2018-19 Encased

• G-001 A 119-ROW AUTOGRAPH BLOCK SITS INSIDE THE PARENT CHECKLIST PAST THE BASE RUN (EXTRACTION_SPEC 5.3), AND IT IS EXCLUDED FROM base_set_size. tcdb parent sid 205974 states Total Cards "219"; cards 1-100 carry the single flag "SN99" with no AU and no RC, and the rows after them carry AU/RC/MEM - https://www.tcdb.com/Gallery.cfm/sid/205974/2018-19-Panini-Encased?PageIndex=11 renders "101 Donte DiVincenzo | Milwaukee Bucks | AU, RC, SN75" and ?PageIndex=19 renders "190 Michael Porter Jr. | Denver Nuggets | AU, RC, SN75" immediately followed by "191 Shai Gilgeous-Alexander | Los Angeles Clippers | AU, MEM, RC, SN99". base_set_size is 100. CONVENTION CHECK AGAINST THE ADJACENT SEASON: encased-bkb-2019-20 makes the same call (base_set_size=100 against a 274-card parent), so the two Encased releases AGREE and this is not a repeat of the Contenders include/exclude divergence EXTRACTION_SPEC 5.3 records. Recorded because the spec asks for the comparison to be made explicitly, not because it disagrees.
• G-002 THE PARENT RENDERS 219 ROWS OVER CARD NUMBERS 1-223, SO FOUR NUMBERS ARE UNISSUED - ALL OF THEM INSIDE THE 101-223 AUTOGRAPH BLOCK, NONE INSIDE THE BASE RUN. Card 220 is directly evidenced as unissued: the last rendered gallery page https://www.tcdb.com/Gallery.cfm/sid/210632/2018-19-Panini-Encased-Red?PageIndex=22 shows nine tiles reading 214, 215, 216, 217, 218, 219, 221, 222, 223 with no 220, reproduced on a second differently-parameterised fetch of the same page (&sort=0) so it is not the documented dropped-tile artefact, and ?PageIndex=23 renders no tiles at all. The other three unissued numbers are LOCATED BUT NOT IDENTIFIED: Gallery paginates by card INDEX, and the index-to-number offset is +1 at ?PageIndex=13 (indices 121-130 render numbers 122-131) and +3 at ?PageIndex=19 (indices 181-190 render numbers 184-193), so exactly one number between 101 and 121 and exactly two between 131 and 183 were never issued. Which ones was not established and would cost roughly six more Gallery fetches. THE BASE RUN IS UNAFFECTED AND IS CONTIGUOUS 1-100, enumerated on two endpoints with an overlap that agrees row for row. No base_card_tiers.csv was written: the base run has no tiers at all, so there is nothing to tile.
• G-003 THE INTERNAL BOUNDARIES BETWEEN THE THREE 29-CARD ROOKIE SIGNATURE SETS ARE NOT EVIDENCED BY ANY SOURCE. Cards 101-190 form one autograph-only band in which every rendered row reads "AU, RC, SN75" (confirmed at Gallery ?PageIndex=11, ?PageIndex=13 and ?PageIndex=19 on https://www.tcdb.com/Gallery.cfm/sid/205974/2018-19-Panini-Encased ). That band holds exactly 87 rows - 90 numbers less the three unissued ones of G-002 - which is exactly 29+29+29, the counts cardboardconnection and checklistcenter give for Rookie Endorsements, Rookie Scripted Signatures and Rookie Notable Signatures. The ARITHMETIC therefore corroborates the three counts, but nothing states where one set stops and the next begins, and the flag column is identical across all 87 rows so it cannot be used to split them. Rookie Jersey Autographs is the one in-parent set whose extent IS evidenced at both ends (191-223, less the unissued 220, = 32). Same shape as gaps G-002 in encased-bkb-2019-20, which could not split its 131-190 band either.
• G-004 A SINGLE tcdb PARALLEL SID CARRIES TWO PRINT RUNS ACROSS CARD-NUMBER RANGES, AND THE BASE-SCOPE ROWS STATE ONLY THE BASE-RUN FIGURE. All four sids in tcdb's "Parallel Sets (4)" section state Total Cards "219" and cover the whole parent, base run PLUS the 101-223 autograph block. The Red sid renders "SN35" on every one of cards 1-100 and "SN25" from card 101 on - https://www.tcdb.com/Gallery.cfm/sid/210632/2018-19-Panini-Encased-Red?PageIndex=10 ("100 Nikola Vucevic | SN35") against ?PageIndex=11 ("101 Donte DiVincenzo | AU, SN25") and ?PageIndex=22 ("223 Omari Spellman | AU, MEM, SN25"). The four base-set-scope rows in parallels.csv therefore carry 35/10/5/1, which is the base-run figure and is exactly what cardboardconnection states in words. WHAT IS NOT WRITTEN: the colour tiers of the 101-223 block. They are real - Red on that block is provably SN25 - but a row for them would have to be scoped to one of the four in-parent sets, and G-003 says three of those four cannot be delimited; and the Gold, Green and Platinum tokens on the block were never enumerated (only their cards 1-100 were). Omitted rather than guessed. Roughly eight Gallery fetches would settle it.
