2019-20 Status

• G-001 base_print_run recorded as unnumbered with basis `assumed`, NOT `sourced`. No source consulted STATES that the 2019-20 Status base cards are unserialised. The positive evidence is indirect but real: tcdb prints a per-row serial flag on every numbered tier of this product (SN125 on White Diamond, SN49 on Green Escher Squares, SN10 on Gold Fireworks, SN1 on Swirl-O-Rama) and prints nothing on the base rows or on Blue/Red; cardboardconnection and checklistcenter both begin the figure list at "White Diamonds - #/125" with Blue and Red carrying none. That is a rule applied, not a statement found, so per EXTRACTION_SPEC 2.1 the basis is `assumed`.
• G-002 release_date is UNKNOWN because no source published a DAY. cardboardconnection states "Release Date: November 2020" and checklistcenter states "Release Date: November 2020" - month precision only, confirmed on a narrow verbatim re-fetch of cardboardconnection which stated the page "does not provide a specific day of the month". The schema field is ISO YYYY-MM-DD, which cannot carry a month-precision date, so writing 2020-11-01 would have invented a day. The MONTH is sourced and is recorded here and in release notes. Note the date is genuinely late for a 2019-20-season product and was deliberately not "corrected" to 2019 - the Asia/T-Mall edition shipped around November 2020. Beckett, which normally carries a precise street date, was unreachable (see G-010).
• G-003 Upper Echelon card count: tcdb states Total Cards "10", renders 10 rows ending "10 Giannis Antetokounmpo Milwaukee Bucks", and all six of its parallel sets independently state Total Cards "10"; checklistcenter states "Upper Echelon Set - 10 Cards". Against this, cardboardconnection groups it in a sentence assigning 30 cards each to "New Beginnings, Status Symbols, Trophy Club and Upper Echelon". sets.csv writes 10. The disagreement is one aggregator sentence against eight independent pages, but it is recorded rather than silently dropped.
• G-004 New Kids on the Court Autographs count spread: tcdb parent Total Cards "24" / 24 rows rendered; tcdb Gold Fireworks parallel "25"; tcdb Green Escher Squares "24"; tcdb Swirl-O-Rama "27"; cardboardconnection 27; checklistcenter 27. sets.csv writes 27 on the agreement of tcdb's own 1/1 parallel with both aggregators. Which specific cards tcdb's parent omits was not identified.
• G-005 Newbie Status Signatures count spread: tcdb parent Total Cards "28" / 28 rows; Gold Fireworks "27"; Green Escher Squares "26"; Swirl-O-Rama "30"; cardboardconnection 30; checklistcenter 30. sets.csv writes 30 on the agreement of tcdb's 1/1 parallel with both aggregators. The omitted rows were not identified.
• G-006 Prominence Signatures count: tcdb's parent page CONTRADICTS ITSELF - Total Cards "7" above 8 rendered rows. Confirmed on a narrow re-fetch under a second slug (ViewSet slug is unvalidated), which returned the same "7" header and the same 8 codes: "PS-AIV, PS-DRO, PS-HOW, PS-JST, PS-JWE, PS-KGA, PS-LBI, PS-MJO". The Gold Fireworks parallel renders a DIFFERENT set of 8 - PS-DWD present, PS-LBI absent - and Swirl-O-Rama renders all 9. Union of tcdb's own rosters = 9, matching cardboardconnection (9) and checklistcenter (9 Cards). sets.csv writes 9. Resolved on rendered rows rather than on any header, but recorded because no single page states 9 as a header.
• G-007 FOUR parallel rows are PROVABLY serial numbered with a per-card VARIABLE print run, and the schema cannot express that state, so all four are written UNKNOWN/UNKNOWN - which is indistinguishable from "nothing was found". The held facts, so they are not lost: New Kids on the Court Autographs Gold Fireworks - rows flagged "AU, SN10" and "AU, SN5"; New Kids on the Court Autographs Green Escher Squares - "AU, SN15" and "AU, SN10"; Newbie Status Signatures Gold Fireworks - "AU, SN10" and "AU, SN5" (SN5 on Ja Morant only); Newbie Status Signatures Green Escher Squares - "AU, SN15" and "AU, SN5". Each is serial_numbered=yes in fact. This is the exact defect EXTRACTION_SPEC 2.3 describes; it needs either a print_run_basis column or permission to pair serial_numbered=yes with print_run=UNKNOWN.
• G-008 Parallel NAME variance not resolved: tcdb titles the BASE tier "White Diamond" (singular) - page title "2019-20 Panini Status - White Diamond Basketball Checklist" - and also uses the singular for Trophy Club and Upper Echelon, but uses "White Diamonds" (plural) for New Beginnings and Status Symbols. Both cardboardconnection ("White Diamonds - #/125") and checklistcenter ("White Diamonds #/125") use the plural throughout. Each parallel row is written with the spelling tcdb gives THAT set, and all White Diamond/White Diamonds rows carry name_ambiguous=yes with ambiguous_with naming the other spelling. Unresolved: whether the product itself prints one spelling and tcdb is inconsistent, or the two spellings are genuinely different tiers.
