2020-21 Certified

• G-001 base_print_run recorded as unnumbered/assumed. No source STATES that the 2020-21 Certified base cards are unserialised. The positive evidence is indirect: tcdb prints a per-row serial flag on every numbered parallel of this release (SN99 Mirror Orange, SN25 Mirror Camo, SN10 Mirror Gold, SN5 Mirror Green, SN1 Mirror Black) and prints nothing on the base checklist, on Mirror Blue or on Mirror Red; and cardboardconnection and checklistcenter both begin the figures at "Orange #/99" with Blue and Red carrying none. That is a rule applied, not a statement found, so the basis is `assumed` per EXTRACTION_SPEC 2.1.
• G-002 release_date 2021-02-17 rests on two sources that may be one opinion, and one of them qualifies it. cardboardconnection prints "Estimated Release Date: February 17, 2021" (https://www.cardboardconnection.com/2020-21-panini-certified-basketball-nba-cards); checklistcenter prints "Release Date: February 17, 2021" unqualified (https://www.checklistcenter.com/2020-21-panini-certified-basketball-card-checklist/). A third source pulls the other way: sportscollectorsdaily, published 2020-12-01, says the product "is scheduled to land online and at hobby shops" on a timing it describes relative to the season start rather than by date (https://www.sportscollectorsdaily.com/2020-21-panini-certified-basketball-preview/). No Panini first-party release-date page was found. The ISO date is written on the two aggregators; the "Estimated" qualifier is recorded here rather than dropped.
• G-003 STRUCTURAL: the three Asia base-set parallels are NOT parallels of the full 200-card base run. tcdb states Total Cards "100" for Asia, Asia Gold and Asia Red alike, and they carry their OWN 1-100 numbering - tcdb renders "[1] Draymond Green" on Asia Gold where base card 1 is Spencer Dinwiddie and Draymond Green is base card 5. The roster is a 100-card selection taken in base order (base 5, 10, 15, 16, 17, 20 = Asia 1-6) and renumbered from 1. They are filed as base-set-scope parallels because tcdb lists them under "Parallel Sets" (https://www.tcdb.com/Inserts.cfm/sid/253172/2020-21-Panini-Certified) and cardboardconnection lists them under the base set ("ASIA PARALLELS: Base Asia, Gold, Red."), but parallels.csv has no column for a parallel's own roster size or numbering scheme, so a consumer reading these three rows will assume 200 cards numbered 1-200 and be wrong on both counts. Unresolved: whether they should instead be catalogued as their own sets.
• G-004 Asia print runs. NO source states a print run for ANY Asia tier in this release - not cardboardconnection ("ASIA PARALLELS: Base Asia, Gold, Red." with no #/ notation, and "ASIA PARALLELS: Light Blue (No Johnson*), Red & Gold." for Fabric of the Game), not checklistcenter (which does not list the Asia tiers at all). The three BASE-scope Asia tiers are written unnumbered/no on indirect evidence only: a full-page tcdb probe of Asia Gold returned "NO ROW ON THE PAGE CARRIES A SERIAL-NUMBER FLAG", and first-page probes of Asia and Asia Red returned none either, on a renderer that prints SN flags across this release. That is the same standard as G-001 and it is weaker for Asia and Asia Red, where only the opening rows were probed. The four SET-scope Asia tiers (Fabric of the Game and Freshman Fabric Asia Light Blue and Asia Red and Gold) were NOT probed for flags, so they are UNKNOWN/UNKNOWN rather than unnumbered - the asymmetry between the two groups is deliberate and is the evidence, not a classification.
• G-005 Fabric of the Game Signatures - the 20-card checklist is named differently by the three sources and no source shows an unqualified base version, exactly as in the 2019-20 sibling (certified-bkb-2019-20 gap G-003). tcdb: "2020-21 Panini Certified - Fabric of the Game Signatures Jersey Number" (and catalogues Asia Light Blue, Asia Red and Gold, Black, Camo, Gold as its parallels, with NO plain row). checklistcenter: "Fabric of the Game Signature Jersey Number" (singular "Signature"). cardboardconnection: "Fabric of the Game Signatures". sets.csv carries the family name common to all three - byte-identical to the sibling release's set_name - and "Jersey Number" is carried as a parallel with value_contested=yes, so no side is silently picked. Unresolved: whether the 20-card checklist IS the Jersey Number version or a plain version all three sources omit.
