2016-17 Gold Standard

• G-001 WHAT "Total Cards: 373" COUNTS - a SIXTH measured unit for this header, and it is neither of the two this season had already decomposed. It is THE HIGHEST CARD NUMBER. The parent renders 438 rows; gallery row index N maps to card number N-65 at every point checked, so 438 = 373 distinct numbers + 65 letter-suffixed VAR rows. It is NOT the row count (that is 438, and the threads-bkb-2016-17 shape would predict a row count) and it is NOT the count of distinct ISSUED numbers (that is also 373 here only because nothing is unissued, so the two readings are not distinguishable on this release - they are distinguishable from the 438 row count). Decomposition closes exactly: 200 base + 38 + 31 + 31 + 38 + 35 = 373. Recorded so the next agent does not add a seventh reading by assuming one of the first five. Evidence: https://www.tcdb.com/Checklist.cfm/sid/139470/2016-17-Panini-Gold-Standard ; https://www.tcdb.com/Checklist.cfm/sid/139470/2016-17-Panini-Gold-Standard?PageIndex=5 ; https://www.tcdb.com/Gallery.cfm/sid/139470/2016-17-Panini-Gold-Standard?PageIndex=40
• G-002 CARD_COUNT NOT RESOLVED ON 38 OF THE 44 INSERT/AUTO SETS. tcdb carries NO card-count column on either of its two set listings for this parent - Inserts.cfm has none, and the parent set-overview "Included Sets:" list has none either (both checked explicitly) - so every card_count costs its own fetch, the documented second-largest cost in this pass. Six were measured off their own pages (Rookie Jersey Autographs 38, RJA Double 31, RJA Triple 31, RJA Jumbo 38, Solid Gold 50, Bullion Brand Logo Tag Rookies 35). The remaining 38 are written with card_count EMPTY, which migrations/002 documents as "NULL = unresolved", rather than guessed. The image-completion percentages in the Inserts.cfm Notes cell are recorded verbatim in each row's notes and permit an arithmetic guess at the count (7.9% correctly implied 3 of 38 on RJA Prime); that is a DERIVED number, not a sourced one, and no count was written from it. Evidence: https://www.tcdb.com/Inserts.cfm/sid/139470/2016-17-Panini-Gold-Standard?sort=2
• G-003 TIER-OR-SET IS UNRESOLVED FOR 18 OF tcdb's 45 INSERT ROWS, AND THIS RELEASE DIVERGES FROM THE SIBLING CONVENTION. tcdb files as separate sids 18 rows whose names are a base set name plus a tier word - 7 in the Rookie Jersey Autographs family (Prime Logo Tags 175899, Double Prime 175900, Double Prime Tags 143866, Triple Prime 175902, Triple Prime Tags 175903, Jumbo Prime 175905, Jumbo Prime Tags 175906) and 11 others ending Black, Platinum or Prime (Golden Jumbos Threads Black, Golden Pairs Black, Golden Quads Black, Golden Trios Black, Newly Minted Memorabilia Black/Duals Black/Quads Black/Triples Black, White Gold Threads Black, Gold Strike Signatures Platinum/Prime, Good as Gold Autographs Platinum/Prime). EVERY OTHER 2016-17 RELEASE IN data/ FILES SUCH TIERS IN parallels.csv WITH scope_type=set - threads-bkb-2016-17 has 45 set-scope parallel rows against 26 sets, nattreasures 75, optic 88, prizm 65. This release files only ONE of them that way (Rookie Jersey Autographs Prime), because that is the only one whose numbering was actually measured: sid 175898 renders cards 201-238, the same numbers and roster as sid 143864, at SN25 against SN199, which is what makes it a tier. The test is load-bearing and not guessable from the name - in the SAME family, Double, Triple and Jumbo are NOT tiers: they occupy their own ranges 239-269, 270-300 and 301-338 and are distinct sets. So the 18 unopened rows are left as sets.csv rows, which is the weaker claim; promoting them to set-scope parallels would assert a tier relationship and a print run that were never measured. THIS IS A RECORDED CONVENTION DIVERGENCE, not a resolved question: a consumer comparing goldstandard-bkb-2016-17 against its sibling 2016-17 releases will see set counts that are not like-for-like. Resolution is 18 fetches, one per sid, reading the first and last card number and the flag column.
