2021-22 Obsidian

• G-001 tcdb.com's ViewSet page for '2021-22 Panini Obsidian' states 'Total Cards: 240', and its paginated checklist continues numbering beyond the corroborated 200-card base into cards 201-240 (all explicitly RC-tagged, mostly overlapping the 151-200 rookie subjects but including at least one player -- Scottie Lewis, card #210 -- absent from the 1-200 range). checklistcenter.com, cardsmithsbreaks.com (full sequential 1-200 checklist matching tcdb 1-200 exactly), and sportscardchecklist.com (base + all 16 base-parallel sets each listed at 200 cards) independently agree the base checklist is 200 cards. No editorial or manufacturer source was found naming or describing a 201-240 subset (e.g. 'Rookie SP' or 'Rookie Variation'). Treated base_set_size as 200 and excluded 201-240 from base_cards.csv; this 40-card tier is NOT represented anywhere in the output files pending a source that names it.
• G-002 Electric Etch Red Flood: existence as a base-set parallel is corroborated by checklistcenter.com, dacardworld.com (Asia Tmall box product page), tcdb.com, sportscardchecklist.com, and multiple eBay listings of actual 2021-22 basketball cards (e.g. 'Magnitude Electric Etch Red Flood' -- note: that specific listing is for the INSERT set Magnitude's own Red Flood parallel, not the base set, so was not used as corroboration for the base parallel). No source found states a print run for the BASE-SET Electric Etch Red Flood specifically. Excluded from parallels.csv rather than guessing a number by pattern-matching against its sibling Flood parallels (Green Flood /25, Yellow Flood /10, Red Yellow Green Flood /99).
• G-003 Electric Etch White Pulsar: existence as a base-set parallel is corroborated by checklistcenter.com, sportscardchecklist.com, tcdb.com, and multiple basketball-specific eBay listings (e.g. 'Electric Etch White Pulsar #12 Ja Morant', 'Kyle Lowry #20 Electric Etch White Pulsar'). One eBay listing described a 2021-22 Obsidian White Pulsar card as 'SSP' (Kareem Abdul-Jabbar). A print run of /11 was found for White Pulsar cards, but only on 2021-22 Panini Obsidian SOCCER listings (Everton/Mane player cards) -- a different licensed product from the same release wave -- and per the hard rule against carrying odds/runs across products, this was NOT applied to the basketball parallel. No basketball-specific print run was found. Excluded from parallels.csv.
• G-004 Electric Etch Blue Finite (#/3): mentioned only once, by checklistcenter.com, alongside its other Electric Etch parallel listings. Could not be corroborated by sportscardchecklist.com's 16-item enumeration of 200-card base-parallel sets (which has no 'Blue Finite' entry), by the official blog.paniniamerica.net preview, by dacardworld.com's hobby/FOTL/Asia box listings, or by any eBay search. Given this project's history of fabricated single-source checklists, this parallel was NOT included in parallels.csv.
• G-005 Card-count disagreements between checklistcenter.com and sportscardchecklist.com for seven non-base sets: Aurora Autographs (49 vs 47), Galaxy Autographs (49 vs 48), Lightning Strike Signatures (27 vs 24), Matrix Material Autos (20 vs 19), Volcanic Signatures (39 vs 37), Tunnel Vision (40 vs 20), Rookie Jersey Ink (40 vs 37). sets.csv records the sportscardchecklist.com figure (a long, internally-varied 126-item enumeration that reads as genuine scraped site data rather than a rounded/uniform guess) as primary and notes the checklistcenter.com alternate in each row's notes field.
• G-006 beckett.com/news returned HTTP 403 on every fetch attempt for this release, and cardboardconnection.com's 2021-22 Obsidian basketball page (both the canonical and /amp URLs) was blocked by robots.txt for automated fetching -- both are the spec's normally-primary sources and could not be used directly. Relied instead on checklistcenter.com, tcdb.com, sportscardchecklist.com, cardsmithsbreaks.com, breakninja.com, dacardworld.com, steelcitycollectibles.com, and the official blog.paniniamerica.net product-preview post for cross-checking.
• G-007 Team assignments for card numbers 188 (Usman Garuba), 190 (Omer Yurtseven), 194 (Sandro Mamukelashvili), 196 (Jericho Sims), 198 (Joe Wieskamp), and 200 (David Duke Jr.) were not independently recoverable from a fetched web source (tcdb.com's paginated checklist truncated before reaching these rows in every retry, and none of these six players appear in tcdb's separately-confirmed 201-240 rookie tier). Their teams were filled from general 2021-22 NBA-season knowledge (all six players' 2021 draft-night/two-way team assignments are well-established public record) rather than from a directly quoted source row; all other 192 of 200 base_cards.csv team assignments were directly confirmed via tcdb.com's checklist table.
