2021-22 Contenders Optic

• G-001 base_print_run recorded as unnumbered / basis assumed. No source STATES that the 2021-22 Contenders Optic base cards 1-100 are unserialised. The evidence is indirect but strong and consistent: tcdb prints a per-row serial flag on every numbered base parallel ("SN99" on Blue, "SN49" on Orange, "SN45" on Blue Wave, "SN25" on Green Pulsar, "SN10" on Gold and Gold Wave, "SN1" on Gold Vinyl and all four Printing Plates) and prints NO FLAG TOKEN ON ANY ROW of the base checklist or of Silver, Red, 75th Anniversary, Red Wave or Black Pandora SSP; and both aggregators open the base ladder with the unfigured "75th Anniversary, Red, Red Wave, Silver, Black Pandora SSP" before "Blue #/99". Per EXTRACTION_SPEC 2.1 that is a rule I applied, not a statement I found, so the basis is assumed, not sourced.
• G-002 BASE SET SIZED AT 100 AGAINST A PARENT "Total Cards" OF 140, AND THE CATALOGUE CONTRADICTS THAT CALL IN THIS SAME SEASON. EXTRACTION_SPEC 5.3's rule is the EXCLUSION rule and it was applied: cards 101-140 are an autograph block filed inside the parent checklist past the base run (every row flagged "AU, RC" against a blank flag column on 1-100; five base parallels stop at 100 while eleven run to 140; both aggregators head a 100-card base set and a separate 40-card Rookie Ticket set; and the same parallel name carries different runs across the boundary). THE DIVERGENCE, WITH EXACT FIGURES. data/contenders-bkb-2021-22 - the sibling product in MY OWN SEASON - sets base_set_size=165 and states in its release notes: "Base set = the Season Ticket numbering, 165 cards: #1-100 Season Ticket (unsigned veterans), #101-145 Rookie Season Ticket Autographs (on-card), #146-165 Rookie Season Ticket Stickers (sticker autographs)." That is the INCLUSION convention, applied to structurally the same object I have excluded. This release sets base_set_size=100 and files cards 101-140 as the auto set "Rookie Ticket" in sets.csv. RE-VERIFIED AGAINST THE REPO ON THIS DATE, three lines of csv.DictReader over data/*/release.csv: contenders-bkb-2019-20 = 165 (include); contenders-bkb-2020-21 = 100 (exclude); contenders-bkb-2021-22 = 165 (include); contenders-bkb-2022-23 = 100 (exclude); contenders-bkb-2023-24 = 100 (exclude); contendersoptic-bkb-2019-20 = 100 (exclude); contendersoptic-bkb-2020-21 = 100 (exclude); this release = 100 (exclude). The split runs by SEASON, not by line, exactly as the corrected EXTRACTION_SPEC 5.3 table states, and the repair needed is over two releases - contenders-bkb-2019-20 and contenders-bkb-2021-22 - not a run of four. NOTE the divergence is now sharper than when the 2020-21 sibling logged it (its G-002): the two releases that disagree are in the SAME SEASON, from the same manufacturer, in the same design family, and a consumer asking "how big is the 2021-22 Contenders base set?" gets 165 from one row and 100 from the other with nothing in either row saying which convention it is under. Flagged, not worked around, and not resolved in either direction.
