2019-20 Obsidian

• G-001 Panini America's own preview blog post (blog.paniniamerica.net) originally titled the product the 'May-Releasing 2019-20 Obsidian Basketball,' but the actual hobby street date used here (2020-07-15) is corroborated by three independent post-release sources instead (checklistcenter.com, tcdb.com, dacardworld.com). Not treated as a disagreement requiring resolution since the blog post predates release and describes an original target date, but noted for completeness.
• G-002 checklistcenter.com lists a 'Vitreous Set: 5 Cards' for 2019-20 Panini Obsidian Basketball. This could not be corroborated on any second source. Multiple sources (breakninja.com's checklist index) instead show 'Vitreous' as a feature of 2020-21 Panini Obsidian Basketball, and an eBay listing found during research ('2020-21 Panini Obsidian Vitreous Electric Etch Contra /9') is explicitly dated 2020-21, not 2019-20. Excluded from sets.csv and parallels.csv as likely a single-source year-confusion error rather than a genuine 2019-20 subset.
• G-003 The Rookie Jersey Autographs set is credited with 41 cards (checklistcenter.com) numbered 201-242, but card #223 does not appear in the published checklist on either checklistcenter.com or tcdb.com (both jump from #222 Tyler Herro to #224 Kevin Porter Jr.), leaving only 40 numbers publicly visible against a stated count of 41. Card #223's player identity is unverified.
• G-004 Card counts for Rookie Jersey Autographs (41), Rookie Auto (39), Jersey Autographs (17), Galaxy Autographs (50), Lightning Strike Signatures (30), Matrix Autographs (50), and Onyx Autographs (30) all come from checklistcenter.com only; no second source with an independent count was found despite confirming each set's existence elsewhere (blog.paniniamerica.net, gogts.net, dacardworld.com, steelcitycollectibles.com, tcdb.com set-name references).
• G-005 No published numeric pull odds (e.g. '1:X packs') or 'X per case' / case-hit language was found for any set, insert, or parallel in this release, across cardboardconnection.com (robots-blocked), beckett.com (403 on every attempt), dacardworld.com, steelcitycollectibles.com, blog.paniniamerica.net, gogts.net, theawesomecardshop.com, and blowoutcards.com. sportscardchecklist.com's note that Pitch Black cards are 'serialized to 99 copies' is a print run, not stated odds, and per the hard rule against inferring odds from print run it was not used to assign a scarcity_tier. As a result scarcity_sets.csv and scarcity_parallels.csv contain no rows.
• G-006 cardboardconnection.com (robots-blocked/DNS-failed on every WebFetch attempt, and a direct curl was blocked with HTTP 403 by the egress proxy) and beckett.com (HTTP 403 on every attempt) could not be fetched directly for this release, despite being the two top-priority sources in the extraction spec. All findings instead rely on checklistcenter.com, tcdb.com, dacardworld.com, steelcitycollectibles.com, theawesomecardshop.com, blog.paniniamerica.net, gogts.net, sportscardchecklist.com, blowoutcards.com, and physical-listing corroboration from comc.com, eBay sold listings, and Beckett Marketplace (item pages, not the blocked beckett.com/news site).
• G-007 3 insert/auto set(s) in this release are documented as carrying their own Electric Etch parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-008 3 insert/auto set(s) in this release are documented as carrying their own Electric Etch parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-009 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-010 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-011 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-012 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-013 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-014 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-015 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-016 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-017 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-018 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-019 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-020 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-021 3 insert/auto set(s) in this release are documented as carrying their own parallel tier, but no source enumerated the colours, so no scoped parallel rows could be created. The tier exists; its members are unknown.
• G-022 Chase/insert checklist pass, 2026-09-11: insert_cards.csv created and all 11 non-base sets resolved COMPLETE against sets.csv card_count -- 391 rows written, 391 catalogued, every set reconciling exactly (Rookie Jersey Autographs 41, Rookie Auto 39, Jersey Autographs 17, Galaxy Autographs 50, Lightning Strike Signatures 30, Matrix Autographs 50, Onyx Autographs 30, Volcanic Signatures 39, Atomic 40, Pitch Black 30, Tunnel Vision 25). Primary source tcdb.com Checklist.cfm; no set required Gallery.cfm truncation recovery. tcdb sid arithmetic reconciles exactly: Inserts.cfm/sid/227999 lists 77 insert sids = 11 tcdb sets x (1 base + 6 Electric Etch parallels). Block boundaries confirmed by probing outward -- ViewSet.cfm/sid/237921 is '2020 Panini National Treasures Collegiate' and /sid/237992 is '2013 Hero Decks Boston Bruins', both foreign products, so the 237922-237991 run is swept to exhaustion; Galaxy Autographs sits on a separate island at 237765-237771 alongside the base parallels and is reachable only via Inserts.cfm, not by sweeping from the main block.
