2016-17 Grand Reserve

• G-003 base_print_run=unnumbered CARRIES basis=assumed, NOT sourced, AND THE DISTINCTION IS THE WHOLE POINT. No source consulted states in words whether 2016-17 Panini Grand Reserve base cards carry a serial number: cardboardconnection's base-set block says only "100 cards." and "PARALLEL CARDS: Vintage." (https://www.cardboardconnection.com/2016-17-panini-grand-reserve-basketball), Panini's own blog describes the hits and not the base cards (https://blog.paniniamerica.net/the-panini-america-quality-control-gallery-2016-17-grand-reserve-basketball/), and blowoutbuzz says nothing either way (https://blowoutbuzz.wordpress.com/2017/05/16/first-buzz-2016-17-panini-grand-reserve-basketball-cards/). What supports the value is a WITHIN-PRODUCT POSITIVE DISCRIMINATIVE TEST RUN ON THREE RENDERERS: Checklist.cfm, ViewSet.cfm and Gallery.cfm with &sort=0 each render AU, MEM and the serial tokens SN35/SN40/SN49/SN75/SN99 on the rows of this product's own sid 154902 Cornerstones, while the same three render no SN or PR token on any of the parent's 100 base rows - and the parent's flag column is demonstrably alive there, because it renders RC on cards 1, 21, 45 and 50. Per EXTRACTION_SPEC 2.1 that is good evidence and still not a source stating it. RESOLVE BY: a Panini sell sheet, a first-look blog post that describes the base cards, or a marketplace listing for a base card that states the numbering in words.
• G-004 rookie_flag IS `yes` ON FOUR ROWS AND `UNKNOWN` ON THE OTHER 96, AND THE CATALOGUE IS SPLIT ON THIS CONVENTION. tcdb renders the RC token on exactly four base rows - 1 Ben Simmons, 21 Malcolm Delaney, 45 Yogi Ferrell, 50 Sean Kilpatrick - read on https://www.tcdb.com/Checklist.cfm/sid/152590/2016-17-Panini-Grand-Reserve?PageIndex=1 and reproduced under ?sort=0. EXTRACTION_SPEC 2.4 permits yes/no only where a source explicitly shows an RC designation, and CAV-003 says absence of an RC token is not a designation, so the other 96 rows are UNKNOWN. I can see that the four are internally consistent with the roster - no other 2016-17 NBA rookie appears in the checklist - but establishing that requires reading players' draft years, which is derivation, not sourcing. THE DIVERGENCE, STATED PLAINLY BECAUSE IT IS A CONTRADICTION INSIDE OUR OWN DATA: of the 22 releases already in data/ for this season, fifteen write `no` on their non-RC base rows (absolute, donruss, goldstandard, limited, nattreasures, noir, optic, paninihoops, prestige, prizm, revolution, select, threads, totallycertified) and seven write UNKNOWN (courtkings, excalibur, flawless, immaculate, impeccable, preferred, spectra). Nothing on either kind of row tells a reader which convention it is under. RESOLVE BY: an owner decision on whether absence of a tcdb RC token may be written as `no`, applied in one pass over the season.
• G-005 PRODUCT-NAME DIVERGENCE: PANINI'S OWN BLOG DROPS "Panini" FROM THE PRODUCT NAME AND USES A SINGLE-YEAR FORM. reference/brands.csv carries product_name "Panini Grand Reserve Basketball" and that is written verbatim in release.csv; it matches the season-independent form of tcdb's heading "2016-17 Panini Grand Reserve" (https://www.tcdb.com/ViewSet.cfm/sid/152590/2016-17-Panini-Grand-Reserve) and of cardboardconnection's H1 "2016-17 Panini Grand Reserve Basketball Cards" (https://www.cardboardconnection.com/2016-17-panini-grand-reserve-basketball). But the first-party post uses BOTH "2017 Grand Reserve Basketball" and "2016-17 Grand Reserve Basketball" in one page, with no "Panini" in either (https://blog.paniniamerica.net/the-panini-america-quality-control-gallery-2016-17-grand-reserve-basketball/), asked and answered directly. This is the same house-style drift the ledger already records for Donruss and Optic, except that here it is the manufacturer doing the dropping rather than an editorial site, which is a new location for it. NOTHING WAS MINTED, brands.csv and pipeline/brands.py were not edited, and no rename is proposed. RESOLVE BY: noting it in the ledger's account of name drift.