• G-003 THE NEAR-TWIN CHECKLIST RELATIONSHIP RECORDED FOR 2020-21 DOES NOT HOLD FOR 2021-22, AND THAT IS ITSELF THE FINDING. Positional (card_number, player_name) overlap of this release's 100 base cards is 0 of 100 against data/contenders-bkb-2021-22/base_cards.csv and 1 of 100 against data/optic-bkb-2021-22/base_cards.csv - the single coincidence being slot 78, Luguentz Dort, in both. The 2019-20 sibling measured 1 of 100 against each of its two neighbours; the 2020-21 sibling measured 78 of 100 against its own season's Contenders and concluded Panini had rebuilt rather than reprinted that checklist. For 2021-22 the two products share no slot at all: Contenders opens 1 Keldon Johnson / 2 Domantas Sabonis / 3 Ja Morant, this product opens 1 Jayson Tatum / 2 Jaylen Brown / 3 Robert Williams III (team-ordered, four Celtics then four Nets). Roster overlap ignoring slot is 72 of 100 names against Contenders and 88 of 100 against Donruss Optic, which is what a shared league-wide veteran pool looks like, not a shared checklist. So this is NOT an EXTRACTION_SPEC 5.2 identical-checklist case and no base-card identity collision exists for this season. UNRESOLVED: why the relationship changed between 2020-21 and 2021-22. No source was found that comments on it.
• G-004 CARD 111 IS MISSING FROM tcdb'S PARENT CHECKLIST BUT THE CARD EXISTS, AND THIS IS THE OPPOSITE OF THE 2020-21 SIBLING'S FINDING. The task prompt warned to watch for an unissued number because the 2020-21 block ran 101-140 with no 127, confirmed three ways. The same three-way test here says the number is ABSENT FROM tcdb, not absent from the product: the parent checklist goes 110b straight to 112 (https://www.tcdb.com/Checklist.cfm/sid/338927/2021-22-Panini-Contenders-Optic?PageIndex=2); Errors.cfm returns exactly 39 variation rows 101b-140b with no 111b (https://www.tcdb.com/Errors.cfm/sid/338927/2021-22-Panini-Contenders-Optic); and Gallery page 12 ends at 110b while page 13 opens at 112 (https://www.tcdb.com/Gallery.cfm/sid/338927/2021-22-Panini-Contenders-Optic?PageIndex=12&sort=0; https://www.tcdb.com/Gallery.cfm/sid/338927/2021-22-Panini-Contenders-Optic?PageIndex=13&sort=0). But a FOURTH test contradicts all three: tcdb's own base-and-block parallel sets render the card. Red page 2: "111 Joshua Primo AU, SN99 San Antonio Spurs" (https://www.tcdb.com/Checklist.cfm/sid/341248/2021-22-Panini-Contenders-Optic---Red?PageIndex=2). Blue page 2: "111 | Joshua Primo | San Antonio Spurs | AU, SN49" (https://www.tcdb.com/Checklist.cfm/sid/341253/2021-22-Panini-Contenders-Optic---Blue?PageIndex=2). Two different sids, two different serial flags, same player. And both aggregators state the Rookie Ticket set is 40 cards - cardboardconnection "40 cards." (https://www.cardboardconnection.com/2021-22-panini-contenders-optic-basketball-nba-cards), checklistcenter "Rookie Ticket Auto Set" / "40 Cards" (https://www.checklistcenter.com/2021-22-panini-contenders-optic-basketball-card-checklist/) - while both state the variation set is 39, which is exactly what you would expect if 111 is issued and is the one card with no variation. 100 + 40 = 140 = the parent's stated Total Cards. sets.csv therefore records Rookie Ticket at 40 and Rookie Ticket Variation at 39. The base run 1-100 is unaffected either way. UNRESOLVED: why tcdb's parent checklist, Errors.cfm and Gallery all omit a card its own parallel sets carry. No source was found that comments on it, and the card-level checklist of the block is out of scope for this pass.
• G-005 The parent "Total Cards" figure of 140 (https://www.tcdb.com/ViewSet.cfm/sid/338927/2021-22-Panini-Contenders-Optic) counts DISTINCT CARD NUMBERS 1-140, not rendered rows: the parent renders 1-100 (100 rows) plus 101-140 (40 rows, one of which tcdb itself drops) plus 101b-140b (39 variation rows), i.e. about 179 rows. The base-and-block parallels state "140" on the same basis while rendering roughly 179 rows each. EXTRACTION_SPEC 3.2 warns that Total Cards counts different things on different lines; recorded here because on this release it happens to equal base_set_size + Rookie Ticket exactly, which makes it look like a corroboration of the base size when it is not one. No value in these files was taken from it.
