2018-19 Immaculate

• G-001 THIS RELEASE HAS NO THIRD STRUCTURAL SOURCE AND EFFECTIVELY ONLY ONE INDEPENDENT NON-tcdb SOURCE. beckett.com was not consulted: it has 302-redirected to maintenance.beckett.com since at least 2026-09-20 and returns no content (reference/caveats.csv CAV-033). checklistcenter.com, the other aggregator this line's sibling seasons leaned on, returned HTTP 404 for its 2018-19 Immaculate page (https://www.checklistcenter.com/2018-19-panini-immaculate-collection-basketball-checklist/), so it supplied nothing at all on this build. Every non-tcdb claim in this release therefore rests on cardboardconnection.com alone (https://www.cardboardconnection.com/2018-19-panini-immaculate-collection-basketball-cards). CAV-033 warns that cardboardconnection and checklistcenter are best read as ONE editorial lineage rather than two sources; here even that pair collapsed to one site. Where this release says 'two sources agree' it means cardboardconnection plus tcdb, which are genuinely independent of each other, and nothing more. The 2019-20 sibling's counts are largely checklistcenter-derived and are NOT available as a cross-check for the same reason.
• G-002 ROOKIE PATCH AUTOGRAPHS card_count IS CONTESTED 38 vs 39; 39 WAS WRITTEN. The set has no sid of its own and is filed inside the parent checklist at cards 101-140 (EXTRACTION_SPEC 5.3 hidden-block shape). FIVE independent renderings of the /99 block show 38 rows with BOTH 127 and 133 absent: https://www.tcdb.com/Checklist.cfm/sid/209949/2018-19-Panini-Immaculate-Collection?PageIndex=2 ; the same page under &sort=0 ; the parent Red sid 217156 PageIndex=2 ; the parent Gold sid 217154 PageIndex=2 ; and Rookie Logoman Autographs sid 217208, which shares the block's numbering and renders 38 rows. AGAINST THAT, four sources say 39: cardboardconnection states "Rookie Patch Autographs - 39 cards, #/99 base" (https://www.cardboardconnection.com/2018-19-panini-immaculate-collection-basketball-cards), and the Premium Edition (sid 217210), Tags (217211) and Tags Premium Edition (217212) sids each state Total Cards: 39. Per CHASE_CHECKLIST_SPEC 1c-ter the independent third party (cardboardconnection) sides with the parallel pages, so the base page is read as short and 39 is written. A reader who needs the count of cards actually rendered at the /99 tier should read 38. Note Rookie Logoman Autographs is recorded as 38 because its own page states 38 and no source states otherwise for THAT set - so this release deliberately carries 38 and 39 on two sets that share one numbering block.
• G-003 CARD 127 IS NOT ISSUED AND CARD 133 IS ISSUED IN SOME TIERS ONLY. Within the Rookie Patch Autographs block 101-140: card 127 appears in NO rendering of any tier examined (parent /99, parent Red, parent Gold, RPA Premium Edition, RPA Tags Premium Edition, Rookie Logoman Autographs). Card 133 is Rodions Kurucs: it is ABSENT from the parent's own /99 rendering under two differently-sorted fetches, and PRESENT in RPA Premium Edition (https://www.tcdb.com/ViewSet.cfm/sid/217210/2018-19-Panini-Immaculate-Collection-Rookie-Patch-Autographs-Premium-Edition?sort=0, rendered "133 Rodions Kurucs AU, MEM, SN24"), in RPA Tags and in RPA Tags Premium Edition. Either Kurucs has no /99 base card, or tcdb's parent page is missing one row. THE EXTRACTION_SPEC 5.5 MIRROR CHECK WAS RUN AND CAME BACK CLEAN: the 2019-20 Immaculate parent (sid 236707) lists no set whose name contains 'Update' or '2018-19', so neither number is hiding in the next season's Update block. Errors.cfm was NOT used to settle this and its silence would prove nothing either way (CAV-020). Unresolved: whether 127 was ever printed at all.
