2018-19 Certified

• G-001 Freshman Fabric Signatures card count disagreement, 39 against 40. tcdb states Total Cards "39", renders 39 rows, and its own Freshman Fabric Signatures Gold parallel sid states Total Cards "39" and renders 39 rows (https://www.tcdb.com/ViewSet.cfm/sid/190458/2018-19-Panini-Certified-Freshman-Fabric-Signatures , https://www.tcdb.com/ViewSet.cfm/sid/190510/2018-19-Panini-Certified-Freshman-Fabric-Signatures-Gold ). The 39 card codes were enumerated individually rather than taken from any header or from a fetch's own tally - FFS-AH, FFS-AS, FFS-BB, FFS-CH, FFS-CS, FFS-DA, FFS-DD, FFS-DEM, FFS-DG, FFS-DM, FFS-EO, FFS-GA, FFS-GT, FFS-HD, FFS-JB, FFS-JE, FFS-JJ, FFS-JO, FFS-JR, FFS-JV, FFS-KBD, FFS-KH, FFS-KK, FFS-LD, FFS-LS, FFS-LW, FFS-MB, FFS-MIB, FFS-MOB, FFS-MPJ, FFS-MW, FFS-OS, FFS-RW, FFS-SGA, FFS-SVI, FFS-TB, FFS-TY, FFS-WC, FFS-ZS - and counted here. checklistcenter heads the same set "2018-19 Panini Certified - Freshman Fabric Signature Set" and lists 40 rows (https://www.checklistcenter.com/2018-19-panini-certified-basketball/ ). 39 is written. CONVENTION DIVERGENCE WITHIN THIS LINE: certified-bkb-2019-20 met the mirror-image case on its Certified Signatures set - tcdb 39 against checklistcenter and cardboardconnection 40 - and wrote 40. That release had two aggregators; this one has one, and this one enumerated the codes. Same shape, opposite call, and nothing in either row tells a reader which convention it is under. Resolve by opening the 40th row checklistcenter claims and naming the player it adds.
• G-002 Base-set parallel NAMES are contested between structural and marketplace sources and the contest is unresolved. Two independent structural checklists print the bare colour: tcdb renders the tiers as "2018-19 Panini Certified - Red", "- Blue", "- Orange", "- Purple", "- Gold", "- Green", "- Black" plus "- Base Mirror" (https://www.tcdb.com/Inserts.cfm/sid/184567/2018-19-Panini-Certified ), and checklistcenter prints "Base Mirror; Red #/299; Blue #/199; Orange #/99; Purple #/49; Gold #/10; Green #/5; Black 1/1" (https://www.checklistcenter.com/2018-19-panini-certified-basketball/ ). Two marketplace catalogues print the Mirror-prefixed form: eBay's catalogue page carries a structured item specific Parallel/Variety "Mirror Blue" and the title "2018-19 Panini Certified - P.J. Tucker #115 Mirror Blue /199" (https://www.ebay.com/p/3056523620 ), and comc files the tier under the set path segment "Panini_Certified_-_Base_-_Mirror_Red". The comc page itself returned HTTP 403 to this environment, so that citation is the weaker of the two. The structural form is written and value_contested=yes is set on the seven colour tiers. NOTE the two marketplace catalogues may be ONE opinion - eBay catalogue data is widely derived - so this is not two independent corroborations of the Mirror form. Resolve by reading the printed card face or a scan.
• G-003 44 of the 54 set-scope parallel print runs are single-sourced. checklistcenter states "Gold #/10; Green #/5; Black 1/1" for each of the 18 insert and autograph sets (https://www.checklistcenter.com/2018-19-panini-certified-basketball/ ), and tcdb independently catalogues a dedicated sid for every one of those 54 tiers (https://www.tcdb.com/Inserts.cfm/sid/184567/2018-19-Panini-Certified ), so the tiers themselves are two-sourced - but only ten of the 54 runs were confirmed by a direct within-sid serial-token probe: 2018 Gold (SN10), 2018 Green (SN5), 2018 Black (SN1), Fabric of the Game Gold (SN10), Fabric of the Game Green (SN5), Fabric of the Game Rookies Black (SN1), Freshman Fabric Signatures Gold (SN10), Materials Gold (SN10), New Generation Jerseys Green (SN5), Rookie Roll Call Black (SN1). Ten of ten agreed with checklistcenter, across six different set families and all three colours. The remaining 44 rows carry source_confidence=single. Resolve by probing the remaining 44 sids; the cost is 44 fetches and the expected yield is low, which is why it was not paid here.