• G-006 Release date written as 2022-12-14. FOUR sources agree: tcdb's parent set page "Release Date: December 14, 2022" (https://www.tcdb.com/ViewSet.cfm/sid/338927/2021-22-Panini-Contenders-Optic); cardboardconnection "Release Date: December 14, 2022" (https://www.cardboardconnection.com/2021-22-panini-contenders-optic-basketball-nba-cards); checklistcenter "December 14, 2022" (https://www.checklistcenter.com/2021-22-panini-contenders-optic-basketball-card-checklist/); gogts.net "Panini Contenders Optic NBA is scheduled to release on Wednesday, December 14, 2022." (https://gogts.net/2021-22-panini-contenders-optic-nba-basketball-cards-checklist/); and cardsmithsbreaks "Release Date: December 14, 2022" (https://cardsmithsbreaks.com/full-checklist/2021-22-panini-contenders-optic-basketball/). No source disagrees. Recorded as a gap only to note that beckett, the usual independent check, is unreachable (G-010), and that the date is eight months after 2021-22 Panini Contenders (2022-04-13) and three months after 2021-22 Donruss Optic (2022-09-02) - i.e. this product shipped a full season late, which is why its base checklist carries post-February-2022-deadline teams (8 Seth Curry Brooklyn Nets, 45 Caris LeVert Cleveland Cavaliers, 50 Tyrese Haliburton Indiana Pacers, 36 Domantas Sabonis Sacramento Kings, 95 CJ McCollum New Orleans Pelicans).
• G-007 license_status was NOT supplied by the task prompt, which specified only product_name, season, manufacturer and league. "licensed (NBA/NBPA)" is written because that is what BOTH prior seasons of this same line carry (data/contendersoptic-bkb-2019-20 and data/contendersoptic-bkb-2020-21). NOTE the two 2021-22 releases this product is most often confused with both carry the narrower "licensed (NBA)": data/contenders-bkb-2021-22 and data/optic-bkb-2021-22. No source was consulted on the NBPA question and no evidence supports either string over the other; the value is INHERITED, not sourced, and the catalogue is inconsistent about it across three Panini NBA products of the same season. This is the third season of this line to log the identical gap.
• G-008 rookie_flag rests on a single source. tcdb renders no (RC) token on any of cards 1-100 - blank flag column across rows 1-84 of the parent checklist (https://www.tcdb.com/Checklist.cfm/sid/338927/2021-22-Panini-Contenders-Optic?PageIndex=1) and no tokens on Gallery tiles 81-100 (https://www.tcdb.com/Gallery.cfm/sid/338927/2021-22-Panini-Contenders-Optic?PageIndex=9; https://www.tcdb.com/Gallery.cfm/sid/338927/2021-22-Panini-Contenders-Optic?PageIndex=10) - on the SAME renderer that prints "AU, RC" on every row of 101-140 (https://www.tcdb.com/Checklist.cfm/sid/338927/2021-22-Panini-Contenders-Optic?PageIndex=2), so absence across the base run is evidenced rather than unresolved. But no other source consulted publishes RC designations for this base set; neither cardboardconnection nor checklistcenter marks rookies on the base checklist. All 100 rows written rookie_flag=no on tcdb's display alone. TEMPORAL CHECK RUN AND PASSED: no 2021 draftee appears in 1-100 - the whole 2021 class sits in the 101-140 block (Cade Cunningham 124, Evan Mobley 138, Jalen Green 110, Scottie Barnes 121, Josh Giddey 130) - and the nearest candidates for a delayed rookie designation, Anthony Edwards (75), Desmond Bane (94), Devin Vassell (100), Saddiq Bey (47), Cole Anthony (65) and Tyrese Haliburton (50), are all 2020 draftees whose RC cards appeared in the 2020-21 release of this same line, so rookie_flag=no is correct for them.