• G-009 No pack odds were found for any set in this release. Neither cardboardconnection, checklistcenter nor breakninja publishes per-set insertion odds for 2019-20 Status, and Beckett (the usual carrier of odds tables) is unreachable - see G-010. stated_odds, odds_evidence_url and odds_evidence_quote are therefore empty on every sets.csv row, and scarcity_tier is left empty (= pass never covered this row, per unknown_semantics.csv) rather than UNKNOWN.
• G-010 beckett.com could not be consulted: https://www.beckett.com/news/2019-20-panini-status-basketball-cards/ 302-redirects to https://beckett-maintenance-page.s3.amazonaws.com/index.html . This is the SAME failure recorded on the certified-bkb-2019-20 pilot (its G-013), so it is a persistent outage, not a transient one, and every release in this season is one structural source shallower than normal. Note also that cardboardconnection and checklistcenter agree closely on the parallel list and may be one opinion counted twice; breakninja was used as a genuinely independent third voice on channel and on the set inventory, and its wording is clearly not copied.
• G-011 cardboardconnection's box configuration line is internally contradictory and was NOT written anywhere. Verbatim, confirmed on a narrow re-fetch: "6 cards per box, 10 packs per box". That cannot be right as printed (it is almost certainly "6 cards per pack, 10 packs per box"), but the correction is a guess, so no pack/box configuration is recorded for this release. No other source consulted published a configuration.
• G-012 Renowned Autographs card-number scheme anomaly: a narrow re-fetch listing every card-number cell returned "2, RA-SMB, RA-KAT, RA-JKD, ..." - one bare "2" among 28 RA-prefixed codes, and the parent page rendered that row as "2 Markelle Fultz AU, SN25". Whether the card genuinely prints "2" or tcdb holds a malformed number was not resolved. Card-level checklists for autograph sets are out of scope for this pass (CHASE_CHECKLIST_SPEC), so this is flagged for that pass rather than resolved here.
• G-013 Possible NON-PACK distribution inside an Asia-exclusive product, unresolved. tcdb renders the Prominence Signatures row "PS-LBI Larry Bird AU, SN25 *Panini Rewards release/sale* Boston Celtics". Panini Rewards is a manufacturer loyalty programme, not the T-Mall box, so at least one card in this release may not have reached collectors through the Asia/T-Mall channel recorded on every parallel row. This is also the card that tcdb's Gold Fireworks parallel roster omits (see G-006). No source was found that states how PS-LBI was distributed.
• G-014 All 36 insert-scope and autograph-scope parallel rows rest on a SINGLE source (tcdb). Neither cardboardconnection nor checklistcenter itemises parallel tiers per insert set - both publish only the six BASE-set tiers - and breakninja states only a box-level "10 Parallels and 10 Inserts". Every one of the 36 was read individually off its own tcdb checklist rather than extrapolated, which mattered: the autograph tiers do NOT follow the base ladder (base Green Escher Squares is #/49, but Prominence and Renowned Green Escher Squares are #/15, and Prominence Gold Fireworks is #/5 against the base #/10). Corroboration for these 36 print runs is nonetheless one source deep. Related: tcdb's insert-set parallel rosters are demonstrably incomplete in places - for several autograph sets the Gold Fireworks and Green Escher Squares tiers render fewer rows than the Swirl-O-Rama 1/1 tier of the same set.
• G-015 channel=asia is a PRODUCT-level value on all 42 parallel rows; no source states a channel for any individual parallel. The product-level exclusivity is genuinely sourced in three independent wordings (cardboardconnection "2019-20 is an Asian/T-Mall exclusive", checklistcenter "exclusive to the Asian market", breakninja "basically an Asian only release"), so channel_basis is sourced-exclusive rather than from-name. Two facts the enum cannot carry, recorded here: (1) T-Mall is a RETAILER DEAL (Alibaba) as well as a region, and channel_taxonomy.csv notes `asia` is a region ORTHOGONAL to trade channel - channel_class is written `regional` and the T-Mall SKU is put in channel_sku, but a row that is regional AND retailer-deal at once cannot be expressed; (2) whether the box is hobby- or retail-configured within Asia was not sourced - retailer listings title it both "T-Mall Box" and "TMALL Hobby Box" and no editorial source settles it.
• G-016 The `unnumbered` value on the ten Blue and Red rows (base set plus New Beginnings, Status Symbols, Trophy Club, Upper Echelon) rests on evidenced ABSENCE, not on a statement. tcdb prints a serial flag on every numbered tier of this product and nothing on Blue/Red, on the same renderer; both aggregators list Blue and Red with no figure while giving figures for the other four base tiers. parallels.print_run has no basis column, so this row cannot record that its `unnumbered` is inferred the same way release.base_print_run_basis records it for the base cards (G-001). Same evidence shape, and one file can label it while the other cannot.
• G-017 ONE parallel_family was MINTED and needs checking against the catalogue: `Escher Squares`, for the Green Escher Squares tier (9 rows). No near-match exists in the 12,803 existing parallel rows (searched: "Escher", "Esch"), and the tier is a genuine named visual treatment rather than a bare colour, so it was minted rather than filed as `other`. The other four families used here already exist and were deliberately NOT re-minted: `Fireworks` (8 existing rows) for Gold Fireworks, `Swirl` (35) for Swirl-O-Rama, `Diamond` (20) for White Diamond/White Diamonds - minting "White Diamond" would have created a silent synonym for `Diamond` - and `other` (5,007) for the plain Blue and Red tiers, which publish no family name at all.

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