• G-005 TWO SET COUNTS DISAGREED BETWEEN tcdb AND THE AGGREGATORS AND WERE SETTLED BY A THIRD PAGE INSIDE tcdb. tcdb's Endorsements base page states Total Cards "43" and renders 43 rows - self-consistent, which EXTRACTION_SPEC warns is not evidence - while cardboardconnection ("Endorsements Autographs (45 cards)") and checklistcenter ("45 Cards - Serial Numbered as Noted") both say 45. Its own Blue FOTL parallel page states 45 and renders 45 rows, and its Bronze FOTL page states and renders 44, so the base page is SHORT and 45 was written. Identically for Scripted Signatures: base page 45/45, aggregators 48, Blue FOTL page states 48 and renders 48. Both counts carry a dated marker in sets.csv. WHAT REMAINS OPEN: the two aggregators are ONE editorial lineage (CAV-033), so "two sources agree" here is one opinion plus a tcdb parallel page, and the parallel page is itself a roster of a scarcer tier that need not equal the base set. On both sets the SCARCEST tier (#/3 Blue FOTL) is the most complete rendering tcdb holds - the CHASE_CHECKLIST_SPEC Recon finding, reproduced here on a second line.
• G-006 SEVEN SET-SCOPE TIERS ARE PROVABLY SERIAL NUMBERED WITH A PER-CARD VARIABLE RUN, AND THE SCHEMA CANNOT SAY SO. EXTRACTION_SPEC 2.3 forces print_run and serial_numbered to move together, so all seven are written UNKNOWN/UNKNOWN, which is indistinguishable from "we found nothing" - the collapse unknown_semantics.csv exists to prevent. The distinct token sets, each enumerated over every rendered row rather than sampled, are: Endorsements Red {AU, SN25, SN15} over 41 of 41 rows (https://www.tcdb.com/ViewSet.cfm/sid/210640/2018-19-Panini-Encased-Endorsements-Red); Scripted Signatures Red {AU, SN15, SN25} over 42 of 42 (https://www.tcdb.com/ViewSet.cfm/sid/210658/2018-19-Panini-Encased-Scripted-Signatures-Red); Jerseys Prime {MEM, SN25, SN18, SN13, SN12, SN11, SN8, SN7, SN5} over 43 of 58 (https://www.tcdb.com/ViewSet.cfm/sid/210642/2018-19-Panini-Encased-Jerseys-Prime); Materials Prime {SN25, SN5, SN2} over 47 of 59 (https://www.tcdb.com/ViewSet.cfm/sid/210646/2018-19-Panini-Encased-Materials-Prime); Rookie Materials Prime {MEM, SN25, SN3, SN2, SN1} over 39 of 39 (https://www.tcdb.com/ViewSet.cfm/sid/210650/2018-19-Panini-Encased-Rookie-Materials-Prime); Substantial Swatches Prime {SN25, SN10} over 20 of 20 (https://www.tcdb.com/ViewSet.cfm/sid/210665/2018-19-Panini-Encased-Substantial-Swatches-Prime); and Substantial Swatches Rookies Prime {MEM, SN25, SN10} over 29 of 29 (https://www.tcdb.com/ViewSet.cfm/sid/210667/2018-19-Panini-Encased-Substantial-Swatches-Rookies-Prime). That is 7 of the 26 set-scope tiers on this release, a 27% per-card-variable rate. checklistcenter independently describes the Jerseys Prime tier in words as "58 Cards - Serial Numbered #/25 or as Noted" - a ceiling plus a variable, which is not a print run and was not written as one. HAD ANY OF THESE BEEN SAMPLED at three rows, five of the seven would have been written as a clean sourced "25".