• G-006 release_date=2017-07-19 HAS TWO SOURCES PRINTING THE SAME FIGURE AND THEIR INDEPENDENCE IS UNTESTED. tcdb's parent renders the UNQUALIFIED line "Release Date: July 19, 2017" and the page was asked directly whether the word "Estimated" appears anywhere on it, answering no (https://www.tcdb.com/ViewSet.cfm/sid/152590/2016-17-Panini-Grand-Reserve). cardboardconnection's PRODUCT page - not its season index, per scope memo section 1a - renders "Release Date: 7/19/2017" (https://www.cardboardconnection.com/2016-17-panini-grand-reserve-basketball). Per scope memo section 7.6 the discriminator for whether two sources are one lineage is PUBLICATION DISTANCE, and I did not measure it for either; both plausibly descend from Panini's sell sheet, so this should be read as one figure with two renderings rather than as corroboration. TWO FORECAST DATES WERE ALSO FOUND AND NEITHER IS RECORDED AS A COMPETING RELEASE DATE: blowoutbuzz published "Arrives: June 28" on 2017-05-16, about six weeks ahead and three weeks wrong (https://blowoutbuzz.wordpress.com/2017/05/16/first-buzz-2016-17-panini-grand-reserve-basketball-cards/); Panini's own blog published "set to arrive in hobby shops across North America next Wednesday" on 2017-07-13, which resolves to 2017-07-19 and therefore AGREES, but a six-day-ahead forecast is still a forecast and is not counted as a second source (https://blog.paniniamerica.net/the-panini-america-quality-control-gallery-2016-17-grand-reserve-basketball/). BOTH SOURCES FILE THE PRODUCT AS 2016-17 despite the July 2017 date - tcdb's heading carries "2016-17" and cc's breadcrumb reads "2016-2017 Basketball Cards" - so the late date is a late release inside the season, not a mis-slotted product. RESOLVE BY: a dated retailer or distributor announcement independent of Panini's sell sheet.
• G-007 A POSITIVE DEBUT STATEMENT WAS FOUND, A 2016-17 BASKETBALL FLOOR IS CLAIMED FOR `grandreserve`, AND THE LIMIT OF THE CLAIM IS STATED HERE. specs/2016-17_SCOPE_MEMO.md section 3a marked this line "Probable debut - NOT PROVEN" because the only evidence was absence from tcdb's 2015-16 and 2014-15 indexes, which section 3b proves with Studio is not a floor. THE POSITIVE STATEMENT EXISTS AND IS FIRST-PARTY. blog.paniniamerica.net, published 2017-07-13, confirmed on two differently-parameterised fetches of the same URL: "It's not often that collectors have the opportunity to get in on the ground floor of a new release, but that's exactly the opportunity at hand for basketball collectors next week when Panini America unveils 2017 Grand Reserve Basketball." and "The newest addition to the company's NBA product portfolio boasts an eclectic mix of designs, card stocks, and technologies - not to mention plenty of on-card autographs." (https://blog.paniniamerica.net/the-panini-america-quality-control-gallery-2016-17-grand-reserve-basketball/). Two further lineages say the same in their own words: cardboardconnection, "The new brand gets underway with three hits per hobby box and this includes at least two total autographs." (https://www.cardboardconnection.com/2016-17-panini-grand-reserve-basketball), and blowoutbuzz, "This new brand delivers a hit in every pack and a new look and energy for the hardwood." (https://blowoutbuzz.wordpress.com/2017/05/16/first-buzz-2016-17-panini-grand-reserve-basketball-cards/). WHAT IS NOT ESTABLISHED: all three statements are scoped to the NBA/basketball portfolio and to this being a new release. NONE names a cross-sport ancestor the way cardboardconnection does for Impeccable ("first entered the hobby with the 2016 Panini Impeccable Football set") or Panini's blog does for Aficionado ("A baseball-only release that burst onto the scene with a one-season run in 1996"), and the brief warns that the ancestor was a different sport for both of those and for Vanguard. A web search for a Panini "Grand Reserve" football or baseball set returned nothing, which per the Studio lesson is evidence of nothing. What DOES exclude the Studio failure mode is the first-party wording itself: "ground floor of a new release" is incompatible with the eight-year-gap return Panini's own blog described for Studio. beckett.com was not opened (CAV-033); a beckett.com/news URL surfaced in search with the title "2016-17 Panini Grand Reserve Basketball to debut this summer", cited as a title only. RESOLVE BY: a first-party or editorial sentence addressing Grand Reserve in another sport, in either direction.