• G-023 CONTRADICTS G-002. G-002 excluded 'Vitreous' from sets.csv as a likely single-source year-confusion error on checklistcenter.com, on the grounds it could not be corroborated and appears as a 2020-21 Obsidian feature. tcdb.com carries a 2019-20 Vitreous set for THIS release: ViewSet.cfm/sid/237978 renders heading '2019-20 Panini Obsidian - Vitreous' with 'Total Cards: 5' -- matching checklistcenter.com's 'Vitreous Set: 5 Cards' exactly -- and Inserts.cfm/sid/227999 lists it under the 2019-20 parent together with six sibling parallel sids (237979 Green, 237980 Orange, 237981 Purple, 237982 Red, 237983 White Mojo, 237984 Yellow), the identical 7-sid shape every other set in this release has. The 77-sid arithmetic only balances if Vitreous is counted. This is a second independent source for a 12th non-base set and G-002 should be re-reviewed. No rows written here: sets.csv is out of scope for this pass and a chase row cannot reference a set that does not exist in sets.csv.
• G-024 CONTRADICTS G-003 on arithmetic. G-003 states that with card #223 absent 'only 40 numbers [are] publicly visible against a stated count of 41'. That is an off-by-one: the range 201-242 inclusive is 42 numbers, so removing 223 leaves 41, not 40. tcdb Checklist.cfm/sid/227999?PageIndex=3 renders 41 rows (201-222 = 22 rows, 224-242 = 19 rows) and states 'Total Cards: 241', which equals the 200-card base plus exactly 41. The set is therefore COMPLETE at 41 with 223 simply unissued, and all 41 rows are written. No missing player remains to identify. (Note: the summarizing fetch of that page self-reported 'Total Rows Listed: 42' while actually listing 41 -- counted by hand, per spec.) G-003 may be closeable.
• G-025 DOCUMENTED PANINI PRINTING ERROR, rows written as the source pairs them. tcdb renders RA-BCL as Kyle Guy and RA-KGY as Brandon Clarke -- the reverse of what the initials suggest (BCL/Brandon CLarke, KGY/Kyle GuY), which reads as a catalogue transposition. It is not. Errors.cfm/sid/237964 annotates both rows verbatim: RA-BCL carries "UER: Brandon Clarke 'Rookie Autograph' card & info on back" and RA-KGY carries "UER: Kyle Guy 'Rookie Autograph' card & info on back". Panini mis-paired the two cards at print; tcdb's pairing is correct and is written verbatim. Flagged because a resolver matching a seller title on initials alone will mis-hit these two cards.
• G-026 Four printed codes are each shared by two different players within one set and are written as two rows each, per spec (chase_cards is unique on set_id + card_number + player_name): GA-DGR = Devean George and Devonte' Graham; MA-GAN = Giannis Antetokounmpo and Greg Anthony; OA-TYG = Thaddeus Young and Trae Young; VS-DTH = Daniel Theis and David Thompson. Each is n=2 with a plausible initials fit for both players, none carries a UER note on tcdb Errors.cfm, and OA-TYG's two rows carry different serial-number flags (SN60 vs SN35) -- so these read as genuine initials collisions, not data-entry smears. OA-TYG (Thaddeus/Trae Young) is named in CHASE_CHECKLIST_SPEC.md as a known-real collision shape on the FB- prefix.
• G-027 NAMING DIVERGENCE, not corrected. sets.csv names the 39-card rookie autograph set 'Rookie Auto'; tcdb names the same set '2019-20 Panini Obsidian - Rookie Autographs' (sid 237964) and marketplace sellers write 'Rookie Autograph'/'Rookie Auto' interchangeably. Rows were written under the sets.csv spelling byte for byte. Recorded per spec as a naming question rather than changing our row. Separately confirmed that 'Rookie Auto' and 'Rookie Jersey Autographs' are genuinely different sets, not duplicates: different numbering (RA- letter codes vs the 201-242 numeric block) and different rosters -- Nicolas Claxton, Talen Horton-Tucker and Terance Mann appear only in Rookie Auto, while Coby White, De'Andre Hunter, Jarrett Culver, Sekou Doumbouya and Tremont Waters appear only in Rookie Jersey Autographs.
• G-028 PARTIALLY RESOLVES G-004. G-004 records that card counts for Rookie Jersey Autographs (41), Rookie Auto (39), Jersey Autographs (17), Galaxy Autographs (50), Lightning Strike Signatures (30), Matrix Autographs (50) and Onyx Autographs (30) came from checklistcenter.com only, with no independent second count. tcdb.com now independently confirms every one of those seven counts via its own ViewSet.cfm 'Total Cards' field and, for each, a rendered checklist whose rows were counted by hand and matched: sid 227999 PageIndex=3 (41), 237964 (39), 237929 (17), 237765 (50), 237936 (30), 237943 (50), 237950 (30). All seven counts are therefore now two-source.

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