• G-009 NO PACK ODDS AND NO SCARCITY TIERS WERE FOUND FOR ANY SET IN THIS RELEASE. checklistcenter publishes only a box-level "Box Break: 1 Auto; 2 Inserts; 2 Parallels" and a box price (https://www.checklistcenter.com/2021-22-panini-contenders-optic-basketball-card-checklist/); cardboardconnection only "Hobby Configuration: 6 cards per pack, 1 pack per box, 20 boxes per case" (https://www.cardboardconnection.com/2021-22-panini-contenders-optic-basketball-nba-cards); gogts.net only a per-box contents line (https://gogts.net/2021-22-panini-contenders-optic-nba-basketball-cards-checklist/). stated_odds, odds_evidence_url and odds_evidence_quote are EMPTY on every sets.csv and parallels.csv row, and scarcity_tier is unset throughout sets.csv. The single exception in parallels.csv is base Black Pandora SSP, where scarcity_tier=SSP is taken from the tier's own printed NAME as tcdb and both aggregators write it ("Black Pandora SSP") - not from any odds or print statement. Note the Rookies-scope tier of the same finish is named "Rookies Black Pandora" by tcdb WITHOUT the SSP suffix while checklistcenter writes "Black Pandora SSP" for it; that row is left with scarcity_tier unset.
• G-010 SOURCE AVAILABILITY AND FETCH BEHAVIOUR. (1) beckett.com is STILL unreachable on this date: https://www.beckett.com/news/2021-22-panini-contenders-optic-basketball-cards/ 302-redirects to https://beckett-maintenance-page.s3.amazonaws.com/index.html - the identical failure recorded on certified-bkb-2019-20, contendersoptic-bkb-2019-20 (its G-015) and contendersoptic-bkb-2020-21 (its G-017), so the outage is persistent across at least four passes and the usual third structural source is gone for this line. NOTE data/contenders-bkb-2021-22 reports having read beckett through an allorigins text proxy; that route was not used here. (2) cardboardconnection and checklistcenter agree verbatim on this product at almost every point checked - the base ladder, the Rookie Ticket ladder, every insert ladder, and every insert count - so they should be treated as plausibly ONE opinion. They are NOT one opinion everywhere: they disagree on Legendary Contenders Autographs (43 against 34) and on Perennial Contenders Autographs (28 against 25). Every count written was checked against tcdb for that reason. (3) gogts.net was consulted as a third aggregator and carries a release date, a configuration line and a partial base parallel list, but no set counts and no per-set ladders (https://gogts.net/2021-22-panini-contenders-optic-nba-basketball-cards-checklist/). cardsmithsbreaks carries only a title and a release date (https://cardsmithsbreaks.com/full-checklist/2021-22-panini-contenders-optic-basketball/). (4) curl and any direct HTTP client are blocked in this environment (the egress proxy rejects CONNECT to www.tcdb.com), so every reading went through the summarizing fetch layer and no raw HTML was ever inspected.
• G-011 SET NAMING: five unresolved naming questions, none of which affects a count. (1) "All Star Aspirations" is written as tcdb prints it and as tcdb names its four parallel sids (https://www.tcdb.com/Inserts.cfm/sid/338927/2021-22-Panini-Contenders-Optic); BOTH aggregators print "All-Star Aspirations" with a hyphen, and data/contendersoptic-bkb-2020-21 carries the hyphenated form for the same set, so the two seasons of this line now differ by one character. (2) "Rookie Ticket Variation" is written SINGULAR because both aggregators print it singular this season ("Rookie Ticket Variation Set Checklist", "Rookie Ticket Variation Set"); data/contendersoptic-bkb-2020-21 carries the PLURAL "Rookie Ticket Variations" because cardboardconnection printed it plural that season, and the 2019-20 sibling carries the singular. The name genuinely moves between seasons on the source; it is written as this season's sources print it. (3) "Pick n Roll" is tcdb's spelling and its page title; checklistcenter prints "Pick 'n' Roll Set". (4) The two block sets have NO NAME ON tcdb at all because they have no sid - the names "Rookie Ticket" and "Rookie Ticket Variation" are cardboardconnection's section headings; checklistcenter calls the first "Rookie Ticket Auto Set". (5) "Veteran Ticket" is tcdb's name, chosen because tcdb also names its eleven parallel sids "Veteran Ticket Blue", "Veteran Ticket Gold Wave" and so on; checklistcenter calls it "Veteran Ticket Auto Set". No name was invented and no existing name was rewritten.