• G-006 Two tiers are PROVABLY serial numbered with a per-card variable run, and the schema cannot say so. (a) Fabric of the Game Signatures Jersey Number: tcdb annotates its 20 rows SN51, SN8, SN11, SN5, SN50, SN27, SN12, SN20, SN1, SN7, SN13, SN34, SN17, SN11, SN19, SN40, SN13, SN34, SN67, SN7 - each card numbered to that player's jersey number. (b) Fabric of the Game Signatures Camo: 18 rows, of which FG-DAF carries SN18 and FG-KJN carries SN22 against SN25 on the other 16. parallels.csv forces print_run and serial_numbered to move together, so both collapse to UNKNOWN/UNKNOWN - indistinguishable from "we found nothing". The held fact is recorded here because the CSV has nowhere to put it. This is the same defect certified-bkb-2019-20 recorded as its G-005 and EXTRACTION_SPEC 2.3 flags as unfixed.
• G-007 Ceiling not resolved, and the catalogue is inconsistent about it. cardboardconnection "Camo #/25 or less (No Giles/Bogdan*)" and checklistcenter "Camoe #/25 or Less" publish Fabric of the Game Signatures Camo as a ceiling; tcdb confirms the ceiling is real (SN18 and SN22 rows against SN25). Per EXTRACTION_SPEC 2.3 a ceiling is not a print run, so this tier is UNKNOWN/UNKNOWN rather than 25. This DIVERGES from older catalogue practice, where several releases wrote the ceiling as an integer with a gap (contenders-bkb-2019-20 Ticket Stub 77, elite-bkb-2019-20 Aspirations 99). It AGREES with the 2019-20 sibling, which took the same UNKNOWN line under this spec (its G-006, 19 tiers). Only ONE tier in this release is a ceiling, against 19 in 2019-20 - the aggregators publish exact figures for every other tier this season.
• G-008 Freshman Fabric Signatures Green is single-source: tcdb alone catalogues it (Total Cards "40", "SN5" on every rendered row, https://www.tcdb.com/Checklist.cfm/sid/261146/2020-21-Panini-Certified-Freshman-Fabric-Signatures-Green); cardboardconnection "PARALLEL CARDS: Red #/99. Blue #/49, Camo #/25, Gold #/10, Black 1/1." and checklistcenter "Parallels: Red #/99; Blue #/49; Camo #/25; Gold #/10; Black 1/1" both omit Green. The print run 5 is a direct reading of tcdb's own flag, not an inference from the sibling tiers. The 2019-20 sibling recorded the identical single-source Green on this same set (its G-007) but had to leave print_run UNKNOWN; two seasons running, one source carries a tier the other two do not.
• G-010 rookie_flag rests on a single source. tcdb renders an RC token on cards 151-200 and on no card in 1-150; no other source publishes per-card RC designations for this base set. It is corroborated only at the aggregate level, by sportscollectorsdaily: "Certified delivers a 200-card checklist with 150 veterans and 50 rookies" (https://www.sportscollectorsdaily.com/2020-21-panini-certified-basketball-preview/), which matches the 150/50 split exactly. Written yes/no on that basis.
• G-011 TEMPORAL CHECK FLAG (EXTRACTION_SPEC 4.3): card 178 Caleb Martin (Charlotte Hornets) is rendered with an RC token inside the 151-200 rookie block, but Caleb Martin went undrafted in 2019 and played the 2019-20 NBA season, so 2020-21 is not his rookie season by playing record. Every other card in 151-200 is a 2020 draftee. The flag is written as tcdb displays it (Gallery.cfm page 18) and the anomaly recorded rather than silently corrected or silently copied. Unresolved: whether Panini designated him RC in this product or tcdb's row is wrong.
• G-012 No pack odds were found for any set in this release. cardboardconnection publishes only box-level break averages ("1 Rookie Autograph and 1 Additional Autograph", "10 Parallels", "8 Inserts/Insert Parallels" for hobby; "1 Exclusive Parallel or Autograph", "1 Exclusive Parallel", "4 Base Asia Cards" for Asia), which are not per-set stated odds. stated_odds, odds_evidence_url and odds_evidence_quote are therefore empty on every sets.csv and parallels.csv row and scarcity_tier is unset throughout.