• G-007 ONE TIER'S TOKEN ENUMERATION IS INCOMPLETE AND ITS PRINT RUN IS WRITTEN ANYWAY. Scripted Signatures Bronze FOTL states Total Cards "48" and only 40 rows survived the fetch (https://www.tcdb.com/ViewSet.cfm/sid/210653/2018-19-Panini-Encased-Scripted-Signatures-Bronze-FOTL), so the token set {AU, SN12} covers 40 of 48 rows, not all of them. 12 was written because every one of the 40 reads SN12 and because the sibling Bronze FOTL on Endorsements reads SN12 on all 44 of its 44 rows. This is the weakest print run in the file: on this very release two tiers that looked uniform over their first forty rows turned out variable, so 8 unread rows are a real exposure. Topping it up needs one Gallery fetch of the tail. Two further tiers are enumerated over a majority but not all rows and are written UNKNOWN anyway, so they carry no exposure: Jerseys Prime (43 of 58) and Materials Prime (47 of 59). [CORRECTED 2026-09-22: the token enumeration for this tier is now COMPLETE over all 48 rows, not 40. https://www.tcdb.com/Gallery.cfm/sid/210653/2018-19-Panini-Encased-Scripted-Signatures-Bronze-FOTL?PageIndex=4 renders ten tiles (SC-KKZ Kyle Kuzma through SC-RTJ Rudy Tomjanovich) and ?PageIndex=5 renders the final eight (SC-SSK Scott Skiles through SC-XMD Xavier McDaniel), every one of the eighteen reading "AU, SN12". With the 40 rows the inline checklist already gave, the distinct token set over all 48 rows is exactly {AU, SN12} and the print run 12 no longer rests on an unread tail.] The gap stays OPEN for the two tiers it also names - Jerseys Prime (43 of 58) and Materials Prime (47 of 59) - whose enumerations are still partial; both are written UNKNOWN so they carry no exposure, but neither token set is complete. With Bronze FOTL closed, the weakest remaining print run in this file is Endorsements Purple at /15: 36 of 36 rows enumerated on one endpoint only, on a tier no source outside tcdb catalogues.
• G-008 SIX PRIME TIERS HAVE A ROSTER THAT DOES NOT MATCH THE SET THEY PARALLEL, AND NO SOURCE EXPLAINS WHY. Measured off each tier's own tcdb Total Cards against its parent set's: Rookie Jerseys Prime 37 against 40; Rookie Materials Prime 39 against 40; Substantial Swatches Rookies Prime 29 against 30; Vaulted Veteran Material Signatures Prime 22 against 26; Materials Prime 59 against 60; and - running the other way - Jerseys Prime 58 against 50. A prime-patch tier covering only the players whose swatch yields a prime piece explains the five that are SMALLER; it does not explain Jerseys Prime being EIGHT CARDS LARGER than the Jerseys set it parallels, and both aggregators agree on that pair of figures (cardboardconnection "Jerseys Relics (50 cards)" and "Jerseys Prime (58 cards)"; checklistcenter "Jersey Relics Set - 50 Cards" and "Jersey Prime Relics Set - 58 Cards"). Written as tcdb renders them; the larger-than-parent case is unexplained. Compare gaps G-012 in encased-bkb-2019-20, which found the same shape in the SMALLER direction only.
• G-009 CLASSIFICATION DIVERGENCE BETWEEN THIS BUILD AND BOTH AGGREGATORS ON THE PRIME TIERS. cardboardconnection and checklistcenter file "Jerseys Prime (58 cards)" / "Jersey Prime Relics Set - 58 Cards" as a SET in its own right, alongside "Jerseys Relics (50 cards)", and list no Prime tier for any of the other seven memorabilia and signature sets that have one on tcdb. This build files all eight Prime tiers as set-scope rows in parallels.csv and none of them in sets.csv, following encased-bkb-2019-20, which files "Label Materials Prime" and its siblings the same way. The consequence is visible: sets.csv has 16 rows where a reader counting the aggregators' headings would expect 17. Recorded rather than resolved - the aggregators are inconsistent with themselves here (a Prime tier is a set for Jerseys and nothing at all for the other seven), which is why their treatment was not followed.
• G-010 SOURCE DEPTH IS ONE SHALLOWER THAN IT LOOKS. beckett.com was not consulted: CAV-033 records it as 302-redirecting to a maintenance page since at least 2026-09-20 and four agents logged it independently, so no fetch was spent re-proving it. That removes the third structural source, and cardboardconnection (https://www.cardboardconnection.com/2018-19-panini-encased-basketball-nba-cards) and checklistcenter (https://www.checklistcenter.com/2018-19-panini-encased-basketball/) agree so closely on this release - identical counts on all sixteen non-base sets - that they are best read as ONE editorial lineage. Wherever a note here says "both aggregators agree", read one opinion. The genuinely independent inputs on this release are tcdb (structure, rosters, every serial token) and gogts.net (https://gogts.net/2018-19-panini-encased-nba-basketball-cards-checklist/), which corroborated only the release date and the box contents.