• G-012 THREE AUTOGRAPH SET COUNTS ARE CONTESTED AND ALL THREE ARE WRITTEN AGAINST AT LEAST ONE SOURCE. (a) Legendary Contenders Autographs written 34. tcdb's own set page states Total Cards "33" AND renders exactly 33 rows (LCA-AMS through LCA-TPA), but its Blue parallel states "34" and renders 34, the extra row being "LCA-DWL | Dominique Wilkins | Atlanta Hawks | AU, SN75" (https://www.tcdb.com/Checklist.cfm/sid/346377/2021-22-Panini-Contenders-Optic---Legendary-Contenders-Autographs-Blue). checklistcenter states 34; cardboardconnection states "43 cards.", which no other source supports and which reads as a digit transposition of 34 but is NOT confirmed as one. So tcdb's parent page is missing a card its own parallel carries - the same defect shape as card 111 in G-004, on a different set. (b) Perennial Contenders Autographs written 28: tcdb states 28 and its Blue parallel states 28 and renders all 28 rows; cardboardconnection states 28; checklistcenter alone states 25. (c) Sophomore Contenders Autographs written 13: tcdb states 13 and renders 13 on the parent AND 13 on the Blue parallel, i.e. two complete row lists; BOTH aggregators state 14. 13 is written because tcdb states it on two independent pages and renders a complete list on both, and because the two aggregators agree verbatim elsewhere on this product and may be one opinion. The identity of the disputed 14th card was not established. All three figures are single-best-supported, not confirmed.
• G-013 THE "Rookies" INSERT SET HAS NO SID OF ITS OWN ON tcdb AND ITS STATUS AS A SET RATHER THAN A PARALLEL RESTS ON ONE SOURCE. tcdb's Inserts.cfm lists seventeen "Rookies ..." sids - 75th Anniversary, Asia, Asia Blue Wave, Asia Gold Wave, Asia Red Wave, Black Pandora, Blue, Gold, Gold Vinyl, Green Pulsar, Orange, four Printing Plates, Red, Silver - and NO unqualified "Rookies" (https://www.tcdb.com/Inserts.cfm/sid/338927/2021-22-Panini-Contenders-Optic; asked twice, the second time explicitly, and the fetch answered "No link exists with text exactly Rookies"). checklistcenter heads "Rookies Set" at "25 Cards" (https://www.checklistcenter.com/2021-22-panini-contenders-optic-basketball-card-checklist/). The set is written at 25 because every one of those seventeen tcdb sids states Total Cards "25" over an identical roster opening 1 Quentin Grimes / 2 James Bouknight / 3 Jeremiah Robinson-Earl and closing 25 Cade Cunningham. COPY CHECK RUN: this is NOT the 101-140 block renamed - Jeremiah Robinson-Earl appears nowhere in 101-140 - and it is NOT the base run renamed. SECOND UNRESOLVED QUESTION: checklistcenter files "Rookies 75th Anniversary" as a SET of its own at 25 cards and does not list it in the Rookies parallel ladder, while tcdb gives it a sid of exactly the same shape as the other sixteen. It is filed here as a PARALLEL at scope "Rookies". If checklistcenter is right, sets.csv is one set short and parallels.csv one row long. THIRD: data/contendersoptic-bkb-2020-21 filed "Rookies Asia" as an insert SET (card_count 25) because that season had no non-Asia counterpart; this season it is filed as a PARALLEL of "Rookies", so the two seasons of this line classify the same-named object differently.
