2017-18 Essentials

• G-001 CHANNEL IS UNKNOWN ON 19 OF 21 PARALLEL ROWS AND ON THE BASE RUN, DELIBERATELY. The only channel evidence found anywhere for this product is PRODUCT-level: cardboardconnection's "with an average of two autographs per Hobby box" and its configuration line "6 cards per pack, 10 packs per box, 12 boxes per case" at https://www.cardboardconnection.com/2017-18-panini-essentials-basketball-nba-cards, plus tcdb's parent Notes "Hobby set is the base set here, and the Retail set is listed as a parallel here. Hobby base cards are printed on thick stock." at https://www.tcdb.com/ViewSet.cfm/sid/171362/2017-18-Panini-Essentials. None of those speaks to any individual colour tier and none asserts exclusivity, so no tier inherited a channel and channel_basis is empty throughout, per EXTRACTION_SPEC 2.3 (channel=UNKNOWN with a non-empty basis is rejected by the database). The two exceptions are the two Retail rows, where tcdb states the trade channel in words. NOT RESOLVED: whether the seven colour tiers are hobby-only, retail-only or both. Note the product ships in BOTH channels - a Retail version exists as its own parallel - so a blanket hobby inheritance would have been positively wrong here, not merely unsupported.
• G-002 THE FLAG COLUMN IS WHOLLY EMPTY ON FIVE BASE TIERS AND THAT IS THE WEAKEST EVIDENCE IN THIS BUILD. Green (sid 176266), Orange (176267), Red (176268), Spiral (176269) and Platinum (176270) render NO token of any kind - not SN, not RC - on either of two differently-varied fetches each (Checklist.cfm, and ViewSet.cfm under a deliberately wrong slug). reference/caveats.csv CAV-036 fourth shape records that an all-blank flag column can carry ZERO information, and the both-directions split test cannot be run WITHIN any of these five sids because none of them has a single annotated row. print_run=unnumbered was written on all five because the schema forbids print_run=UNKNOWN at base-set scope and because a SAME-ENDPOINT SIBLING CONTRAST is available and positive: on Checklist.cfm, in this same product and over this same 200-card run, Silver (176271) renders SN99 on all 200 rows, Purple (176272) renders SN1 on all 200 rows and Retail (176265) renders RC and AU - so the renderer demonstrably does emit tokens for this product's tiers when they exist. WHAT WOULD SETTLE IT: cardboardconnection and checklistcenter both list these five tiers with no print run, which EXTRACTION_SPEC 2.1 says is not a statement of fact, so neither is counted; a marketplace seller title for a Green/Orange/Red/Spiral/Platinum card showing an absent or present serial would settle it, and none was sought in this build. THE DIRECTION OF RISK IS TOWARDS A FALSE VALUE: if any of the five is in fact numbered, this build has written a confident false unnumbered on it. Recorded rather than hidden.
• G-003 COUNT DISAGREEMENT, RESOLVED TO 40, WITH BOTH AGGREGATORS ON THE OTHER SIDE. tcdb states "Total Cards: 40" and renders 40 contiguous rows DD-1..DD-40 on THREE renderings - https://www.tcdb.com/Checklist.cfm/sid/176245/2017-18-Panini-Essentials-Dynamic-Duos, https://www.tcdb.com/ViewSet.cfm/sid/176245/essentials-dynamic-duos-second-probe and https://www.tcdb.com/Gallery.cfm/sid/176245/2017-18-Panini-Essentials-Dynamic-Duos?PageIndex=4 - with DD-40 (John Wall / Bradley Beal) present on all three and the numbering contiguous with no hole. cardboardconnection states "Dynamic Duos: 39 cards" at https://www.cardboardconnection.com/2017-18-panini-essentials-basketball-nba-cards and checklistcenter states "Dynamic Duos Set - 39 Cards" at https://www.checklistcenter.com/2017-18-panini-essentials/. 40 is written on the rendered-rows rule (EXTRACTION_SPEC 3.3, "Count the rendered rows"). Two aggregators agreeing on one figure that a third source refutes by enumeration is the shared-upstream signature, not corroboration. UNRESOLVED: which card the aggregators are dropping, and whether one of the 40 is in fact not issued.