• G-004 base_print_run=269 RESTS ON ONE LINEAGE. tcdb renders the token "SN269" on every one of the 200 base rows and on all 65 VAR rows, and it is corroborated across three tcdb renderings - Checklist.cfm?PageIndex=1 default sort, the same page under ?sort=0 agreeing token for token on all 72 overlapping rows, and Gallery.cfm tiles on PageIndex 8, 9, 10, 18, 19 and 20 - but three endpoints of one source is still ONE SOURCE. TWO EDITORIAL SOURCES WERE ASKED AND NEITHER STATES A BASE PRINT RUN: cardboardconnection's product page gives "PARALLEL CARDS: AU #/79, Black #/15." and "200 cards." but no figure for the base cards themselves, and gogts.net returns NOT STATED on base set size, base print run and parallels alike. basis is written as "sourced" rather than "assumed" because tcdb's SN269 is a positive per-row serial statement, not an inference from an absence - the EXTRACTION_SPEC 2.1 "assumed" carve-out is about concluding unnumbered from a blank flag column, which is the opposite direction. A second lineage for /269 is still wanted. Evidence: https://www.tcdb.com/Checklist.cfm/sid/139470/2016-17-Panini-Gold-Standard ; https://www.tcdb.com/Checklist.cfm/sid/139470/2016-17-Panini-Gold-Standard?sort=0 ; https://www.cardboardconnection.com/2016-17-panini-gold-standard-basketball ; https://gogts.net/2016-17-panini-gold-standard-nba-basketball-cards-checklist/
• G-005 SIXTY-FIVE VARIATION ROWS CANNOT BE EXPRESSED AND ARE NOT IN base_cards.csv. The base run carries a photo/team variation layer rendered by tcdb as letter-suffixed card numbers flagged VAR, all at SN269: 1b 1c 1d, 5b, 25b, 37b, 42b, 43b, 53b 53c 53d 53e 53f 53g, 55b 55c 55d, 58b, 66b, 71b, 73b, 76b, 82b, 87b, 95b, 96b, 97b, 108b 108c, 115b, 118b, 119b, 120b, 122b, 129b, 130b, 133b, 139b, 145b, 155b, 167b, 171b, 175b, 177b 177c 177d 177e 177f, 179b 179c, 183a 183b, 188b 188c 188d 188e 188f, 194b 194c 194d 194e 194f, 198b 198c 198d - 65 rows, my own count, and 438 rendered rows minus 373 numbers = 65 independently. cards.card_number is INTEGER NOT NULL (migrations/002) so none of them can load, and per EXTRACTION_SPEC 2.4 a non-integer base card number is a finding rather than a value. They are NOT a sets.csv row either: tcdb gives them no sid and no set name, and this parent's Inserts.cfm and Included Sets listings both close at 45 and 47 entries with no variation set among them. Note 183 takes the suffix "a" as well as "b" (183, 183a, 183b), so a checker matching only /b$/ will miss one. Errors.cfm was NOT consulted on this parent - per CHASE_CHECKLIST_SPEC it is the only source for Nb variation tokens and the right next step, and per CAV-020 its silence would prove nothing. All 65 are listed here so the layer is recoverable without re-enumerating the run. Evidence: https://www.tcdb.com/Checklist.cfm/sid/139470/2016-17-Panini-Gold-Standard ; https://www.tcdb.com/Checklist.cfm/sid/139470/2016-17-Panini-Gold-Standard?PageIndex=2 ; https://www.tcdb.com/Checklist.cfm/sid/139470/2016-17-Panini-Gold-Standard?PageIndex=3
• G-006 THE 339-373 BLOCK IS NOT ATTRIBUTED TO ANY NAMED SET - 35 cards, flag column "RC, SN25", NO AU and NO MEM. Four of the five blocks past the base run map to named sids whose own pages were opened and whose number spans match exactly (201-238 Rookie Jersey Autographs, 239-269 RJA Double, 270-300 RJA Triple, 301-338 RJA Jumbo). The fifth does not. First row card 339 Brandon Ingram, last row card 373 Diamond Stone; the roster includes Ben Simmons (340), Domantas Sabonis (350), Dejounte Murray (356) and Patrick McCaw (372), who appear in NO autograph block, so it is not a tier of one. The absent AU and MEM tokens are DISCRIMINATIVE rather than dropped: the same page renders AU