• G-014 THE SAME cardboardconnection PAGE RETURNED TWO DIFFERENT "PARALLEL CARDS" LINES FOR THE SAME HEADING ACROSS TWO FETCHES, AND ONLY THE NARROW RE-FETCH EXPOSED IT (EXTRACTION_SPEC 4.1). Asked for the Rookie Ticket Set ladder, one fetch of https://www.cardboardconnection.com/2021-22-panini-contenders-optic-basketball-nba-cards returned "PARALLEL CARDS: Red Wave, Red #/99, Blue #/49, Orange #/25, Gold #/10," and a later narrow fetch of the same page returned "PARALLEL CARDS: Red, Silver, Black Pandora SSP, Blue #/99, Orange #/49, Green Pulsar #/25, Gold #/10, Gold Vinyl 1/1, Printing Plates 1/1." - which is the BASE-SET ladder, attributed to the Rookie Ticket heading. The first reading is the correct one: it matches checklistcenter's Rookie Ticket ladder word for word AND matches tcdb's own per-row serial flags on the block (Red is "AU, SN99" on row 111, Blue is "AU, SN49" on row 101, Orange "AU, SN25", Gold "AU, SN10", Gold Vinyl "AU, SN1", Red Wave unflagged). No value in parallels.csv was written from the second reading. RECORDED BECAUSE THE FAILURE IS INVISIBLE FROM THE VALUE: both strings are real text from that page, and a single fetch of either would have looked like a clean citation. Consequence for this release: every cardboardconnection ladder quoted in parallels.csv that could be cross-checked against tcdb flags or against checklistcenter was cross-checked; the Rookie Ticket Variation ladder is the one that could only be partially cross-checked (see G-016).
• G-015 FOUR PARALLEL TIERS ARE PROVABLY SERIAL NUMBERED WITH A RUN THAT VARIES BY CARD, AND THE SCHEMA CANNOT CARRY IT. EXTRACTION_SPEC 2.3 forces print_run=UNKNOWN / serial_numbered=UNKNOWN on these, which is indistinguishable from "we found nothing at all". The held facts, read directly off tcdb rather than inferred: (1) Rookie Ticket Variation Red - tcdb page 2 of the Red set renders "101b Franz Wagner AU, SN49, VAR" and "102b Corey Kispert AU, SN75, VAR" on the same page and reports the page's distinct tokens as SN99, SN75, SN49 (https://www.tcdb.com/Checklist.cfm/sid/341248/2021-22-Panini-Contenders-Optic---Red?PageIndex=2); cardboardconnection states it in words, "Red #/75 or #/49". (2) Legendary Contenders Autographs Blue - 30 of 34 rows carry "AU, SN75" and four carry "AU, SN49" (LCA-CLY Clyde Drexler, LCA-HOW Hakeem Olajuwon, LCA-KGA Kevin Garnett, LCA-KML Karl Malone). (3) Perennial Contenders Autographs Blue - "PCA-ADV Anthony Davis AU, SN49" against "PCA-AEF Anthony Edwards AU, SN75" on the same page, six rows at SN49 and twenty-two at SN75. (4) The PARENT autograph sets themselves are variable in the same way (Legendary Contenders base rows run SN149/SN99/SN75/SN49; Perennial base rows SN99/SN75; Sophomore base rows SN149/SN99) - which is a property of the SET, and EXTRACTION_SPEC 2.2 says sets.csv cannot hold a set's own print run, so those figures exist nowhere in the CSVs at all. A ceiling is not a print run, so none of these was written as an integer. value_contested=yes is set on the three parallel rows.
