2019-20 Impeccable

• G-001 THE SCHEMA CANNOT EXPRESS "SERIAL NUMBERED, RUN UNKNOWN" AND THIS RELEASE LOST THAT FACT ON 21 OF 53 PARALLEL ROWS. tcdb renders a per-row serial flag on every one of these tiers, so each is PROVABLY serial numbered, but the run differs card to card, so print_run/serial_numbered collapse to UNKNOWN/UNKNOWN - indistinguishable from "nothing was found". The 21 rows: Holo Silver and Holo Gold on each of Elegance Veteran Jersey Autographs, Elegance Retired Jersey Autographs, Canvas Creations, Extravagance Autographs, Illustrious Ink, Immortal Ink, Impeccable Shots Signatures, Impeccable Stars Signatures, Impeccable Victory Signatures and Indelible Ink (20 rows), plus Stainless Stars Autographs Purple. Worked example: Elegance Retired Jersey Autographs Holo Silver (tcdb sid 228106) renders ER-CBK Charles Barkley "AU, MEM, SN5" beside ER-KAJ Kareem Abdul-Jabbar "AU, MEM, SN10". Each row's visual_notes carries the observed flags so the fact is not lost entirely. A `print_run_basis` column, or permitting serial_numbered=yes with print_run=UNKNOWN, would recover all 21.
• G-002 CEILING PRINT RUNS NOT RESOLVED, AND THIS DIVERGES FROM EARLIER CATALOGUE PRACTICE. cardboardconnection publishes the Holo tiers of ten letter-coded autograph sets as ceilings - verbatim "PARALLEL CARDS: Holo Silver #/25 or less, Holo Gold #/10 or less, Platinum 1/1." on Elegance Veteran Jersey Autographs, Elegance Retired Jersey Autographs, Canvas Creations, Extravagance Autographs, Illustrious Ink, Immortal Ink, Impeccable Shots Signatures, Impeccable Stars Signatures, Impeccable Victory Signatures and Indelible Ink. Per EXTRACTION_SPEC 2.3 a ceiling is not a print run, so all 20 tiers are UNKNOWN rather than 25/10. This is the OPPOSITE value from what several existing catalogue rows carry on identical evidence (Contenders Ticket Stub 77 "recorded as the maximum", Season Ticket Stub 50, Elite Red/Aspirations/Status 99 "the stated ceiling"), and nothing in either row tells a reader which convention it is under. Flagged for the expected repair pass. Independently, tcdb PROVES all 20 of these tiers are variable rather than merely capped, so the ceiling is genuinely not a run in this product.
• G-003 Stainless Stars Autographs Purple: sources disagree on the Zion Williamson print run and the tier is not a single figure. tcdb sid 228159 renders the ordinary rows "AU, SN49" and the Zion Williamson row "AU, SN35"; cardboardconnection states "Purple #/49 (Zion #/30*)". Both agree Zion is an exception; they disagree on whether it is 35 or 30. Recorded UNKNOWN/UNKNOWN with value_contested=yes rather than picking a side. Note the non-autograph Stainless Stars Purple has no such exception - tcdb renders Zion at SN49 there and cardboardconnection states a flat "Purple #/49" - so the two sets genuinely differ and this is not a transcription error.
• G-004 rookie_flag is UNKNOWN on all 100 base cards. tcdb prints NO (RC) token on any base tile 1-100, while printing "RC" on every tile 101-180 of the SAME parent checklist - so the renderer is demonstrably capable of the token here and withholds it on the base run. That pattern would normally support rookie_flag=no as evidenced absence, but it cannot be right: cards 88 Zion Williamson, 90 Ja Morant, 82 RJ Barrett, 1 Luka Samanic, 2 Romeo Langford and many others are 2019-20 rookies, and an eBay catalogue title for card 90 reads "2019-20 Panini Impeccable - Ja Morant #90 Holo Gold /10 (RC)". tcdb is therefore simply not populating RC on this base set. No consulted source publishes RC designations for it, so all 100 are UNKNOWN per EXTRACTION_SPEC 2.4 rather than a false "no". Resolving this needs a source that marks rookies on the base checklist.