• G-008 TWO SETS RENDER ONE MORE ROW THAN THEIR OWN tcdb HEADER STATES, ON TWO ENDPOINTS EACH, AND THE UNIT CANNOT EXPLAIN IT. Difference Makers (sid 154866) states "Total Cards: 39" and renders 40 rows over 40 DISTINCT card numbers with no duplicate - the numbers 2-50 with 1, 6, 9, 28, 32, 34, 35, 37, 43 and 48 absent - identically on https://www.tcdb.com/Checklist.cfm/sid/154866/2016-17-Panini-Grand-Reserve-Difference-Makers?sort=0 and on https://www.tcdb.com/ViewSet.cfm/sid/154866/2016-17-Panini-Grand-Reserve-Difference-Makers , the second asked to list every card number in ascending order and to answer yes or no for cards 1, 6, 28 and 50 individually. Legendary Cornerstones (sid 154894) states "Total Cards: 19" and renders 20 rows over the CONTIGUOUS numbers 1-20, again identically on https://www.tcdb.com/Checklist.cfm/sid/154894/2016-17-Panini-Grand-Reserve-Legendary-Cornerstones?sort=0 and on https://www.tcdb.com/ViewSet.cfm/sid/154894/2016-17-Panini-Grand-Reserve-Legendary-Cornerstones , the second asked specifically whether a row for Alex English exists and answering "Yes, Card #20". In both cases the header counts one fewer than the distinct numbers rendered, so the usual unit explanations (distinct numbers versus rows, variation rows duplicating a number) do not apply - this is the CAV-044 shape where the header is short of its own object. card_count is my own count of the rendered rows in both. cardboardconnection states "39 cards." for Difference Makers, agreeing with the header; for Legendary Cornerstones it states nothing, because it folds the set into another (G-009). RESOLVE BY: a third enumeration, e.g. a complete Gallery.cfm page sweep of both sids, or a marketplace listing for a 40th Difference Makers card.
• G-009 cardboardconnection FOLDS TWO tcdb SETS INTO ONE AND ITS COUNT IS THE SUM, WHICH LOOKS LIKE A DISAGREEMENT AND IS NOT. cc lists "Cornerstones Quad Jersey Autographs" at "59 cards" (https://www.cardboardconnection.com/2016-17-panini-grand-reserve-basketball) while tcdb files Cornerstones (sid 154902, header 40, 40 rows numbered 1-40) and Legendary Cornerstones (sid 154894, header 19, 20 rows numbered 1-20) as separate sids with separate tier families. 40 + 19 = 59 exactly, using tcdb's OWN Legendary header rather than my enumerated 20 - so cc's 59 descends from the same 19 that G-008 shows is one short, and cc is NOT an independent confirmation of either figure. Panini's own copy supports the folding: it describes a single "Cornerstones Jersey Autographs insert that features veterans, rookies and legends" (https://blog.paniniamerica.net/the-panini-america-quality-control-gallery-2016-17-grand-reserve-basketball/), where "legends" is Legendary Cornerstones and "rookies" is Rookie Cornerstones. This release follows tcdb and files three separate sets, because that is the structure the sids and the tier families actually have. RESOLVE BY: nothing outstanding; recorded so a consumer reconciling against cc does not read a 19-card hole.