• G-004 base_print_run recorded as unnumbered with basis assumed, NOT sourced. No source found states in words that the 2018-19 Certified base cards are unserialised. The evidence is indirect and within-parent: the parent checklist renders no serial token on any of the 200 base rows (https://www.tcdb.com/ViewSet.cfm/sid/184567/2018-19-Panini-Certified ) while seven of the eight base parallel sids in the same parent render SN299, SN199, SN99, SN49, SN10, SN5 and SN1 respectively. EXTRACTION_SPEC 2.1 is explicit that this shape is assumed, not sourced, however strong it looks. Both sibling seasons of this line carry the same value under the same basis. Resolve with a pack-insert, sell-sheet or box-back statement.
• G-005 THE 2018-19 SCOPE MEMO IS WRONG ABOUT WHICH SEASON THIS LINE RETURNS FROM, and the error runs straight into a brand the memo elsewhere tells builders to exclude. Memo section 1 records Certified as "return of the NBA brand for the first time since the 2010-11 season" and warns that "Certified reaches to 2010-11 at least". Checked directly against tcdb season indexes: year/2009 returns "2009-10 Panini Certified" (https://www.tcdb.com/ViewAll.cfm/sp/Basketball/year/2009 ), whose own page renders the heading "2009-10 Panini Certified", Total Cards "200" and Release Date "February 3, 2010" (https://www.tcdb.com/ViewSet.cfm/sid/9882/2009-10-Panini-Certified ). year/2010 returns exactly ONE product containing "Certified" and it is "2010-11 Panini Totally Certified" (https://www.tcdb.com/ViewAll.cfm/sp/Basketball/year/2010 ), whose page renders the heading "2010-11 Panini Totally Certified", Total Cards "185" and Release Date "July 6, 2011" (https://www.tcdb.com/ViewSet.cfm/sid/57627/2010-11-Panini-Totally-Certified ). Totally Certified is a DIFFERENT brand. year/2008 and year/2011 return no product containing "Certified" at all (https://www.tcdb.com/ViewAll.cfm/sp/Basketball/year/2008 , https://www.tcdb.com/ViewAll.cfm/sp/Basketball/year/2011 ); year/2012 returns only "2012-13 Panini Totally Certified" and year/2013 only "2013-14 Panini Totally Certified" (https://www.tcdb.com/ViewAll.cfm/sp/Basketball/year/2012 , https://www.tcdb.com/ViewAll.cfm/sp/Basketball/year/2013 ). So the span finding is: this line DOES run backwards, the memo is right that 2018-19 is a return and not a debut, and the season it returns from is 2009-10, not 2010-11. Independent corroboration of the return itself, from a retailer rather than from tcdb: "Certified makes its return to basketball for the first time in nearly a decade!" (https://www.steelcitycollectibles.com/i/2018-19-panini-certified-basketball-hobby-box ), which fits a 2009-10 predecessor (nine seasons) at least as well as a 2010-11 one. The memo attributes its wording to cardboardconnection, which this environment cannot reach (see G-012), so the original phrasing could not be re-read at source and it is possible the memo quotes it accurately and cardboardconnection is the party conflating Certified with Totally Certified. Either way the memo sentence should not be relied on as written. Earlier seasons are OUT OF SCOPE for this build and nothing was written for them.
• G-006 Guard against a plausible misreading of the "10th Anniversary Autographs" set as evidence about this line's own age. It is Panini's tenth anniversary as an NBA licensee, not Certified's tenth year: Steel City Collectibles describes "Panini's 10th Anniversary celebration in basketball cards" (https://www.steelcitycollectibles.com/i/2018-19-panini-certified-basketball-hobby-box ), and the ten cards are retired legends - tcdb renders the first as 10-AI (Allen Iverson) and the last as 10-SO (Shaquille O'Neal), all flagged AU, SN15 (https://www.tcdb.com/ViewSet.cfm/sid/190455/2018-19-Panini-Certified-10th-Anniversary-Autographs ). Recorded because a later pass counting backwards from "10th Anniversary" would land on 2008-09 and be wrong.