• G-005 tcdb internal count contradiction on Elegance Retired Jersey Autographs. The parent set (sid 228104) and its Holo Gold entry (sid 228105) both state Total Cards "9", but the Holo Silver (sid 228106) and Platinum (sid 228107) entries for the same set state Total Cards "27" while rendering the same ER- letter codes. cardboardconnection (9 cards) and checklistcenter ("Elegance Retired Jersey Auto Set - 9 Cards") both say 9, so 9 is written. Read as an incomplete or mis-seeded tcdb entry on the two 27-card pages, but not positively explained by any source. Same shape as the pilot release's Mirror Camo anomaly.
• G-006 tcdb internal count contradiction on Extravagance Autographs: the parent (sid 228112), the Holo Silver entry (sid 228114) and the Platinum entry (sid 228115) all state Total Cards "28", but the Holo Gold entry (sid 228113) states "29". cardboardconnection (28) and checklistcenter ("Extravagance Auto Set - 28 Cards") agree on 28, so 28 is written. The identity of the extra Holo Gold row was not established.
• G-007 TWO REAL 40-CARD SETS ARE ABSENT FROM tcdb Inserts.cfm AND EXIST ONLY AS A NUMBERED BLOCK INSIDE THE PARENT CHECKLIST (EXTRACTION_SPEC 5.3 / the documented Inserts.cfm omission failure mode). Rookie Autographs occupies cards 101-140 and Elegance Rookie Jersey Autographs occupies cards 141-180 of parent sid 217707; neither has a sid of its own and neither appears among the 70 rows Inserts.cfm lists. They were recovered from Gallery tiles (101-140 flagged "AU, RC, SN99"; 141-180 flagged "AU, MEM, RC, SN99") and the block-to-name mapping was then SOURCED, not inferred, from checklistcenter, which renders "Rookie Auto Set" as 101 Kevin Porter Jr. to 140 Rui Hachimura and "Elegance Rookie Jersey Auto Set" as 141 Goga Bitadze to 180 Kyle Guy. Both are in sets.csv and both are EXCLUDED from base_set_size. An agent working only from Inserts.cfm would miss 80 cards and two sets; an agent trusting the parent Total Cards "180" would overstate base_set_size by 80.
• G-008 Channel is a product-level inference for all 53 parallel rows. The only channel evidence found is cardboardconnection's box configuration ("9 cards per pack, 1 pack per box, 3 boxes per case") and breakninja's ("5 Autographed Cards" and "1 Silver Insert, Gold Insert or Stainless Stars" per box). No source states a channel for any individual parallel, and no retail, blaster, mega or international SKU was found for 2019-20 Impeccable - but the ABSENCE of such SKUs was not positively confirmed. All rows carry channel=hobby with channel_basis=from-note and channel_confidence=single. The single exception with any name-level channel evidence is Silver NBA Logo FOTL - see G-016.
• G-009 beckett.com could not be consulted: https://www.beckett.com/news/2019-20-panini-impeccable-basketball-cards/ 302-redirects to https://beckett-maintenance-page.s3.amazonaws.com/index.html . Beckett is the usual third structural source, so cross-checking depth on set counts and parallel print runs is one source shallower than normal. This reproduces the pilot release's finding, i.e. the outage is not release-specific.
• G-010 cardboardconnection and checklistcenter agree almost verbatim on the base line ("100 cards. Serial numbered #/99." against "100 Cards - Serial Numbered #/99"), on the parallel ladder, and on every set count, and may be ONE opinion counted twice (EXTRACTION_SPEC 5.4). Where this release claims "three sources agree", the third is tcdb, which is structurally independent of both - counts come from its own Total Cards headers and print runs from its per-row serial flags. Where only the two aggregators support a value, source_confidence reflects that. breakninja was consulted as a fourth source but publishes no base size or print runs; it did independently confirm the set NAMES including Elegance Rookie Jersey Autographs, and that the checklist is organised by NBA teams.
• G-011 No pack odds were found for any set in this release. cardboardconnection and breakninja publish only box-level configuration ("9 cards per pack, 1 pack per box, 3 boxes per case"; "5 Autographed Cards" per box), which is not a per-set stated odds figure, so stated_odds/odds_evidence_url/odds_evidence_quote are empty on every sets.csv row and scarcity_tier is unset throughout.