• G-010 channel=hobby ON ALL 24 PARALLEL ROWS IS A PRODUCT-LEVEL PRESENCE CLAIM, NOT A PER-TIER ONE, AND channel_basis SAYS SO. The evidence is Panini's own blog - "On average, each hobby box of 2017 Grand Reserve Basketball (eight cards per pack, three packs per box) will deliver at least two autographs and one additional autograph or memorabilia card." (https://blog.paniniamerica.net/the-panini-america-quality-control-gallery-2016-17-grand-reserve-basketball/) - with cardboardconnection giving a single Product Configuration, "8 cards per pack, 3 packs per box, 8 boxes per case", and "three hits per hobby box" (https://www.cardboardconnection.com/2016-17-panini-grand-reserve-basketball), and blowoutbuzz giving one SRP for the same pack (https://blowoutbuzz.wordpress.com/2017/05/16/first-buzz-2016-17-panini-grand-reserve-basketball-cards/). No source consulted names a retail, blaster, mega, factory-set or international SKU for this product, and NONE claims exclusivity, so `sourced-presence` is the strongest basis the evidence supports (CAV-010) and channel_confidence is `multi` because three sources place it in hobby boxes. I HAVE NOT SOURCED ANY TIER'S CHANNEL INDIVIDUALLY. The status-bkb-2017-18 lesson says declining to inherit a product-level hobby note onto tiers it does not cover is the right call; here the note covers the whole product because the product has exactly one documented SKU, so the value is written and this gap records that the support is product-level. RESOLVE BY: a per-tier distribution statement, or confirmation that no non-hobby SKU of this product exists.
• G-011 tcdb's Inserts.cfm CARRIES NO USABLE NOTES COLUMN ON THIS PARENT, SO EVERY PRINT RUN IN THIS RELEASE COST ITS OWN PER-SID ROW SWEEP. https://www.tcdb.com/Inserts.cfm/sid/152590/2016-17-Panini-Grand-Reserve was asked directly whether a column headed "Notes" exists: it does not. The right-hand cell on 42 of the 43 Insert Sets rows carries an image-completion percentage ("13.3% of images added" and the like), exactly one is EMPTY (Unbreakable, sid 154891), and the single Parallel Sets row carries a percentage too ("2.0% of images added" against Vintage). So the column is a percentage column, there is effectively one blank, and CAV-036's both-directions split test is UNRUNNABLE rather than merely failed - a blank here carries zero information about numbering, and the one blank sits against a set my own sweep shows carries no serial token anyway. This is a thirteenth parent for the CAV-036 tally and the modal worthless state again. NOTE for the caveat record: this parent has NO sibling season anywhere in the catalogue, so the season-reversal finding appended to CAV-036 on 2026-09-29 could not be tested here in either direction. RESOLVE BY: nothing; recorded as a measurement.