• G-004 THREE INTERIOR NUMBERS ARE UNISSUED IN THE 201-240 AUTOGRAPH BLOCK, AND THE 5.5 MIRROR CHECK CANNOT BE RUN FOR THIS LINE. 222, 227 and 239 are absent from the block on TWO independently-scoped enumerations: the Hobby parent at https://www.tcdb.com/Checklist.cfm/sid/171362/2017-18-Panini-Essentials?PageIndex=3 and the Retail parallel at https://www.tcdb.com/Checklist.cfm/sid/176265/2017-18-Panini-Essentials-Retail?PageIndex=3, each rendering 37 rows over 201-240 with exactly those three missing. 200 + 37 = 237 reconciles tcdb's parent header exactly. EXTRACTION_SPEC 5.5 requires checking the NEXT season's parent for an Update block of THIS season before concluding a number is unissued. THAT CHECK IS NOT AVAILABLE HERE: Panini Essentials has no 2018-19 basketball release - specs/2017-18_SCOPE_MEMO.md section 8a admits the line scoped to 2017-18 only, there is no essentials-bkb-2018-19 in data/, and no such product appears in tcdb's season index. So the conclusion rests on the two enumerations alone and the one documented escape route could not be closed off. Errors.cfm/sid/171362 returns "No errors or variations have been added for this set yet." - per CAV-020 that silence is worthless and NO conclusion is drawn from it either way, including about a variation layer.
• G-005 2017-18 IS THIS LINE'S FLOOR AS FAR AS THIS BUILD MEASURED, AND HERE IS EXACTLY WHAT THAT DOES AND DOES NOT ESTABLISH. MEASURED NEGATIVE: https://www.tcdb.com/ViewAll.cfm/sp/Basketball/year/2016 lists NO entry whose name contains "Essentials", and the same render carries a working positive control ("2016-17 Panini Prestige", sid 146982), so the absence is a reading of the page rather than a failed fetch. This reproduces the Phase 0 finding in specs/2017-18_SCOPE_MEMO.md section 2c independently. WHAT IT DOES NOT ESTABLISH: tcdb's season index is an editorial index; section 5a of that memo records that a season scope bounded from ViewAll.cfm under-counts by at least one release because box sets are filed as subsets of a parent rather than as their own parent, so an earlier Essentials could in principle sit inside another product's Inserts.cfm and be invisible here. No 2016-17 or earlier parent was swept for that, and no aggregator index was checked below 2017-18. NO FLOOR IS CLAIMED IN release.csv and none should be inferred; section 8a explicitly claims no floor for any of the five lines admitted on 2026-09-27. SEPARATELY, A CONFUSABLE TO KEEP IN VIEW: "Essentials" is also a sub-brand BLOCK inside Panini Chronicles (named as such in reference/caveats.csv CAV-020 and CAV-021), and data/chronicles-bkb-2017-18 exists in this same season. A search for "2017-18 Panini Essentials" that returns a Chronicles Essentials card is a different object; no row of this release was taken from or written into any chronicles directory.
• G-006 PER-CARD VARIABLE PRINT RUN, INEXPRESSIBLE IN THE SCHEMA, AND INDEPENDENTLY CORROBORATED. The 20 rows carry AU on every row with SN99 on 18 and SN49 on exactly 2, enumerated over all 20 rows and confirmed on two endpoints: https://www.tcdb.com/Checklist.cfm/sid/176239/2017-18-Panini-Essentials-Claim-to-Fame-Autos and https://www.tcdb.com/ViewSet.cfm/sid/176239/essentials-claim-to-fame-second-probe (a deliberately wrong slug). The two SN49 rows are Kobe Bryant - which tcdb renders with the bare ordinal "1" as its card number and which also carries EXCH - and CF-KDR Kevin Durant. checklistcenter states the SAME exception independently: "Claim to Fame Autos Set - 20 Cards - Serial Numbered #/99 (*Bryant, Durant #/49)" at https://www.checklistcenter.com/2017-18-panini-essentials/. sets.csv has no print_run column and parallels.csv cannot hold serial_numbered=yes with print_run=UNKNOWN (EXTRACTION_SPEC 2.3), so this census exists only here and in the sets.csv note. An unhedged flat #/99 would be false on two of twenty cards.