• G-008 Per-set parallel-tier repair pass (2026-08-27, targeted): looked for insert/autograph colour parallel tiers for these non-base sets and found none to source -- Mist, Unbreakable, Vitreous, Volcanix. No source consulted (checklistcenter.com, tcdb.com) during the 2026-08-27 targeted parallel-tier repair pass showed a per-set colour ladder for these inserts, consistent with this project's own prior single-source-confirmation-not-found notes on the same rows. Left unannotated.
• G-009 Rookie Jersey Autographs (card_count 40) has NO subset sid on tcdb.com. tcdb's Inserts.cfm listing for the basketball parent (sid 313254) enumerates 96 insert sids + 16 parallel sids over a perfectly contiguous block 323813-323924 (112 = 96 + 16), and the sids immediately outside that block belong to other products (323812 = '2021 Wild Card Alumination - Starbright Black', 323925 = '2021 G.I. Joe Deck-Building Game'), so the block is bounded and nothing is hiding in it. Alphabetically the set would sit at 323885, which is Rookie Jersey Ink. The 40 cards were instead found filed INSIDE the parent checklist as a contiguous own block at 201-240, past the 200-card base run, every row flagged 'AU, MEM, RC, SN99' (221 Bones Hyland and 224 Keon Johnson flagged SN49) - read from https://www.tcdb.com/Checklist.cfm/sid/313254/2021-22-Panini-Obsidian?PageIndex=3 and corroborated card-for-card by https://www.tcdb.com/Gallery.cfm/sid/313254 pages 21-24. THE RESIDUAL UNCERTAINTY IS THE SET NAME, NOT THE NUMBERS. The numbers are independently corroborated: an eBay seller free-text title, "EVAN MOBLEY 21'-22' Panini Obsidian RPA /99 #223" (https://www.ebay.com/itm/354424393860), matches this block's 223 Evan Mobley, and its title shape is plainly seller-written rather than eBay catalogue form. But the mapping of the block to the NAME 'Rookie Jersey Autographs' rests on two eBay catalogue entries - '2021-22 Panini Obsidian - Rookie Jersey Autograph Jalen Johnson #219 /99 (AU, MEM, RC)' (https://www.ebay.com/p/27056510188) and '... Jalen Suggs #237 Electric Etch Yellow /10 (AU, MEM, RC)' (https://www.ebay.com/itm/267238025879) - both of which carry the tcdb flag string '(AU, MEM, RC)' and are therefore tcdb-seeded rather than independent. The supporting circumstantial evidence is strong (exactly 40 cards against card_count 40; every row autograph + memorabilia + rookie + /99; the only sets.csv set left unaccounted for after all 13 others mapped to their own sid), but no non-tcdb-derived source was found that names this block. Rows were written under the sets.csv name; a reviewer wanting a fully independent name attestation should treat that one field as open. cardboardconnection.com remains robots-blocked and beckett.com/news now 302-redirects to a maintenance page, so both of the normally-primary independent checklists were unavailable (see G-006).
• G-010 RESOLVES THE SUBSTANCE OF G-001. G-001 records that tcdb.com's ViewSet page for '2021-22 Panini Obsidian' states 'Total Cards: 240' and continues numbering 201-240 past the corroborated 200-card base, all RC-tagged, with no source found naming the tier. That 201-240 tier is the Rookie Jersey Autographs set (40 cards, matching sets.csv card_count 40), filed inside the parent checklist because Panini numbered it as an own block continuing the base run rather than with a letter code. Every one of the 40 rows carries the flags 'AU, MEM, RC, SN99' - autograph, memorabilia, rookie, serial-numbered /99 - which is not a base-card flag set; see https://www.tcdb.com/Checklist.cfm/sid/313254/2021-22-Panini-Obsidian?PageIndex=3. G-001's decision to EXCLUDE 201-240 from base_set_size and base_cards.csv was therefore correct and should stand; the 40 cards are now carried in insert_cards.csv under 'Rookie Jersey Autographs'. Note the one player G-001 singles out as absent from 1-200, Scottie Lewis at #210, is present in this block. Naming caveat per G-009. No file outside insert_cards.csv and gaps.csv was touched, so release.csv and G-001's own text are left as-is for a human to amend.