• G-008 No pack odds were found for any set in this release. stated_odds, odds_evidence_url and odds_evidence_quote are empty on all 19 sets.csv rows and all 62 parallels.csv rows, and scarcity_tier is empty rather than UNKNOWN, matching both sibling seasons. checklistcenter publishes a box break ("2 Autographs; 1 Memorabilia; 10 Inserts; 7 Parallels") but not per-set odds (https://www.checklistcenter.com/2018-19-panini-certified-basketball/ ); tcdb does not carry odds at all (https://www.tcdb.com/Inserts.cfm/sid/184567/2018-19-Panini-Certified ). Note the box break says 7 parallels per box against an EIGHT-tier base ladder; that is not reconciled here and is not treated as a count of tiers.
• G-009 Set and parallel NAME forms diverge between sources and between this release and its own siblings, and the divergence is a resolver hazard rather than a data error. tcdb renders "Signed Sealed Delivered", "Certified Future" and "New Generation Jerseys"; checklistcenter heads the same three "Signed, Sealed, Delivered Auto Set", "Certified Futures Set" and "New Generations Jersey Set" (https://www.checklistcenter.com/2018-19-panini-certified-basketball/ ). tcdb's forms are written. checklistcenter also misspells the product itself in one heading ("2018-19 Panini Ceritifed Basketball"). Against the siblings: certified-bkb-2020-21 carries the set "Signed Sealed and Delivered" (with "and"), and both certified-bkb-2019-20 and certified-bkb-2020-21 carry base parallels named "Mirror Blue", "Mirror Red", "Mirror Orange", "Mirror Camo", "Mirror Gold", "Mirror Green" and "Mirror Black" where this release writes "Base Mirror", "Red", "Blue", "Orange", "Purple", "Gold", "Green" and "Black". Nothing was renamed in any direction; see G-002 for the evidence behind the names written here.
• G-010 Set-level print runs are HELD but not expressible, the recurring sets.csv gap EXTRACTION_SPEC 2.2 describes. tcdb renders per-row serial tokens that are uniform across a whole set on four of the memorabilia sets - Fabric of the Game, Fabric of the Game Rookies, Materials and New Generation Jerseys all render "MEM, SN149" on every row - and checklistcenter independently prints "#/149" for each of those four (https://www.checklistcenter.com/2018-19-panini-certified-basketball/ ). "10th Anniversary Autographs" renders "AU, SN15" on every row. sets.csv has no print_run column and one is not in scope, and these are properties of the set rather than parallel tiers, so no parallels.csv row was invented to carry them. The figures live in each set's notes in words. Three further sets render per-CARD variable serials - Choice Signatures (SN199, SN99, SN49, SN25, SN15), Signed Sealed Delivered (the same five tokens) and Freshman Fabric Signatures (SN149, SN99) - which is the state the schema cannot express at all. Certified Potential Autographs and Rookie Roll Call render AU with NO serial token on any row, and the eight plain inserts render no token at all.
• G-011 Sid interleaving inside this parent is REAL and both boundaries were proved, so the season-level "easy regime" claim in the 2018-19 scope memo fails on a third parent as well. The Certified block runs 190455-190538 and both boundaries are foreign: sid 190454 renders "2009-10 Playercards Premium Serie (DEL) - DEL Elite Performers Printing Plates Black" (German hockey) and sid 190539 renders "1988 Star Mark McGwire (Aqua)" Glossy (baseball) (https://www.tcdb.com/ViewSet.cfm/sid/190454/x , https://www.tcdb.com/ViewSet.cfm/sid/190539/x ). Inserts.cfm accounts for 80 sids in that range - 72 under "Insert Sets (72)" and 8 under "Parallel Sets (8)" (https://www.tcdb.com/Inserts.cfm/sid/184567/2018-19-Panini-Certified ) - leaving exactly four unaccounted, and all four were probed and are foreign: 190496 = "1994 Magic: The Gathering HarperPrism Book Promos", 190503 = "1989 Star Kirby Puckett" Glossy, 190513 = "1989 Star Kevin McReynolds" Glossy, 190526 = "1993 Magic: The Gathering International Collectors' Edition" (https://www.tcdb.com/ViewSet.cfm/sid/190496/x , https://www.tcdb.com/ViewSet.cfm/sid/190503/x , https://www.tcdb.com/ViewSet.cfm/sid/190513/x , https://www.tcdb.com/ViewSet.cfm/sid/190526/x ). 84 sids in range, 80 Certified, 4 foreign, zero unexplained. The sid enumeration for this release is therefore complete and closed, which is the reason to record it.