• G-012 THE PER-CARD VARIABLE RUN CENSUS - THE VALUES THE SCHEMA CANNOT HOLD. EXTRACTION_SPEC 2.3 forces print_run=UNKNOWN and serial_numbered=UNKNOWN on a tier that is PROVABLY serial numbered with a run that varies per card, which is indistinguishable in the CSV from having found nothing. Nine parallel rows in this release are in that state and this gap carries what was measured, tier by tier, so the fact survives. All were read over EVERY rendered row, never sampled. Cornerstones Granite sid 154903, 40 of 40 rows, tokens AU MEM and SN15 SN18 SN20 SN25. Cornerstones Quartz 154906, 40 of 40, AU MEM and SN25 SN30 SN49 (SN30 on cards 29, 32, 39). Grand Autographs Granite 154869, 40 of 40, AU and SN15 SN25. Legendary Cornerstones Granite 154895, 16 of 16, AU MEM and SN15 SN23 SN25 (SN23 on card 12). Legendary Cornerstones Marble 154896, 18 of 18, AU MEM and SN5 SN8 SN10 - sixteen rows SN10, card 11 SN5, card 17 SN8, confirmed row for row on Checklist.cfm?sort=0 and ViewSet.cfm. Legendary Cornerstones Quartz 154898, 19 of 19, AU MEM and SN15 SN25 SN35 SN49. Reserve Materials Marble 154882, 73 of 73, MEM and SN6 SN10 - seventy-two rows SN10, card 11 James Harden SN6, confirmed on both endpoints. Reserve Signatures Granite 154885, 49 of 49, AU and SN15 SN25. Rookie Cornerstones Marble 154910, 40 of 40, AU MEM and SN8 SN10 - thirty-nine rows SN10, card 134 Dario Saric SN8, confirmed on both endpoints. NINE FURTHER SETS ARE VARIABLE AT SET LEVEL, where the figure has nowhere to live either because sets.csv has no print_run column (see G-013): Cornerstones 40 rows AU MEM SN35 SN40 SN49 SN75 SN99; Difference Makers 40 rows AU SN10 SN25 SN35 SN75 SN99; Grand Autographs 40 rows AU SN35 SN49 SN99; Highly Revered 23 rows AU SN25 SN35 SN49 SN60 SN99; Legendary Cornerstones 20 rows AU MEM SN34 SN35 SN49 SN60 SN99; Reserve Signatures 50 rows AU SN20 SN25 SN35 SN49 SN75 SN99; Team Slogans 20 rows MEM SN1 SN2; The Ascent 39 rows AU SN25 SN35 SN75; Upper Tier Signatures 42 rows AU SN10 SN25 SN30 SN60 SN99. TWO METHOD POINTS WORTH THE RECORD. First, three of the nine parallel tiers differ on a SINGLE ROW out of 18, 40 and 73, so a sample of any size short of the whole tier writes a clean sourced-looking flat integer - the CAV-050 shape exactly. Second, and this is the sharper one, PANINI'S OWN FIRST-PARTY COPY WOULD HAVE WRITTEN TWO OF THOSE WRONG VALUES: it states "Marble (numbered to 10)" with no "or less" qualifier while attaching "or less" to base, Quartz and Granite (https://blog.paniniamerica.net/the-panini-america-quality-control-gallery-2016-17-grand-reserve-basketball/), and both Legendary Cornerstones Marble and Rookie Cornerstones Marble are provably not flat. Variability at the 1/1 end also reproduces here, against the clause CAV-050 refutes: Team Slogans carries SN1 on six rows and SN2 on fourteen, confirmed row for row on two endpoints, with blowoutbuzz independently stating the ceiling, "limited to no more than two copies" (https://blowoutbuzz.wordpress.com/2017/05/16/first-buzz-2016-17-panini-grand-reserve-basketball-cards/). RESOLVE BY: a print_run_basis column on parallels, or a sibling table holding per-card runs. Until then this is a schema problem, not missing research.
