Prizm /249 Mitchell Robinson #99" (ebay.com/itm/234295332626) and "2020-21 Donruss Optic Premium Box Set #151 Anthony Edwards rookie card" (goldin.co) - both agree with base_cards.csv at those numbers (99 Mitchell Robinson, 151 Anthony Edwards). Neither listing page was fetched. (c) scarcity_tier for the autograph set is UNKNOWN - no source states a tier - and its stated_odds "1 per box" rests on one box-contents bullet, "1 Rated Rookie Autograph". (d) The base cards' finish is described as "an exclusive Kaleidoscope finish" by cardboardconnection (https://www.cardboardconnection.com/2020-21-donruss-optic-premium-box-set-basketball-cards) while marketplace titles for the same cards say "Scope", "Premium Scope" and "Prizm Silver Scope"; the finish name was not resolved and is written nowhere. (e) cardboardconnection and checklistcenter are known to copy each other (EXTRACTION_SPEC 5.4) and Beckett was not used, so where both are cited they may be one opinion; the genuinely independent corroboration for the print run is DA Card World plus the tcdb per-row SN249 flags.
• G-015 METHOD / SPEC DEFECT recorded for reproducibility, since it changes the documented fetch plan. (a) EXTRACTION_SPEC 3.2 says the slug is unvalidated on ViewSet.cfm and validated on Gallery.cfm, and says nothing about Checklist.cfm. Measured this pass: Checklist.cfm VALIDATES the slug too - https://www.tcdb.com/Checklist.cfm/sid/281727 (no slug) and https://www.tcdb.com/Checklist.cfm/sid/281727/x (wrong slug) both returned 404 pages, while the canonical slug on the same sid rendered normally. ViewSet.cfm/sid/<N>/x rendered correctly on every sid tried and was therefore the only usable tool for discovering an unknown sid. (b) EXTRACTION_SPEC 3.2 says Checklist.cfm?PageIndex=N returns 100 rows per page. The SERVER does paginate at 100 (page 1 = cards 1-100, page 2 = 101-200), but the summarizing fetch layer truncates each page at roughly row 74 of page 1 and row 171 of page 2, and it also REFUSES a request to reproduce a whole page (it answered "I need to respectfully decline reproducing the entire checklist table"). Asking for a bounded range of about 30 rows at a time succeeds. The workable budget for a 200-card run was therefore 4 bounded Checklist asks plus 6 Gallery pages, not 2 fetches. (c) tcdb's Inserts.cfm for a parent with a large insert inventory (https://www.tcdb.com/Inserts.cfm/sid/264676/2020-21-Donruss-Optic) is truncated by the fetch layer before the Parallel Sets heading, so the spec's "Inserts.cfm is your sets.csv backbone" does not hold for base parallels on a product this size; sid discovery had to be done by probing ViewSet.cfm across a sid range.

If you know a source that answers one of these, tell us — name the question's label and where to look.