• G-012 Two sources this catalogue normally relies on were unreachable and their absence narrows this release. Beckett 302s to https://beckett-maintenance-page.s3.amazonaws.com/index.html - the eighth consecutive pass on which it has been down, and it has a 2018-19 Certified page (https://www.beckett.com/news/2018-19-panini-certified-basketball-cards/ ) that was never read. cardboardconnection is refused by this environment with a robots.txt error on both URL forms tried (https://www.cardboardconnection.com/2018-19-panini-certified-basketball-nba-cards and .../2018-19-panini-certified-basketball-cards ), which matters more than usual here: the 2018-19 scope memo's span claim and its release date are both attributed to cardboardconnection, and neither could be re-read at source (see G-005). CONSEQUENCE FOR THE COUNTS: the structural spine of this release is tcdb and the only independent aggregator that answered is checklistcenter, so "two sources agree" throughout this release means tcdb plus checklistcenter and nothing else. The usual third structural source is gone.
• G-013 The fetch layer invented a gloss for a tcdb flag token on this release, exactly as EXTRACTION_SPEC 3.4 warns. Asked for the flag tokens on the Signed Sealed Delivered set it returned the token UER together with the expansion "Unedited Error Report", which appears nowhere on the page and is not what UER means on tcdb (https://www.tcdb.com/ViewSet.cfm/sid/190460/2018-19-Panini-Certified-Signed-Sealed-Delivered ). The literal token UER is recorded in that set's notes; the gloss was discarded and no meaning is written. The same response also glossed AU, MEM, SN and RC unprompted. Recorded as a live reproduction of a documented failure mode, not as a finding about the product.
• G-014 EXTRACTION_SPEC 5.3 was checked and there is nothing to exclude, and the check is recorded because the section demands the negative be evidenced rather than assumed. The block was sized off the last rendered page and never off the parent header: Gallery page 20 ends at card 200, Gallery page 21 fetched with a varied URL renders no tiles (https://www.tcdb.com/Gallery.cfm/sid/184567/2018-19-Panini-Certified?PageIndex=21&sort=0 ), Checklist.cfm?PageIndex=3 renders no rows, and Errors.cfm returns "No errors or variations have been added for this set yet." (https://www.tcdb.com/Errors.cfm/sid/184567/2018-19-Panini-Certified ). No AU/MEM/SN-flagged block sits past 200 and no variation set exists. base_set_size=200 therefore agrees with certified-bkb-2019-20 and certified-bkb-2020-21, but was re-derived here and not carried from them. For the record against the catalogue-wide split the spec names: this release is under the exclusion rule and has nothing to exclude, so it takes no side in the contenders-bkb-2019-20 / contendersoptic-bkb-2019-20 disagreement. [CORRECTED 2026-09-20: THE Errors.cfm SENTENCE QUOTED ABOVE IS NOT EVIDENCE AND THE CONCLUSION DRAWN FROM IT DOES NOT HOLD. Measured 2026-09-20 on tcdb sid 182809: Errors.cfm returns that identical sentence for a set whose own renderings carry UER on at least two cards (card 260 Mo Bamba, "UER: Statline says Undrafted, was actually drafted 6th", confirmed under &sort=0; card 67 Justin Jackson, "Career stat line is that of Kosta Koufos"). The endpoint returns the same string on a set that provably HAS errata and on one that appears not to, so it discriminates nothing and its silence is uninformative. WHAT SURVIVES: any claim here about base_set_size, about the base run being contiguous, or about no numbered block sitting past the base run, because those rest on the independent contiguous enumeration also cited in this note - the last rendered Gallery page, the empty page after it, and the per-row flag column. WHAT DOES NOT SURVIVE: the narrower claim that NO VARIATION SET EXISTS. A variation set renders as b-suffixed rows (101a/101b) that a base-run enumeration is not designed to surface, and Errors.cfm was the documented route to it, so that claim is now unsupported rather than disproved - it may still be true. Not re-fetched; recorded so it is not read as settled. See reference/caveats.csv CAV-020.]