• G-016 SEVENTY-THREE OF THE 129 PARALLEL ROWS CARRY A PRINT RUN THAT WAS NOT READ OFF THAT TIER'S OWN PAGE. Each such tier is evidenced to EXIST by its own tcdb sid in the parent inventory (https://www.tcdb.com/Inserts.cfm/sid/338927/2021-22-Panini-Contenders-Optic), and its print run comes from an aggregator ladder that was CALIBRATED against tcdb before being trusted (EXTRACTION_SPEC 3.3). Calibration points actually fetched, all passing: All Star Aspirations Blue Cracked Ice "SN75"; Hoop Dreams Red Cracked Ice NO token; Superstars Gold "SN10"; Uniformity Gold Vinyl "SN1"; Veteran Ticket Red "AU, SN35"; Veteran Ticket Orange "AU, SN15"; Veteran Ticket Gold Wave "AU, SN10"; Veteran Ticket Red Wave NO token; 1984 Tribute Autographs Blue "AU, SN25"; 1984 Tribute Autographs Red Wave NO token; Rookies Blue "SN99"; Rookies Silver, Rookies Red, Rookies Asia, Rookies 75th Anniversary all NO token; Rookies Asia Blue Wave "SN45"; Legendary Contenders Blue, Perennial Contenders Blue, Sophomore Contenders Blue all read row by row. Rows carrying source_confidence=single are exactly the unfetched ones. THE UNIFORMITY IS A SIGNAL, NOT PROOF (EXTRACTION_SPEC 4.5): cardboardconnection prints an identical four-item ladder under all nine ordinary insert headings, and tcdb independently lists exactly those four sids under each of the nine, which is why it is judged real rather than template-fill. The Rookie Ticket Variation ladder is the weakest: checklistcenter publishes no separate ladder for it, so its Printing Plate rows rest on cardboardconnection alone plus the parent plate sets' Total Cards extent of 140. Nothing here was promoted to a stronger basis than the evidence supports.
• G-017 FIFTEEN PARALLEL ROWS ARE WRITTEN channel=UNKNOWN WITH AN EMPTY channel_basis, ON PURPOSE. They are the Wave, Pandora and Pulsar finishes at every scope: base Red Wave, Black Pandora SSP, Blue Wave, Gold Wave, Green Pulsar; Rookie Ticket Red Wave and Gold Wave; Rookie Ticket Variation Red Wave and Gold Wave; 1984 Tribute Red Wave and Gold Wave; Veteran Ticket Red Wave and Gold Wave; Rookies Black Pandora and Green Pulsar. The only channel evidence for this product is the product-level hobby-box note, which supports "the product ships in hobby boxes" and nothing about an individual tier. That note was allowed to carry channel=hobby / channel_basis=from-note on the ordinary ladder (110 rows), but it was NOT extended to these fifteen, because in the 2019-20 release of this same line the Wave finishes were a China/T-Mall exclusive and carry channel=china there, and because checklistcenter files Red Wave, Blue Wave #/45 and Gold Wave #/10 under a SEPARATE heading "Asia Parallels" for the Rookies set this season (https://www.checklistcenter.com/2021-22-panini-contenders-optic-basketball-card-checklist/). Writing hobby on them would be the inference EXTRACTION_SPEC 1.5 forbids. ENCODING NOTE: channel=UNKNOWN with channel_basis EMPTY, because migrations/003_parallels.sql enforces (channel = 'UNKNOWN') = (channel_basis IS NULL) - data/contendersoptic-bkb-2020-21 made the same judgement for the same reason and encoded it as channel_basis=unsupported on six rows, which the database rejected. The reasoning was right; only the encoding was wrong, and it is corrected here. Note the consequence honestly: this release uses two different treatments of the SAME product-level note, and a consumer cannot tell from the rows alone which tiers were judged and which were inherited. ALSO UNRESOLVED: no retail, blaster, mega, FOTL, First Off The Line or international SKU was found for this product, but the ABSENCE of such SKUs was not positively confirmed by any source. [REVIEWED 2026-08-19: still open; 15 of 129 parallels lack a sourced channel.]