• G-013 SIX SETS ARE UNIFORMLY SERIAL NUMBERED AND sets.csv CANNOT HOLD THE FIGURE. EXTRACTION_SPEC 2.2 is explicit that there is no print_run column here and that the number must not be put in card_count, so these figures live only in each set row's notes, in words, and here. Every one was read over EVERY rendered row. Hickory sid 154872, 5 of 5 rows MEM SN39, flat 39, corroborated by cardboardconnection "5 cards. Serial numbered #/39." Local Legends 154875, 10 of 10 AU SN25, flat 25, corroborated "10 cards. Serial numbered #/25." Number Ones 154877, 10 of 10 AU SN1, flat 1/1, corroborated "Serial numbered 1/1." One-of-a-Kind 154878, 10 of 10 AU SN1, flat 1/1, corroborated "Serial numbered 1/1." Reserve Materials 154879, 75 of 75 MEM SN35, flat 35, confirmed on a second endpoint and corroborated "75 cards. Serial numbered #/35." Rookie Cornerstones 154907, 40 of 40 AU MEM SN99 over card numbers 101-140, flat 99, confirmed on a second endpoint and corroborated "40 cards. Serial numbered #/99." All cc quotes from https://www.cardboardconnection.com/2016-17-panini-grand-reserve-basketball. A reader filtering parallels.csv for these figures finds nothing, because the numbering is a property of the set and not of a parallel tier. RESOLVE BY: a print_run column on sets, or a documented convention for carrying a set-level run.
• G-014 CARD-NUMBER FINDING FOR THE CHASE PASS: Local Legends (sid 154875) RENDERS TWO NUMBERING SCHEMES IN ONE COLUMN. https://www.tcdb.com/Checklist.cfm/sid/154875/2016-17-Panini-Grand-Reserve-Local-Legends?sort=0 renders ten rows of which five carry bare ordinals (1 Larry Bird, 3 Allen Iverson, 5 Kobe Bryant, 6 Kevin Durant, 7 Stephen Curry) and five carry letter codes (LL-OR Oscar Robertson, LL-MJ Magic Johnson, LL-AD Anthony Davis, LL-JW John Wall, LL-PG Paul George). Per CHASE_CHECKLIST_SPEC a page mixing real codes and bare ordinals is the dominant premium pattern and a single real code settles the scheme for the set, so the five LL- codes are writeable and the five ordinals are placeholders to be withheld with the players named. This release writes no insert_cards.csv - that is a separate pass - so the finding is recorded here rather than acted on. The same sid's Onyx tier (154876) renders the identical mixed pattern on the identical players. RESOLVE BY: the chase pass, with a marketplace listing title for one of the five ordinal players.
• G-015 CARD-NUMBER FINDING FOR THE CHASE PASS: Startups (sid 154888) RENDERS THREE ROWS AS A BARE PREFIX WITH AN EMPTY SUFFIX. https://www.tcdb.com/Checklist.cfm/sid/154888/2016-17-Panini-Grand-Reserve-Startups?sort=0 renders 20 rows against a header of "Total Cards: 20"; seventeen carry real codes (ST-BB, ST-CJ, ST-DJ and so on) and three render the literal prefix "ST-" with nothing after it, on Andrew Wiggins, Wade Baldwin IV and Doug McDermott. CHASE_CHECKLIST_SPEC names this token shape as more dangerous than TBD because it reads as a real code, and says both writing it literally and repairing it with an invented suffix are fabrication. Recorded here, not acted on, because this release writes no insert_cards.csv. RESOLVE BY: the chase pass, from a marketplace listing title for one of the three named players.
• G-016 Vintage=unnumbered IS SINGLE-SOURCE AND RESTS ON A DISCRIMINATIVE TEST RATHER THAN ON A STATEMENT. Vintage (sid 154893) is the release's only base-set parallel. Its own flag column is EMPTY on every rendered row on three endpoints - https://www.tcdb.com/Checklist.cfm/sid/154893/2016-17-Panini-Grand-Reserve-Vintage?sort=0 (84 rows), https://www.tcdb.com/ViewSet.cfm/sid/154893/2016-17-Panini-Grand-Reserve-Vintage (81 rows) and https://www.tcdb.com/Gallery.cfm/sid/154893/2016-17-Panini-Grand-Reserve-Vintage?PageIndex=1&sort=0 (10 tiles), each asked directly whether any row carries a token beginning SN or PR and each answering no. PER CAV-042 THAT ABSENCE IS WORTH NOTHING BY ITSELF, and on this sid it is demonstrably lossy: 154893 also drops the RC token that the parent renders on card 1 Ben Simmons, which is a new measured instance of the flag column differing between two sids of one product. What licenses `unnumbered` is that THE SAME THREE RENDERERS emit AU, MEM and SN tokens on the sibling sid 154902 Cornerstones while emitting none on 154893 and none on the parent's 100 base rows. source_confidence is `single`: cardboardconnection names the tier ("PARALLEL CARDS: Vintage.") and prints NO figure for it (https://www.cardboardconnection.com/2016-17-panini-grand-reserve-basketball), and an editorial checklist omitting a figure is not a statement of fact. Note also that CAV-041 would have left NO LEGAL ROW for this tier had the probe come back variable, since print_run=UNKNOWN is forbidden at base-set scope - the row exists only because the discriminative test came back clean. RESOLVE BY: a source stating in words whether Vintage parallels are serial numbered.