• G-007 PER-CARD VARIABLE PRINT RUN, INEXPRESSIBLE IN THE SCHEMA, AND INDEPENDENTLY CORROBORATED. The 37 rows carry AU on every row with SN99 on 36 and SN57 on exactly one, DG-DBD Dragan Bender - enumerated over all 37 rows and confirmed on two endpoints: https://www.tcdb.com/Checklist.cfm/sid/176264/2017-18-Panini-Essentials-Destined-for-Greatness and https://www.tcdb.com/ViewSet.cfm/sid/176264/essentials-destined-second-probe (a deliberately wrong slug). checklistcenter states the SAME exception independently: "Destined for Greatness Sigs Set - 37 Cards - Serial Numbered #/99 (*Bender #/57)" at https://www.checklistcenter.com/2017-18-panini-essentials/. A flat #/99 would be false on that card. SN57 is an unusual figure and both sources give it, so it is recorded as read and not "repaired" towards a rounder number.
• G-008 PER-CARD VARIABLE PRINT RUN, WRITTEN AS UNKNOWN/UNKNOWN, AND AN UNHEDGED EDITORIAL FIGURE REFUTED. The 30 rows carry AU on every row with SN35 on 29 and SN22 on exactly one, CE-AMG Alonzo Mourning - enumerated over all 30 rows and confirmed on two endpoints: https://www.tcdb.com/Checklist.cfm/sid/176260/2017-18-Panini-Essentials-Called-to-Excellence-Gold and https://www.tcdb.com/ViewSet.cfm/sid/176260/essentials-cte-gold-second-probe (a deliberately wrong slug), which reproduces CE-AMG = "AU, SN22" and CE-KBR = "AU, SN35, EXCH". checklistcenter states a flat "Gold #/35" at https://www.checklistcenter.com/2017-18-panini-essentials/ and THAT FIGURE IS REFUTED on CE-AMG. print_run and serial_numbered are UNKNOWN/UNKNOWN and value_contested=yes: the tier is PROVABLY serial numbered with a variable run, which EXTRACTION_SPEC 2.3 says the schema cannot express, so UNKNOWN here does NOT mean "nothing was found". THIS IS THE ONLY VARIABLE TIER OF THE TWELVE ON THE THREE LADDERS. Unlike the two set-level exceptions above, checklistcenter does NOT carry the SN22 exception, so this one rests on tcdb alone (two endpoints of one source is still one source) and source_confidence is single.
• G-009 A UNIFORM PRINT RUN THAT sets.csv CANNOT HOLD. All 37 rows of the 201-240 block are flagged "AU, SN99" at https://www.tcdb.com/Checklist.cfm/sid/171362/2017-18-Panini-Essentials?PageIndex=3, and both aggregators state it in words - cardboardconnection "Rookie Autos: 37 cards, #/99" at https://www.cardboardconnection.com/2017-18-panini-essentials-basketball-nba-cards and checklistcenter "Rookie Autos Set - 37 Cards - Serial Numbered #/99" at https://www.checklistcenter.com/2017-18-panini-essentials/. Three sources, one figure, no exception found on any row. sets.csv has no print_run column (EXTRACTION_SPEC 2.2) and /99 here is the set's own uniform numbering rather than a parallel tier, so no parallels.csv row was invented to carry it: the figure lives in the sets.csv note and here. This is the state EXTRACTION_SPEC 2.2 says to record rather than force into card_count.
• G-010 SET-NAME DIVERGENCES BETWEEN tcdb AND THE AGGREGATORS, RECORDED RATHER THAN RESOLVED. tcdb's headings are used throughout, because a resolver has to match what tcdb prints and because tcdb is the structural source here. The divergences: tcdb "Called to Excellence" vs cardboardconnection "Called to Excellence Autographs" vs checklistcenter "Called to Excellence Autos Set"; tcdb "Claim to Fame Autos" vs cardboardconnection "Claim to Fame Signatures"; tcdb "Destined for Greatness" vs cardboardconnection "Destined for Greatness Signatures" vs checklistcenter "Destined for Greatness Sigs Set"; tcdb "True Potential" vs cardboardconnection and checklistcenter both "True Potential Signatures"; tcdb "Kobe's All-Rookie Team" and "Rock the Rim" vs checklistcenter's "Kobe's All-Rookie Team Die-Cut Set" and "Rock the Rim Die-Cut Set". No name was changed and no set was merged or split to fit a source. SEPARATELY, FOR THE CHASE PASS: tcdb renders the Kobe Bryant row of Claim to Fame Autos with the bare ordinal "1" while the other 19 rows carry CF- letter codes - the mixed-code/bare-ordinal shape CHASE_CHECKLIST_SPEC documents. Under its one-real-code-settles-the-scheme rule the scheme is CF- letter codes and that "1" is a placeholder, so the Kobe row's card_number should be withheld and the player named, not written as 1 and not "repaired" to an invented CF- suffix. insert_cards.csv was NOT written by this build - it is a separate pass under CHASE_CHECKLIST_SPEC.