• G-004 THREE SETS WHERE tcdb'S OWN BASE PAGE IS EXACTLY ONE CARD SHORT OF ITS OWN PARALLEL PAGES AND OF cardboardconnection. Heralded Signatures: base page states 39, Gold (217164) and Platinum (217165) both state 40, cardboardconnection states 40 - written as 40. Immaculate Ink: base page (216989) states 19, Gold (216992) and Platinum (216993) both state 20, cardboardconnection states 20 - written as 20. Modern Marks: base page (217182) states 37, Gold (217184) and Platinum (217185) both state 38, cardboardconnection states 38 - written as 38. The same shape three times in one release is itself the finding: it is the CHASE_CHECKLIST_SPEC 1c-ter 'base page is short, parallels recovered it' pattern, not three coincidences. The missing player in each set was NOT identified - that is a card-level question for the chase pass, and naming one would be derivation.
• G-005 FIVE SET COUNTS WHERE tcdb AND cardboardconnection DISAGREE AND tcdb WAS WRITTEN, plus one where cardboardconnection was written. tcdb figure first: Immaculate Moments 25 vs 24; Logoman Autographs 54 vs 53; Patch Autographs 46 vs 50; Shadowbox Signatures 46 vs 43; Sneaker Swatches Signatures 47 vs 46. tcdb was preferred in these five because its header and its rendered rows agree with each other (and for Patch Autographs and Sneaker Swatches the figure is also echoed by the set's own Gold/Platinum parallel sid), whereas cardboardconnection publishes a single unsourced number. Separately, Dual Rookie Logoman Autographs is written as 29 on the agreement of tcdb's header AND cardboardconnection, even though the one fetch of that page self-reported 30 rendered rows - a summarising fetch's own arithmetic is not evidence (EXTRACTION_SPEC 3.2) and was not allowed to outvote two independent statements. Each of these six would be settled by counting rendered rows on a second differently-varied fetch.
• G-006 EIGHTEEN PARALLEL ROWS CARRY print_run=UNKNOWN, AND SEVENTEEN OF THEM ARE PROVABLY SERIAL NUMBERED WITH A PER-CARD VARIABLE RUN, AND THE SCHEMA CANNOT SAY SO. EXTRACTION_SPEC 2.3 forbids serial_numbered=yes with print_run=UNKNOWN, so the seventeen variable-run tiers collapse to UNKNOWN/UNKNOWN (the eighteenth, Standout Memorabilia Red, is UNKNOWN for a different reason - see G-007), which is indistinguishable from 'we found nothing'. UNKNOWN ON THESE ROWS DOES NOT MEAN UNNUMBERED - every one of them renders SN tokens on every row. The tiers and their measured DISTINCT token sets, each enumerated across a whole page rather than sampled: Materials Gold {SN10, SN6}; Patch Autographs Jersey Number {varying}; Patch Autographs Premium Edition {varying}; Patch Autographs Red {SN15, SN25}; Patch Autographs Tags Premium Edition {SN3, SN1}; Premium Patch Autographs Gold {SN10, SN5, SN3}; Premium Patch Autographs Red {SN25, SN15, SN14, SN10, SN5}; Remarkable Memorabilia Red {SN25, SN24}; Remarkable Rookie Jerseys Gold {SN10, SN5}; Remarkable Rookie Jerseys Red {SN25, SN24}; Rookie Patch Autographs Jersey Number {varying}; Rookie Patch Autographs Tags {SN3, SN24}; Sneaker Swatches Signatures Gold {SN10, SN5, SN3}; Sneaker Swatches Signatures Jumbo {SN10, SN7, SN6, SN5, SN4}; Sneaker Swatches Signatures Jumbo Red {SN5, SN2}; Sneaker Swatches Signatures Red {SN25, SN21, SN14, SN10, SN7, SN5}; Swatches Red {SN25, SN21}; plus Standout Memorabilia Red (see G-007). SEVENTEEN OF FIFTY SET-SCOPE TIERS - 34% - carry a variable run. Writing the modal token would have produced seventeen clean-looking sourced integers, every one of them wrong. cardboardconnection independently describes seven of these with 'or less' ceiling wording, which corroborates the shape; A CEILING IS NOT A PRINT RUN. This is the modelling gap of caveats.csv CAV-023 and CAV-028, not missing research.