• G-011 BEARS ON G-005 - tcdb.com is a third, independent voice on all seven disputed card counts and it agrees with sportscardchecklist.com (the figure sets.csv records) in every case, not with checklistcenter.com. Read as the 'Total Cards:' set-info line on a deliberately out-of-range Gallery.cfm PageIndex, then confirmed by counting the rows actually rendered across every Gallery page: Aurora Autographs 47 rendered / 'Total Cards: 47' (ccc says 49); Galaxy Autographs 48 / 48 (ccc 49); Lightning Strike Signatures 24 / 24 (ccc 27); Matrix Material Autos 19 / 19 (ccc 20); Volcanic Signatures 37 / 37 (ccc 39); Tunnel Vision 20 / 20 (ccc 40); Rookie Jersey Ink 37 / 37 (ccc 40). Header and rendered rows agree in all seven, and all seven equal sets.csv. tcdb was not consulted for sets.csv's counts (sets.csv's own notes attribute them to sportscardchecklist.com and checklistcenter.com), so this is genuine third-party corroboration rather than a shared upstream. G-005 can be closed in sets.csv's favour. sets.csv was NOT edited - that is a human decision. Base Gallery URL shape: https://www.tcdb.com/Gallery.cfm/sid/<sid>?PageIndex=99.
• G-012 RUN SUMMARY, chase-checklist pass 2026-09-11. All 14 previously-unstarted insert/autograph sets were resolved in full: 430 rows written, and every set reconciles EXACTLY with its sets.csv card_count (47/47, 48/48, 24/24, 19/19, 28/28, 40/40, 37/37, 37/37, 40/40, 10/10, 30/30, 10/10, 20/20, 40/40). NO set is partial and no row was withheld, so there is no per-set shortfall gap. Source: tcdb.com Gallery.cfm page-by-page for the card_number tokens (13 sets) and Checklist.cfm on the parent sid 313254 for Rookie Jersey Autographs; tcdb.com ChecklistByLastName.cfm for team_or_school, whose roster matched the Gallery roster name-for-name in all 13 cases (its card-number column is a row index and was not used for any number). team_or_school is filled for all 430 rows. THREE NUMBERING SCHEMES ARE IN USE IN ONE RELEASE and they were derived per set from rendered evidence, never carried across sets or seasons: (1) letter-coded, with a DIFFERENT prefix and prefix length per set - AUR- (Aurora Autographs), GA- (Galaxy Autographs), LSS- (Lightning Strike Signatures), MMA- (Matrix Material Autos), RES- (Rookie Eruption Signatures), RJI- (Rookie Jersey Ink), VS- (Volcanic Signatures); (2) plain 1..N for the six plain inserts Magnitude, Mist, Pitch Black, Supernova, Tunnel Vision, Unbreakable; (3) a numeric OWN BLOCK at 201-240 for Rookie Jersey Autographs (see G-009/G-010) - an autograph/memorabilia set that is NOT letter-coded, which is the opposite of the usual Panini habit. Suffixes are not derivable from initials (AUR-ART = Artis Gilmore, AUR-CAZ = Cazzie Russell and AUR-BOB = Bob Dandridge are first-name-based; RJI-JHO = Keon Johnson; LSS-FOX = Rick Fox); every code was read off the page individually. TWO DUPLICATE PRINTED CODES were found and both rows written verbatim per spec: GA-DRO on both Dennis Rodman and Duncan Robinson, and RJI-JPR on both Jason Preston and Joshua Primo. In both cases a narrow verbatim re-fetch returning ViewCard cids showed the cids running strictly sequential in display order (19423405-14 and 19421584-93), which DISARMS the cid-adjacency smear test - adjacency carries no information on a sequential page - so neither row was withheld and neither code was repaired; both suffixes are a plausible initials fit for each player, consistent with a genuine collision. Marketplace calibration passed on three sets: Magnitude #23 Luka Doncic (https://www.ebay.de/itm/156694269565), Pitch Black No.26 Scottie Barnes (https://www.ebay.de/itm/376610238862) and Rookie Jersey Autographs #223 Evan Mobley (https://www.ebay.com/itm/354424393860), all seller free-text titles. Fabrication checks all run: copy check vs base_cards.csv shows only incidental same-number overlap (max 5 of 40 rows, star-heavy rosters drawing on the same player pool) with no ordered prefix match; no temporal impossibility (every subject is a 2021-22-era player or a retired legend, and every rookie is a 2021 draftee); no suspicious uniformity (14 sets, 9 distinct roster lengths, 8 distinct numbering prefixes); narrow verbatim re-fetch run on two pages and matched what was written. SEPARATELY: the 50 pre-existing rows (Vitreous 1-10, Volcanix 1-40) were NOT edited, but were re-verified against tcdb.com Gallery.cfm sid 323916 and sid 323924 and match card-for-card, number-for-number. Their notes cite breakninja.com as the source, which CHASE_CHECKLIST_SPEC.md lists as a site that must never supply a card_number; it happens to be correct here because both are plain 1..N inserts, which is exactly the per-set-type split the spec describes - the aggregators renumber what is coded and leave what is genuinely ordinal alone. The rows stand; the sourcing note is worth a human look.

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