• G-011 NO PACK ODDS EXIST FOR ANY SET IN THIS RELEASE. Panini publishes no numeric odds for its basketball products (CAV-015), and the only published figures are box-level averages - cardboardconnection's "1 Graded Autograph and 1 Additional Autograph" plus "2 Memorabilia Cards" per box, and gogts.net's "delivers (2) Autographed Cards including (1) Graded Autographed Card and (2) Memorabilia Cards in Every Box!" - which are not per-set 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 left EMPTY rather than UNKNOWN, meaning the scarcity pass never covered these rows (CAV-005).
• G-012 rookie_flag=no ON ALL 100 BASE CARDS RESTS ON A WITHIN-PRODUCT POSITIVE CONTROL, NOT ON A SOURCE THAT PRINTS RC DESIGNATIONS. tcdb renders no RC token on any of cards 1-100 while rendering "RC" on every row from 101 onward in the SAME flag column of the SAME parent (https://www.tcdb.com/Gallery.cfm/sid/205974/2018-19-Panini-Encased?PageIndex=10 against ?PageIndex=11), so the absence is discriminative rather than an unresolved blank - the strongest form of the test EXTRACTION_SPEC 6 describes. It is still ONE source: no second source publishes RC designations for this base set, and the roster being entirely veterans was not independently confirmed. CAV-003 records that rookie_flag=UNKNOWN is the normal state for Panini releases, so a full set of "no" is worth re-checking if a second source ever appears. Supporting evidence that the base run holds no 2018 draft class: the 2018-19 rookies (Doncic, Ayton, Trae Young, Bagley) appear only from card 101 onward and in the coded chase sets, never in 1-100.
• G-013 THE 2018-19 SCOPE MEMO FILES ENCASED UNDER A HEADING READING "No `brands.csv` row (11)" AND THAT HEADING IS WRONG FOR THIS LINE TOO. reference/brands.csv carries brd_01M2B8M6BJKF4BVCZ1A77FRZ25, Encased, encased, Panini Encased Basketball, basketball, Panini - and data/encased-bkb-2019-20 was already built against it before this release existed. This is the SEVENTH of the eleven rows under that heading to be disproved, after Contenders Optic, Absolute Memorabilia, Dominion, Impeccable, Cornerstones and Status. The 2018-19 product_name matches that row exactly once the season prefix and cardboardconnection's trailing "Cards" are dropped, so nothing was minted and there is no brand-identity question to decide. The memo body already marks the heading unusable; this row records the seventh instance so the count is right.
• G-014 STRUCTURE REVERSED AGAINST THE ADJACENT SEASON ON THREE AXES, AND NOTHING WAS INHERITED. (1) BASE PARALLEL COUNT: four here (Red, Gold, Green, Platinum - tcdb heads its own section "Parallel Sets (4)" and cardboardconnection prints "Red #/35, Gold #/10, Green #/5, Platinum 1/1.") against SIX in encased-bkb-2019-20 (Purple, Red, Blue, Gold, Green, Platinum). Purple and Blue exist in 2018-19 only at INSERT scope, on Endorsements and Scripted Signatures. (2) PRINT RUN: Red is /35 here and /25 in 2019-20; the /35 slot belongs to Purple there. An agent carrying the sibling ladder across would have written 25 for Red and invented a Purple and a Blue at base scope. (3) SET INVENTORY AND PREFIXES: the memorabilia sets are wholly renamed (Jerseys / Materials / Rookie Jerseys / Rookie Materials here; Label Materials / Rookie Label Materials / Legendary Swatches there), and Scripted Signatures is coded SC- here where 2019-20 codes it SS- - in 2018-19 SS- belongs to Slabbed Signatures instead, so a resolver keyed on prefix across seasons mis-hits. WHAT DID NOT REVERSE: base_set_size 100 and base_print_run 99 are the same in both seasons, and both exclude an in-parent autograph block.
• G-015 NO VARIATION-SET CHECK WAS POSSIBLE AND THE ABSENCE OF ONE IS UNSUPPORTED RATHER THAN DISPROVED. https://www.tcdb.com/Errors.cfm/sid/205974/2018-19-Panini-Encased returns "No errors or variations have been added for this set yet." Per CAV-020 that sentence is worthless as evidence: the endpoint returns the identical string on a set that provably HAS errata (sid 182809, card 260 Mo Bamba) and on one that appears not to, so it discriminates nothing. It is cited here as a fetch made and NOT used to conclude anything. WHAT SURVIVES WITHOUT IT: base_set_size, the contiguity of cards 1-100, and the boundary at 100/101, all of which rest on the independent contiguous enumeration on two endpoints with an agreeing overlap. WHAT DOES NOT: the claim that no b-suffixed variation set exists. A variation set renders as duplicate base numbers that a contiguous enumeration is not designed to surface, and it would also be invisible in the parent Total Cards here, since 219 already reconciles exactly to 100 base rows plus the 119-row block.