• G-012 CORRECTION TO THE ENDPOINT NOTES: Gallery.cfm validates its slug, and on a SUBSET sid the valid slug is the FULL product-plus-subset name, not the bare subset name. Gallery.cfm/sid/228097/Holo-Silver?PageIndex=11 returned "The page you attempted to view does not exist", while Gallery.cfm/sid/228097/2019-20-Panini-Impeccable-Holo-Silver?PageIndex=11 returned the correct 10 tiles. This matters because Gallery on a parallel subset sid is the ONLY route found to that parallel's rows past ~row 80 - ViewSet.cfm truncates there and ViewSet.cfm?PageIndex=2 does not paginate (it re-rendered rows 1-79). Without the full slug the parallel coverage of cards 101-180 could not have been evidenced at all.
• G-013 CROSS-SPORT HAZARD WORSE THAN A HYPHEN TEST CAN CATCH: a "2019-20 Panini Impeccable Premier League" SOCCER product exists under the IDENTICAL hyphenated season string as this basketball release, alongside the single-year "2019 Panini Impeccable Football". Observed in search results including the eBay catalogue title "2019-20 Panini Impeccable Premier League - All-Time Legends Autographs Thierry Henry #AT-1 Gold /10 (AU)", which also shows the soccer product using a structurally identical "Gold /10" tier name. Season format is therefore not a sport discriminator for this line even within a single year. Every figure in this release was taken from rows carrying an NBA team, from tcdb sid 217707 reached via ViewAll.cfm/sp/Basketball/year/2019, or from cardboardconnection/checklistcenter pages whose URLs and rosters are explicitly NBA.
• G-014 Silver NBA Logo FOTL is classified inconsistently by the two sources that carry it and is recorded as a parallel, not a set. tcdb gives it its own sid (228154) in Inserts.cfm, i.e. treats it as a set, with Total Cards "49" and every row "SN15". cardboardconnection treats it as a parallel of Silver NBA Logo: "PARALLEL CARDS: FOTL #/15." Because its checklist is the same 49 cards as its parent and the only source that classifies it calls it a parallel, it is written to parallels.csv scoped to Silver NBA Logo. If the catalogue convention is that any tcdb sid becomes a sets.csv row, this row needs moving - flagged rather than decided.
• G-015 The two inline autograph blocks have per-card variable runs at their own base tier, which sets.csv has nowhere to record. In Rookie Autographs (cards 101-140) tcdb flags every tile "SN99" except card 108 Zion Williamson, flagged "SN49". In Elegance Rookie Jersey Autographs (cards 141-180) every tile is "SN99" except card 143 Zion Williamson, flagged "SN75". The parallel tiers over both blocks ARE uniform (Holo Silver 25, Holo Gold 10, Platinum 1, verified tile by tile including the Zion cards), so the variation is confined to the base tier of each set. Recorded here because sets.csv carries no print-run column.
• G-016 ENCODING TENSION BETWEEN THE SPEC AND THE TAXONOMY, resolved in favour of the taxonomy and flagged. EXTRACTION_SPEC 2.3 lists `FOTL` among the legal values of the `channel` column, but reference/channel_taxonomy.csv classifies FOTL with axis=box-sku and notes it is "First Off The Line - a pre-order allocation of a hobby product, not a separate trade channel", and the spec separately directs that box type belongs in `channel_sku`. The Silver NBA Logo FOTL row is therefore written channel=hobby, channel_sku=FOTL, channel_class=box-type, channel_basis=from-name rather than channel=FOTL. If the loader expects channel=FOTL this row needs flipping; the two documents should be reconciled either way.
• G-017 The five base parallels do not all cover the same cards, and the difference is evidence rather than an error. tcdb states Total Cards "100" for Silver (sid 228099) and Gold (sid 228095) but "180" for Holo Silver (228097), Holo Gold (228096) and Platinum (228098). Reading the 180-card sets tile by tile shows why: those three also parallel the 101-180 autograph block, while Silver and Gold stop at the 100-card base run. This is corroborated by cardboardconnection, which gives the base set five parallels but gives Rookie Autographs and Elegance Rookie Jersey Autographs only Holo Silver, Holo Gold and Platinum. The Silver/Gold count of 100 is what independently sizes the base set against the parent's misleading "180".

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