• G-013 Source depth. beckett.com could not be consulted: https://www.beckett.com/news/2020-21-panini-certified-basketball-cards/ 302-redirects to https://beckett-maintenance-page.s3.amazonaws.com/index.html - the same failure certified-bkb-2019-20 recorded as its G-013, so the usual third structural source is gone for this line in both seasons. That leaves cardboardconnection and checklistcenter, which agree almost verbatim on every parallel line and are plausibly ONE opinion counted twice (EXTRACTION_SPEC 5.4). Where tcdb could test them it confirmed them (Rookie Roll Call Camo SN25; 2020 Mirror Gold SN10; and cardboardconnection's "(No Giles/Bogdan*)" matched tcdb's 18-row Camo roster exactly), so they are calibrated, not merely trusted. Separately the two disagree on hobby box price: sportscollectorsdaily says "around $100" while checklistcenter says "Price: Approx. $450USD Per Hobby Box"; neither figure is written to any column.
• G-014 CATALOGUE-LEVEL: no evidence was found that Certified shipped a 2021-22 basketball product. tcdb ViewAll.cfm/sp/Basketball/year/2021 returns no entry whose name contains "Certified", while ViewAll.cfm/sp/Basketball/year/2020 returns exactly one ("2020-21 Panini Certified", sid 253172). This supports the task prompt's belief that 2020-21 is the line's last basketball season, but it is a single index's silence - absence from one catalogue is not a manufacturer statement of discontinuation, and no Panini announcement was located. Recorded so the catalogue does not later read this as a confirmed end-of-line.
• G-015 license_status was NOT supplied by the task prompt, which gave only product_name, season, manufacturer and league=NBA. EXTRACTION_SPEC 2.1 defines license_status as a prompt-given value, not a researched one. "licensed (NBA/NBPA)" is written because that is what the 2019-20 sibling carries for the same line and every base row renders an NBA team, but it is inherited, not given and not sourced. Flagged so a reviewer can confirm or correct it rather than assume it was specified.
• G-017 Set-name punctuation varies across sources for "Signed Sealed and Delivered": tcdb renders it unpunctuated in both the set title and the sid slug (https://www.tcdb.com/Checklist.cfm/sid/261167/2020-21-Panini-Certified-Signed-Sealed-and-Delivered); cardboardconnection writes "Signed, Sealed and Delivered"; checklistcenter writes "Signed, Sealed, and Delivered". tcdb's form is written. Likewise checklistcenter pluralises "Freshman Fabrics Signatures" against tcdb and cardboardconnection's "Freshman Fabric Signatures" (tcdb's form written), and cardboardconnection suffixes "Rookie Roll Call Autographs" and checklistcenter "Rookie Roll Call Auto" against tcdb's bare "Rookie Roll Call" (tcdb's form written, matching the sibling release). No name was normalised silently.
• G-019 cardboardconnection contradicts itself on one Asia tier name. Its parallel lines read "ASIA PARALLELS: Light Blue (No Johnson*), Red & Gold." (Fabric of the Game) and "ASIA PARALLELS: Light Blue, Red & Gold." (Freshman Fabric), but its prose sentence reads "The Asia boxes contain exclusive Red & Blue and Light Blue parallels for Freshman Fabric and Fabric of the Game." - Red & Blue against Red & Gold, on the same page. tcdb names the subsets "Asia Red and Gold" (sids 261130 and 261141), agreeing with the parallel lines two-to-one, so "Asia Red and Gold" is written and the rows carry value_contested=yes. No "Asia Red and Blue" subset exists on tcdb.
• G-020 Print runs for 15 of the 16 insert-set Mirror tiers (2020, Gold Team, Gold Team Rookies, The Mighty x Camo/Gold/Green/Black) are SINGLE-SOURCE. Only checklistcenter publishes a figure for them ("Mirror Parallels: Camo #/25; Gold #/10; Green #/5; Black 1/1" under each of the four set headings, https://www.checklistcenter.com/2020-21-panini-certified-basketball-card-checklist/); a narrow re-fetch of cardboardconnection returned "NOT PRESENT" for a parallel line under all four of those headings. tcdb corroborates exactly one of the sixteen: "2020-21 Panini Certified - 2020 Mirror Gold" states Total Cards "30" with "SN10" on every row, which is why that single row carries source_confidence=multi and the other fifteen carry single. The remaining fifteen figures were not independently verified and could each be checked with one tcdb fetch.

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