• G-016 TWO ROW-LEVEL ANOMALIES RECORDED AS THE PAGES RENDER THEM, NOT REPAIRED. (1) On Jerseys Prime (https://www.tcdb.com/ViewSet.cfm/sid/210642/2018-19-Panini-Encased-Jerseys-Prime) one row - Elfrid Payton - renders "MEM" with NO serial token at all while its neighbours render "MEM, SN25". A missing token on one row of a renderer that is documented to drop flag tokens in both directions is not evidence that the card is unnumbered; the tier is UNKNOWN on other grounds anyway. (2) On Substantial Swatches Prime (https://www.tcdb.com/ViewSet.cfm/sid/210665/2018-19-Panini-Encased-Substantial-Swatches-Prime) one row reads "DW-DDR | DeMar DeRozan" on a checklist whose other nineteen rows all carry the SW- prefix - the CHASE_CHECKLIST_SPEC "another set's letter code on this set's checklist" shape. Neither was altered. A later card-level pass must not "fix" either one.
• G-017 parallel_family IS "other" ON 27 OF 30 ROWS AND NO FAMILY WAS MINTED. The three non-"other" values are derived, not judged: the repo's own ordered FAMILIES list in pipeline/fix_schema.py maps "Platinum" to Platinum, which fires on the three Platinum rows. Everything else - Red, Gold, Green, Purple, Blue FOTL, Bronze FOTL, Prime - falls through to "other", which is correct: this line publishes no visual family name anywhere. tcdb, cardboardconnection and checklistcenter all name the tiers by colour or by "Prime", and there is no Encased equivalent of Prizm's "Prizm" or Certified's "Mirror", so minting one would have invented a name the product does not use. This matches encased-bkb-2019-20, whose gaps G-017 reached the same conclusion. NOTE the open taxonomy question in CAV-021: Prime/Super Prime is a genuine patch-quality ladder that the catalogue collapses into "other" across 217 rows, and minting a Prime family in this one release would split those rows rather than fix them - so it was deliberately not minted here either. [CORRECTED 2026-09-22: the tally in this row's prose is wrong and the derived values are re-counted here. parallel_family is "other" on 24 of the 30 rows, not 27; "Platinum" on FOUR rows, not three (base-set scope plus Endorsements, Scripted Signatures and Slabbed Signatures); and "Bronze" on TWO - the Bronze FOTL tiers of Endorsements and Scripted Signatures, where the ordered FAMILIES list matches the substring "Bronze" inside "Bronze FOTL". Every value is still purely derived from pipeline/fix_schema.py and none was judged, so the conclusion that no family was minted is unchanged. But the Bronze case is worth flagging to the CAV-021 taxonomy question the row already raises: "Bronze FOTL" is a First-Off-The-Line BOX tier, not a visual finish, so assigning it the visual family "Bronze" is exactly the category error CAV-021 records for threads-bkb-2018-19's "Premium". It was left as the repo's own rule produces it rather than overridden in one release.]
• G-018 CARD-LEVEL CHECKLISTS FOR THE INSERT AND AUTOGRAPH SETS ARE OUT OF SCOPE FOR THIS PASS (CHASE_CHECKLIST_SPEC) AND insert_cards.csv WAS DELIBERATELY NOT WRITTEN. What a later chase pass can take for free, gathered incidentally while sizing the sets: the prefix inventory is Endorsements EN-, Scripted Signatures SC-, Slabbed Signatures SS-, Legendary Swatch Signatures LS-, Vaulted Veteran Material Signatures VV-, Jerseys J-, Materials M-, Rookie Jerseys RJ-, Rookie Materials RM-, Substantial Swatches SW-, Substantial Swatches Rookies SR-. Digit-bearing name suffixes are PRINTED and must not be repaired: "SS-MB3" (Marvin Bagley III), "SR-MB3", "SR-LW4" (Lonnie Walker IV), "SR-JE3" (Jacob Evans III) - the same encoding the 2018-19 scope memo records for Prizm. On Endorsements and Scripted Signatures the MOST COMPLETE roster tcdb holds is the Blue FOTL parallel page, not the base page. The four in-parent sets at 101-223 are numeric, not letter-coded, and any plain integer written onto them needs marketplace corroboration first.

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