and MEM freely on rows 201-338, and a second differently-varied fetch (Gallery.cfm?PageIndex=42&sort=0) returned "RC, SN25" on cards 346-355 and reported neither AU nor MEM present anywhere on it - the CAV-042 both-directions check. Two candidates were opened and ELIMINATED: Solid Gold (sid 143897) is 50 cards numbered 1-50 at SN1, and Bullion Brand Logo Tag Rookies (sid 143870) is 35 cards numbered 1-35 at SN1 - the same COUNT of 35 but its own numbering, ending on Kay Felder not Diamond Stone, so the matching count is a coincidence and not an identification. So this is either a sid-less block filed in the parent (EXTRACTION_SPEC 5.3) or one of the 38 sets not opened (G-002). It is EXCLUDED from base_set_size either way, which is the decision that mattered, and that exclusion does not depend on naming it. Evidence: https://www.tcdb.com/Checklist.cfm/sid/139470/2016-17-Panini-Gold-Standard?PageIndex=5 ; https://www.tcdb.com/Gallery.cfm/sid/139470/2016-17-Panini-Gold-Standard?PageIndex=42&sort=0
• G-007 rookie_flag=no ON ALL 200 BASE CARDS RESTS ON A WITHIN-SID DISCRIMINATIVE TEST, NOT ON AN RC TOKEN PER ROW. CAV-003 says Panini checklists usually do not print RC designations, which would force UNKNOWN. This parent DOES print them: tcdb renders RC on all 173 rows from 201 to 373 and on none of the 265 rows covering cards 1-200, in the same sid and in the same fetches - the strongest form of the positive discriminative test in EXTRACTION_SPEC 6. Consistent with it, no member of the 2016 draft class appears anywhere in 1-200; the 2016-17 rookies are all at 201 and above. That consistency is a sanity check, not the evidence. If a later pass finds tcdb's RC token is set-scoped rather than card-scoped on this line, all 200 rows revert to UNKNOWN. Evidence: https://www.tcdb.com/Checklist.cfm/sid/139470/2016-17-Panini-Gold-Standard ; https://www.tcdb.com/Checklist.cfm/sid/139470/2016-17-Panini-Gold-Standard?PageIndex=3 ; https://www.tcdb.com/Checklist.cfm/sid/139470/2016-17-Panini-Gold-Standard?PageIndex=5
• G-008 CHANNEL IS UNKNOWN ON ALL THREE PARALLEL ROWS AND channel_basis IS DELIBERATELY EMPTY. No source read states a trade channel, box SKU or region for any tier of this product: tcdb carries no channel field, and cardboardconnection's product page was asked for the parallels and returned "PARALLEL CARDS: AU #/79, Black #/15." with no channel statement. Per EXTRACTION_SPEC 2.3 and migrations/003's basis_iff_channel_known, declining to assign a channel means channel=UNKNOWN with channel_basis EMPTY - "unsupported" would be rejected by the database and is not a way to say the channel itself is unknown. channel_confidence is also left empty per CAV-038 rather than carrying an unpoliced value. channel_class is the derived "unknown", which per CAV-013 means the box-type question has no box-type answer, not that it was skipped. Evidence: https://www.cardboardconnection.com/2016-17-panini-gold-standard-basketball
• G-009 NO FLOOR IS CLAIMED FOR goldstandard AND NONE SHOULD BE INFERRED FROM THIS RELEASE. 2016-17 is not this line's first season: the 2016-17 scope memo section 3a records Gold Standard present in tcdb's own 2015-16 index at sid 118994 and in its 2014-15 index at sid 102645, so at least two further seasons exist beneath this one. Those two sids were NOT re-verified by this build - the claim is carried from the scope memo, not measured here. Per the memo's section 3b Studio lesson a floor needs a POSITIVE STATEMENT and absence from adjacent indexes is not one; no such statement was sought or found for this line either way, so the depth below 2016-17 is unmeasured rather than bounded. brands.csv admitted goldstandard on 2026-09-28 with no floor claimed.