• G-018 THE FOUR "Rookies Asia" ROWS CARRY channel=asia ON THE SET NAME ALONE (channel_basis=from-name, channel_confidence=single). No source consulted states where these cards were sold. tcdb gives "Rookies Asia" and its Blue Wave, Gold Wave and Red Wave sub-tiers sids, names and counts but carries NO page note of any kind - specifically NOT the "China-Exclusive retail product. Cards are NOT autographed." note that the 2019-20 sibling's Rookies T-Mall pages carried. checklistcenter groups them under the heading "Asia Parallels" (https://www.checklistcenter.com/2021-22-panini-contenders-optic-basketball-card-checklist/) and says nothing about distribution. NOTE the 2019-20 equivalent was described as a T-Mall (China) product, which is NARROWER than "Asia"; if these tiers are likewise China-only then channel=china would be the correct value and channel=asia overstates the region. This is the same unresolved question data/contendersoptic-bkb-2020-21 logged as its G-009, unchanged a season later. [REVIEWED 2026-08-19: still open; 15 of 129 parallels lack a sourced channel.]
• G-019 PARALLEL FAMILIES USED, reported for checking per EXTRACTION_SPEC 2.3. No family was minted in this pass. The families used are Wave, Pandora, Pulsar, Vinyl, Printing Plate, Cracked Ice and other - every one of them inherited from the two prior seasons of this line (data/contendersoptic-bkb-2019-20 and data/contendersoptic-bkb-2020-21), so the three seasons stay comparable. NOTE that "Pandora", "Printing Plate" and "Cracked Ice" were themselves minted by those siblings and have still not been reconciled against the rest of the catalogue's 12,803 parallel rows. ONE JUDGEMENT WORTH FLAGGING: "75th Anniversary" is filed under family "other" rather than given a new family of its own. It is a league-anniversary commemorative finish new to this season, it appears at two scopes (base-set and Rookies), and a case could be made for minting "Anniversary"; it was not minted because minting a family for a one-season finish is how a family becomes a synonym for another one.
• G-020 ENDPOINT BEHAVIOUR MEASURED AGAIN, CONFIRMING THE CORRECTED EXTRACTION_SPEC 3.2 AND ADDING ONE NUMBER. Checklist.cfm?PageIndex=1 on the parent sid delivered rows 1-84 of 100 and the fetch reported "Last row visible: 84" - i.e. the truncation presented itself as the end of the page, exactly as the spec says. That is an 84th-row truncation, at the top of the 74-86 band the spec records from six agents. PageIndex=2 delivered the whole 101-140 block plus its 101b-140b variations (about 79 rows) WITHOUT truncating, so the truncation is a body-size effect, not a fixed row index. Cards 85-100 were recovered from Gallery pages 9 and 10 (https://www.tcdb.com/Gallery.cfm/sid/338927/2021-22-Panini-Contenders-Optic?PageIndex=9; https://www.tcdb.com/Gallery.cfm/sid/338927/2021-22-Panini-Contenders-Optic?PageIndex=10), 10 tiles each, with the 81-84 overlap agreeing row for row against Checklist page 1 - so the whole 100-card base run cost 3 fetches, not the 10 Gallery alone would have cost. Total cost of this release: about 55 fetches, of which roughly 30 were per-set card_count and per-tier print-run probes - the second-largest cost, exactly as EXTRACTION_SPEC 3.2 warns about Inserts.cfm carrying no card-count column. Separately: the &sort=0 cache-busting trick from the spec was used on Gallery pages 12 and 13 and on two Checklist re-reads, and it worked - the re-read returned a materially different row set from the first ask on the same sid.

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