• G-007 STANDOUT MEMORABILIA RED IS THE ONE TIER IN THIS RELEASE THAT A SINGLE EXTRA FETCH WOULD PROBABLY CLOSE. One fetch of https://www.tcdb.com/ViewSet.cfm/sid/217230/2018-19-Panini-Immaculate-Collection-Standout-Memorabilia-Red rendered 16 of a stated 20 rows, every one SN25. The confirming second, differently-varied fetch (Checklist.cfm?sort=0) was queued and lost to a session rate limit, and was never run. print_run was left UNKNOWN rather than written as 25, because a token seen on one fetch over 80% of a tier is not an enumerated distinct-token set, and this release has already produced seventeen tiers whose own rows disagree - including this set's OWN base page, which mixes SN99 and SN49. Its sibling tiers Standout Memorabilia Gold (20/20, all SN10) and Platinum (20/20, all SN1) were fully enumerated and are flat, so 25 is the likely answer; likely is not sourced. ACTION: one fetch of Checklist.cfm/sid/217230/...?sort=0.
• G-008 THE FOUR BASE-SET PRINTING PLATE TIERS ARE WRITTEN AS print_run=1 ON PARTIAL COVERAGE, AND THE ALTERNATIVE WAS REMOVING THEM. Printing Plates Black, Cyan, Magenta and Yellow (sids 553824-553827) each state Total Cards: 138. SN1 was read on 20 of 138 rows for Black (Gallery pages 1 and 10) and on 10 of 138 for each of Cyan, Magenta and Yellow (Gallery page 1). The distinct token set is therefore NOT fully enumerated for any of the four. print_run=1 was written rather than the rows being removed because EXTRACTION_SPEC section 6 forbids print_run=UNKNOWN at base-set scope, SN1 is positively rendered on every row read, no contrary token appears on any plate row anywhere in the release, and the separate The Standard Printing Plates Yellow sid (354520) independently renders SN1 on 50 rows. cardboardconnection's parallel line does not mention printing plates at all, so all five plate rows are single-source (tcdb only). SEPARATE FINDING IN THE SAME AREA: tcdb lists a Printing Plates YELLOW for The Standard and no Black, Cyan or Magenta sid for that set, which is an asymmetry no source explains - three plate sids may simply be uncatalogued.
• G-009 NO CHANNEL IS ASSERTED ON ANY OF THE 58 PARALLEL ROWS, WHICH IS A DELIBERATE DIVERGENCE FROM ALL SIX SIBLING SEASONS. Every row carries channel=UNKNOWN with an EMPTY channel_basis (the shape migrations/003_parallels.sql requires) and channel_class=unknown from the published map. REASON: no source consulted states a trade channel, an exclusivity claim or a box SKU for this release in a form that could be quoted verbatim. cardboardconnection's configuration line came back as "6 cards per pack, 1 pack per box, 5 boxes per case" with "3 Autographs" and "2 Memorabilia Cards" - the word 'hobby' appeared only in the summarising fetch's own connective prose, never inside quotation marks, and EXTRACTION_SPEC 4.1 says to reject anything outside them. THE COST IS REAL AND IS STATED HERE RATHER THAN HIDDEN: immaculate-bkb-2019-20 through 2024-25 all write channel=hobby with channel_basis=unsupported - i.e. a known channel nobody sourced - so a consumer comparing seasons of this line will see this season alone report UNKNOWN. That is a difference in agent convention, not a difference between the products. SECOND ITEM: cardboardconnection's wording "Premium FOTL #/24" implies the four 'Premium Edition' tiers (Patch Autographs Premium Edition, Patch Autographs Tags Premium Edition, Rookie Patch Autographs Premium Edition, Rookie Patch Autographs Tags Premium Edition) are First-Off-The-Line channel products, which would be channel=FOTL, channel_class=promotion. That reading also did not arrive inside quotation marks and is recorded here rather than written. One verbatim re-fetch of the cardboardconnection page would settle both items for the whole release.