• G-010 SID BOUNDARIES PROVED BOTH WAYS ON BOTH BLOCKS, AND THE UPPER BOUNDARY OF THE MAIN BLOCK IS A CROSS-SPORT TRAP. This release's 47 child sids fall in two blocks, 143862-143899 and 175898-175906, taken from tcdb's own two closed listings rather than from a sweep. Block A lower: sid 143861 renders "1995 Fleer Ultra Ralston Foods Spider-Man", "Total Cards: 5" - foreign, non-sport. Block A upper: sid 143900 renders "2016 Donruss Optic - Aqua", "Total Cards: 200" - A DONRUSS OPTIC SET ONE SID ABOVE THE BLOCK'S TOP, and `optic` is a live brands.csv line, so anyone sweeping upward from 143899 lands on a page that reads as a plausible Donruss Optic release. Same shape as the 148294 "2016 Panini Encased - Gold" trap recorded next to Prestige. Its season prefix is the single-year "2016" rather than "2016-17", which is the tell that it is not the basketball product; the fetch also glossed it as football, and that gloss is recorded as a gloss, not relied on. Block B lower: sid 175897 renders "2016 Topps Apex MLS - Crest Jumbo Relics Green", "Total Cards: 45" - soccer. Block B upper: sid 175907 DOES NOT EXIST and renders the tcdb HOME PAGE ("Cards Listed: 35,440,084"), the documented non-existent-sid signature.
• G-011 set_type ON 38 SETS WAS ASSIGNED FROM THE SET NAME, NOT FROM OPENING THE SET. set_type is a NOT NULL enum with no UNKNOWN value (base|insert|auto), so a value had to be written for every row. Rule applied: `auto` where the tcdb name contains an explicit signing word (Autographs, Signatures, Scripts, Graphs) - 20 of the 38 - and `insert` otherwise, as the CONSERVATIVE value, because `auto` is the stronger claim and EXTRACTION_SPEC 2.2 has no relic value so an unsigned jersey, patch or logo-tag set is an `insert` regardless. The 18 named `insert` this way include Golden Pairs, Golden Trios, Golden Quads, Mother Lode, Gold Standard, the Newly Minted Memorabilia family, Bullion Brand Logo Tag and the Threads sets; any of them could carry signatures and would then be `auto`. Six rows are NOT in this bucket - their set_type was read off the flag column on their own pages. Resolution is the same 38 fetches as G-002 and should be done with it.
• G-012 CAV-036 ON THIS PARENT: THE Inserts.cfm NOTES COLUMN IS PRESENT IN ONE SECTION AND WORTHLESS IN THE OTHER, AND A STRUCTURE-ONLY ASK GOT IT WRONG. Measured per section, as CAV-036 requires. Parallel Sets (2 rows): the cell carries clean PROSE serial statements, "Serial numbered to 79" and "Serial numbered to 15", and both were confirmed against the sets' own pages (SN79 on 200 rows, SN15 on 200 rows) and against cardboardconnection - usable, in the prose format, with no blank cells. Insert Sets (45 rows): 42 of 45 cells are image-completion percentages only, there are ZERO blank cells, and the both-directions split test is therefore UNRUNNABLE, not merely failed. The 3 exceptions are the hoopspremium fourth shape - a semicolon list carrying a real token after the percentage: "20.0% of images added; Serial numbered to 99, 47 or 25" (Gold Scripts), "15.0% of images added; Serial numbered to 75 or 25" (Gold Standard), "0.0% of images added; Serial numbered to 15" (White Gold Threads Black) - so the spec's own worked-example regex anchored at the start of the cell extracts nothing from exactly the three cells that carry data. ALSO RECORDED AS A METHOD FINDING: a first fetch asking only for the page's STRUCTURE answered that the page "does not contain a column headed Notes or similar", which is FALSE - the per-section ask one fetch later returned the two prose serial statements from that very column. A structure-only question is not a safe way to decide whether the Notes column exists; ask the section for its cells. No print run was written from this column for any insert set.