• G-011 reference/brands.csv AND pipeline/fix_schema.py DISAGREE ABOUT WHETHER THIS LINE HAS A VISUAL FAMILY, AND NOTHING WAS ADDED TO EITHER. The ordered FAMILIES list in pipeline/fix_schema.py has no pattern matching this line's ladder, so family_of() sends 18 of this release's 21 parallel rows to "other": every Green, Orange, Red, Spiral, Silver, Purple, Gold and Retail row. Only the four Platinum rows (one base-set, three set-scope) land on a real family, "Platinum", and they do so because "Platinum" happens to be in the list for other lines. parallel_family was DERIVED by lifting CHANNEL_CLASS and family_of out of pipeline/fix_schema.py by AST, per CLAUDE.md and CAV-034, so the repo still holds exactly one definition, and NO entry was added to FAMILIES (CAV-039: 268 existing rows already carry hand-written families that family_of() does not produce, and minting a pattern in one release would split rather than fix them). THE CANDIDATE WORTH A DECISION IS "Spiral": it is a genuine printed finish name, not a colour and not a box format, and it is exactly the shape the list already holds for Swirl, Velocity and Shimmer. "Silver", "Gold", "Purple", "Green", "Orange" and "Red" are plain colours and arguably belong in "other" by design; "Retail" is a trade channel and duplicates a fact already carried by channel=retail, which is the CAV-021 complaint in concrete form. Raised for the taxonomy owner; not acted on.
• G-012 BASE PRINT RUN IS assumed, NOT sourced, AND THE DISTINCTION IS DELIBERATE. base_print_run=unnumbered with base_print_run_basis=assumed. No source read in this build states in words that the base cards carry no serial number. The evidence is indirect but positive: on Checklist.cfm the Silver tier renders SN99 on all 200 rows of this same base run and the Purple tier renders SN1 on all 200, while no base row on either of the parent's two endpoints carries any SN token. EXTRACTION_SPEC 2.1 names exactly this evidence shape and says it is "assumed", not "sourced" - a source that says so in quotable words would be needed to promote it, and none exists. The related fact that IS sourced in words concerns the Retail parallel, not the base: tcdb's Notes for sid 176265 reads "Printed on thin stock, except for rookie autographs which are thick stock but are NOT serially numbered."
• G-013 ONE tcdb SID CARRIES TWO SCOPES, AND TWO tcdb HEADERS ON ONE PRODUCT COUNT TWO DIFFERENT THINGS. Retail is a single sid, 176265, whose header reads "Total Cards: 240" and which spans BOTH the 1-200 base run and the 201-240 Rookie Autos block. Two parallels.csv rows therefore rest on it: a base-set-scope row and a set-scope row for Rookie Autos. They are genuinely different objects - a Retail base card and a Retail rookie autograph - and both are evidenced, but a consumer counting Retail cards by summing parallel rows will double-count the sid unless it reads the scopes. THE HEADER DIVERGENCE IS THE SHARPER FINDING: the Hobby parent sid 171362 states "Total Cards: 237" and the Retail sid 176265 states "Total Cards: 240" for the SAME card population - 200 base plus 37 autos over numbers 1-240 with 222, 227 and 239 unissued. 237 counts ISSUED numbers; 240 counts the HIGHEST number. Neither header is stale and neither is wrong; they count different units, one sid apart, inside one product. This adds a sixth reading to the five units EXTRACTION_SPEC already records for Total Cards, and it is the first time the divergence has been measured BETWEEN TWO SIDS OF ONE PRODUCT rather than between a header and a row count. UNRESOLVED: whether the Retail version genuinely holds 37 or 40 rookie-autograph cards - the rendered rows say 37, matching Hobby, while its own header would allow 40.

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