• G-010 THE Inserts.cfm NOTES-COLUMN TEST CANNOT BE RUN ON THIS PARENT AND WAS NOT REPORTED AS RUN. EXTRACTION_SPEC 3.2 describes a Prizm-shaped parent whose Notes column splits into annotated (print run stated) and blank (unnumbered) rows. This parent is NOT that shape: on https://www.tcdb.com/Inserts.cfm/sid/209949/2018-19-Panini-Immaculate-Collection, 94 of the 95 listed rows carry an image-completion percentage ("12.0% of images added") and exactly one carries a real serial statement - Logoman Autographs, "Serial numbered "One of One"". There are no blank cells, so a blank cell means nothing and the discriminative test has no purchase. This is the sixth documented shape for that column. Every print run in this release was therefore taken from within-sid row-level serial tokens instead, which is the documented fallback.
• G-011 SID ENUMERATION RECONCILED EXACTLY, WHICH BOUNDS THE SEARCH AND DOES NOT PROVE THE COMPOSITION. Inserts.cfm listed 87 insert-section entries and 8 parallel-section entries. Those 87 decompose without remainder into 39 base sids plus 48 set-scope parallel sids, and the 8 into 4 colour parallels plus 4 printing plates. The release owns FOUR separate sid neighbourhoods, not one: 216989-216993 (the Immaculate Ink family), 217153-217237 (everything else, 85 sids, every one accounted for with no interior gap and no foreign product interleaved), 354520 (The Standard Printing Plates Yellow) and 553824-553827 (the four base printing plates). BOTH BOUNDARIES OF BOTH MAIN BLOCKS WERE PROVED FOREIGN by ViewSet.cfm probes: 216988 = "1988-89 Panini Spanish Liga Stickers", 216994 = "2013-14 Panini Rookie Anthology - 2013-14 Panini Prizm Update: Prizms", 217152 = "2019 Panini Phoenix - Veteran Materials Super Prime Gold", 217238 = "1911-14 General Baking (D304)". The 217152 probe also satisfies CHASE_CHECKLIST_SPEC's mandatory 'probe below the bottom of the listing' check for hidden base parallels. Per that spec a balanced count can reconcile while being wrong because errors cancel, so this is recorded as a bound, not a proof. What it does NOT cover: sets tcdb files under a SIBLING product's parent, which no amount of enumeration of this parent can surface.
• G-012 rookie_flag IS UNKNOWN ON ALL 100 BASE CARDS AND THAT IS A DELIBERATE REFUSAL, NOT A COVERAGE HOLE. tcdb renders the RC token on every row of the 101-140 Rookie Patch Autographs block and on no row of the 1-100 base run, which is a within-sid discriminative signal that the base cards carry no RC designation - and the base run does contain 2018-19 rookies (card 88 Miles Bridges, card 3 Thomas Bryant). Absence of a token is not a statement of fact, and CAV-003 records that most Panini checklist sources simply do not print RC designations. All six sibling Immaculate seasons likewise carry rookie_flag=UNKNOWN on all 100 base cards, so this is consistent with the line.