• G-013 NEGATIVE FINDING, RECORDED DELIBERATELY: THIS PARENT CARRIES NO PRIOR-SEASON SUBSETS. EXTRACTION_SPEC 5.5 and the agent brief both require prior-season blocks filed under this parent to be written into sets.csv verbatim with their season prefix and flagged, never gapped away - 2016-17 National Treasures carries 68 such rows from five products and three seasons, and two orchestrator briefs have been reversed on this point. It was checked here in both directions and there are NONE: the Inserts.cfm Insert Sets section was asked directly whether any set name begins with a season other than 2016-17 and answered no across all 45 rows, and the parent set-overview "Included Sets:" list renders all 47 entries under the "Panini Gold Standard -" prefix with no foreign product or season among them. So there is nothing to file, and that is a measured absence rather than an unexamined one. THE MIRROR WAS NOT CHECKED: the 2017-18 Gold Standard parent was not opened for a 2016-17 Update block, which EXTRACTION_SPEC 5.5 names as the checkable direction in which this conclusion could still be incomplete. Note no number in 1-373 is unissued here, so there is no missing card for an Update block to hold.
• G-014 PRINT RUNS EXIST FOR SEVERAL INSERT SETS AND sets.csv CANNOT HOLD THEM - plus two CEILINGS that must not become integers. EXTRACTION_SPEC 2.2 is explicit that sets.csv has no print_run column and that a uniformly numbered set's figure can only live in notes in words. Figures found and deliberately not written as values: Rookie Jersey Autographs is "#/199 or less" per cardboardconnection - a CEILING, and per EXTRACTION_SPEC 2.3 a ceiling is not a print run, so no 199 was written anywhere (tcdb independently renders SN199 on all 38 of its rows, which does support a uniform 199 for THAT tier, but the set row still cannot carry it). Gold Scripts reads "Serial numbered to 99, 47 or 25" and Gold Standard reads "Serial numbered to 75 or 25" - MULTI-VALUED statements, i.e. tcdb flagging its own per-tier or per-card variation, which is the EXTRACTION_SPEC 2.3 expressiveness gap arriving before any tier row was even written. White Gold Threads Black reads "Serial numbered to 15". None of these sets' own pages was opened, so none is a parallels.csv row; resolving them is part of G-003.
• G-015 WHAT THIS BUILD DID NOT DO, STATED PLAINLY. (1) Errors.cfm was not fetched on the parent or any subset, so the 65-row variation layer in G-005 has no suffix source other than Checklist.cfm and tcdb's own b/c/d/e/f/g tokens are recorded as tcdb's disambiguators, not as characters claimed to be printed on the cards. (2) No marketplace corroboration was attempted - no comc path segment, goldin lot title or eBay title - which EXTRACTION_SPEC 3.4 requires before writing plain integers onto an autograph set; no integer was written onto any autograph set here, so the requirement was not triggered, but the control was not exercised either. (3) checklistcenter, checklistinsider, sportscardchecklist and waxstat were not swept, so the only independence test run on this product is tcdb against cardboardconnection; per CAV-037 as corrected those two should be read as ONE lineage by default and a checkable divergence looked for per product - here they AGREE on every checkable figure (release date 12/30/2016, base 200 cards, AU #/79, Black #/15) with no figure differing, which CAV-037 names as a shared-upstream signature rather than as independence. source_confidence=multi on the AU and Black rows should be read with that caveat. (4) blog.paniniamerica.net was not fetched, although the 2016-17 scope memo section 7.6 records a first-look or QC gallery existing for Gold Standard specifically - that is the cheapest unspent first-party source for the /269 base run in G-004 and for the channel in G-008. (5) base_card_tiers.csv was deliberately NOT written: the base run is uniform at /269 over all 200 cards, so there is nothing to tile.

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