• G-017 FAMILIES TAXONOMY ASYMMETRY, REPORTED NOT REPAIRED: THREE OF THIS PRODUCT'S FOUR STONE FINISHES FALL TO `other` WHILE THE FOURTH HAS A FAMILY. parallel_family was derived by AST-lifting family_of from pipeline/fix_schema.py, adding nothing (CAV-039). The derived values are Marble -> Marble (a whole-word match on an existing pattern at list index 152) and Granite -> other, Onyx -> other, Quartz -> other, plus Tag -> other and Vintage -> other; none of the six matches any pattern as a non-whole-word substring, so NO ordered-substring defect of the Ink/Pink, Cosmo/Cosmos, Metallized/Meta, Apprentice/Ice or Gunmetal/Meta class is present here. pipeline/check_families.py was run read-only after writing parallels.csv and reports zero defect candidates across the corpus. THE ASYMMETRY IS THE FINDING: Granite, Marble, Onyx and Quartz are one four-step stone-finish ladder inside one product, applied identically across seven of its sets, and the derivation splits them one-to-three. That is CAV-021's open taxonomy question in a concrete new instance. I did NOT add Granite, Onyx or Quartz to FAMILIES - CAV-019 requires a corpus dry-run before any new pattern and CAV-039 records that 268 hand-written families already exist that the derivation would destroy. The derived values are written as derived. RESOLVE BY: an owner decision on whether FAMILIES gains the three siblings, taken together with the 268 rows.
• G-018 WHAT THIS BUILD DID NOT DO. (1) MARKETPLACE CORROBORATION WAS NOT ATTEMPTED ONCE. No print run or card number here rests on a marketplace listing, and scope memo section 9 records that nobody has tested marketplace corroboration on 2016-17 at all. The six flat integers in parallels.csv and the six set-level flat figures in G-013 would each be strengthened by one seller title; the nine UNKNOWN rows would not, because no single listing can refute a per-card variable run. (2) beckett.com was not opened, confirmed 302ing to a maintenance page (CAV-033); one beckett.com/news title is cited in G-007 as a title only. (3) checklistcenter, checklistinsider, sportscardchecklist and waxstat were not swept, so NO INDEPENDENCE TEST WAS RUN ON ANY FIELD - cardboardconnection is the only editorial lineage here and CAV-037 as corrected says independence must be asked per product and per field. Where a value rests on cc alone or on tcdb alone the row says `single`. (4) No insert_cards.csv was written; that is the chase pass, and G-014 and G-015 hand it two live card-number findings. (5) The 2017-18 parents were NOT checked for a 2016-17 Grand Reserve Update block - the EXTRACTION_SPEC 5.5 mirror - which matters less here than elsewhere because the base run is contiguous 1-100 with no hole for an Update block to fill, but it was not looked at. (6) The pack-odds question is closed only for two sets: cardboardconnection states "Case hit" for All Systems Go and "1:7 cases" for Closing Statements and nothing for the other eighteen (https://www.cardboardconnection.com/2016-17-panini-grand-reserve-basketball), which per CAV-015 is unpublished rather than unresearched for a Panini product.

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