• G-013 CARD 29 IS FILED AS Gary Harris / Denver Nuggets IN THE MIDDLE OF AN ORLANDO MAGIC RUN and was written verbatim rather than 'repaired'. The base run is ordered by team in reverse-alphabetical team order; cards 28, 30 and 31 are Nikola Vucevic, Aaron Gordon and Jonathan Isaac, all Orlando Magic, and 77 is Nikola Jokic (Denver). A Denver player at 29 breaks the pattern and looks exactly like a transcription slip. It was checked on a second endpoint and is what tcdb renders: https://www.tcdb.com/Checklist.cfm/sid/209949/2018-19-Panini-Immaculate-Collection?PageIndex=1 and https://www.tcdb.com/Gallery.cfm/sid/209949/2018-19-Panini-Immaculate-Collection?PageIndex=3 both render "29 - Gary Harris - Denver Nuggets - SN99". Unresolved: whether the card itself is a Panini error, a tcdb error, or simply an out-of-pattern card. Errors.cfm was not consulted because its silence would prove nothing (CAV-020) - but it is worth ONE fetch on the parent sid, since when it RETURNS rows it is decisive.
• G-014 CARD-NUMBER TOKENS THAT THE CHASE PASS MUST NOT TAKE AT FACE VALUE. (1) Logoman Autographs (sid 217175) renders the literal string "TBA" in the card-number column on at least one row (Rudy Gay) - an unwritable placeholder, not a code. (2) Several sets render BARE ORDINALS mixed with real letter codes in one column - Dual Patches Jersey Number renders 1, 3, 9, 21 alongside DP-JCJJ / DP-KHAW / DP-KIKT; Patches Jersey Number renders 1 (Jayson Tatum) alongside PJ-JHD; Sneak Peek - Tongue Logo renders 2 (Dikembe Mutombo) alongside SP-JST; Sneak Peek - Heel Logo renders 64 (Jason Kidd) alongside SP-RAL. Per CHASE_CHECKLIST_SPEC 1b that mixed shape is the dominant premium-line pattern and a single real code settles the scheme for the whole set: the coded rows are writable and the bare ordinals are not. (3) The Standard prints the digit-bearing suffixes this season is documented for - ST-LW4 (Lonnie Walker IV), ST-MB3 (Marvin Bagley III), ST-JE3 (Jacob Evans III). They are printed, not typos; do not repair them. No insert_cards.csv was written by this build - that is a separate pass.
• G-015 SET-LEVEL PRINT RUNS EXIST FOR THIS PRODUCT AND sets.csv HAS NOWHERE TO PUT THEM. Per EXTRACTION_SPEC 2.2 there is no print_run column on sets.csv and card_count must not be used for one. Figures stated by a source but not written as structured data: Logoman and Logoman Autographs and Rookie Logoman Autographs and Dual Rookie Logoman Autographs are 1/1 sets (tcdb renders SN1 on every row of each, and Inserts.cfm annotates Logoman Autographs "Serial numbered "One of One""); all four Sneak Peek sets render MEM, SN1 on every row; Team Slogans renders MEM, SN5 on all 20 rows. Those are properties of the sets, recorded here and in each set's notes in words only. The remaining sets are per-card variable on their own base rows (cardboardconnection's "#/25-99", "#/10-49" style ranges) and have no set print run to record at all.
• G-016 THE 2018-19 BASE PARALLEL LADDER IS STRUCTURALLY DIFFERENT FROM EVERY SIBLING SEASON AND THIS RELEASE CONTRADICTS THE LINE'S APPARENT PATTERN. All six Immaculate seasons already in the catalogue carry base Red at /49. 2018-19 carries base Red at /35, sourced three independent ways: tcdb renders SN35 on all 100 base rows of sid 217156; cardboardconnection states "PARALLEL CARDS: Red #/35, Blue #/25, Gold #/10, Platinum 1/1."; and a retailer listing title for a card in hand reads "2018-19 Immaculate Danilo Gallinari Base Ruby Red 3/35". Blue /25, Gold /10 and Platinum 1/1 DO match the siblings. Two further structural differences from 2019-20: this season's base parallel set includes four Printing Plates tiers that 2019-20 does not carry, and this season's Red and Gold sids extend over the Rookie Patch Autographs block at 101-140 where they carry DIFFERENT print runs (/25 and /10) from the base rows - so one tcdb set name spans two print runs at two scopes. A consumer comparing Red across seasons of this line by name alone will be wrong about 2018-19 twice over.

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