• G-015 The digit-bearing name-suffix warning does NOT hold for this product, and carrying it here would have invited a wrong repair in the opposite direction. 2018-19 Prizm prints SS-MB3 and RS-MB3 for Marvin Bagley III and SS-LW4 / RS-LW4 for Lonnie Walker IV. The same two players in this release carry NO digit: tcdb renders FFS-MB (Marvin Bagley III) and FFS-LW (Lonnie Walker IV) in Freshman Fabric Signatures, alongside FFS-JE (Jacob Evans III) and FFS-RW (Robert Williams III) (https://www.tcdb.com/ViewSet.cfm/sid/190458/2018-19-Panini-Certified-Freshman-Fabric-Signatures?sort=0 ). The encoding is a Prizm convention, not a 2018-19 season convention. No code was normalized in either direction; chase card numbers are out of scope for this pass and none was written.
• G-016 The Inserts.cfm Notes column on this parent carries IMAGE-COMPLETION PERCENTAGES, not print runs, so the blank-annotation test could not be run and is NOT reported as run and passed. The column reads "71.3% of images added", "0.0% of images added" and so on across both headings (https://www.tcdb.com/Inserts.cfm/sid/184567/2018-19-Panini-Certified ); seven insert rows render an EMPTY cell, and those blanks mean no images have been added, not that the tier is unnumbered - Certified Future, Certified Stars, Energizers, Gold Team, Lasting Impressions, Priority Mail and The Mighty are the blank rows and all seven have Black, Gold and Green parallels that probe as SN1/SN10/SN5. Writing unnumbered off that blankness would have been FALSE, not merely unsupported. This is the Contenders / Hoops Winter shape of the column. Every print run in parallels.csv comes from a within-sid serial-token probe or from checklistcenter instead.
• G-017 The base parallel LADDER differs structurally from both later seasons of this line, which is a finding rather than a contradiction but will read as one to a consumer comparing the run. 2018-19 has EIGHT base tiers: Base Mirror (unnumbered), Red #/299, Blue #/199, Orange #/99, Purple #/49, Gold #/10, Green #/5, Black 1/1. certified-bkb-2019-20 has SEVEN: Mirror Blue and Mirror Red both unnumbered, Mirror Orange #/99, Mirror Camo #/25, Mirror Gold #/10, Mirror Green #/5, Mirror Black 1/1. certified-bkb-2020-21 has TEN: those seven plus Asia, Asia Gold and Asia Red. So 2018-19 is the only season of the three with Blue and Purple as numbered tiers, the only one with a #/299 and a #/199 tier, and the only one with a single unnumbered tier. Camo does not exist in 2018-19 at any scope. Nothing was carried from a sibling; every tier here was probed on its own sid.
• G-018 Panini Totally Certified exclusion check, run explicitly as the 2019-20 build did. No set, parallel or card written in this release carries that name. tcdb ViewAll.cfm/sp/Basketball/year/2018 returns exactly one product whose name contains "Certified" and it is "2018-19 Panini Certified", sid 184567 (https://www.tcdb.com/ViewAll.cfm/sp/Basketball/year/2018 ). The parent's own Inserts.cfm lists 72 insert rows and 8 parallel rows and none contains the string "Totally" (https://www.tcdb.com/Inserts.cfm/sid/184567/2018-19-Panini-Certified ). The build also asserts in code that reference/brands.csv contains no row whose any field contains "Totally Certified", so a future brand row of that name would abort this build rather than silently merge. NOTE the live hazard this check exists for: the adjacent seasons on tcdb ARE Totally Certified (2010-11, 2012-13, 2013-14) and the scope memo's span sentence points a reader straight at one of them - see G-005.
• G-019 A fetch's own arithmetic disagreed with a documented figure and neither was used. The 2018-19 scope memo records that ViewAll.cfm/sp/Basketball/year/2018 returned "87 products"; the fetch run for this build reported "Total number of products listed on the page: 77" for the same URL (https://www.tcdb.com/ViewAll.cfm/sp/Basketball/year/2018 ). Ten apart. Nothing in this release depends on either number - the page was used only to establish that exactly one product on it contains the string "Certified" and to read its sid - and the discrepancy is recorded because EXTRACTION_SPEC 3.2 says a summarising fetch's tally is not evidence and this is a clean instance of it. Resolve, if it matters, by enumerating the product names rather than asking for a count.

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