2016-17 Aficionado

• G-001 CHANNEL UNRESOLVED ON ALL 20 PARALLEL ROWS. No source consulted - tcdb parent, tcdb Inserts.cfm, all 34 child sids, cardboardconnection's product page or Panini's own first-look post - states a trade channel, a box SKU, a regional edition or any exclusivity for any tier of this product. cardboardconnection gives only a product configuration ("8 cards per pack, 10 packs per box, 12 boxes per case") and Panini's blog repeats it ("eight cards per pack, 10 packs per box, 12 boxes per case"); neither says hobby, retail or anything else. Every row therefore carries channel=UNKNOWN with an EMPTY channel_basis and an EMPTY channel_confidence, which is what EXTRACTION_SPEC 2.3 requires when a channel is declined rather than unsupported. Resolving this needs a sell sheet or a pack-odds wrapper, neither of which surfaced.
• G-002 base_print_run=unnumbered carries basis=assumed, NOT sourced, and the distinction is the point. No source states in words that the 150 base cards carry no serial number. What exists is a within-product positive discriminative test: tcdb renders no serial token on any of the 150 parent rows across Checklist.cfm pages 1 and 2 and Gallery.cfm pages 8, 9, 10, 11, 12 and 15, while the SAME renderer gives SN99 on all 150 rows of Artist's Proof Red (sid 149134) and SN1 on all 150 rows of Artist's Proof Gold (sid 149135). EXTRACTION_SPEC 2.1 is explicit that exactly this evidence shape is "assumed" and that an agent should not talk themselves into "sourced". The same reasoning and the same basis apply to all three rows of base_card_tiers.csv. Resolving it needs a source that says so in quotable words.
• G-003 rookie_flag=UNKNOWN ON THE 50 ROWS 101-150, AND THE REASON IS MEASURED. tcdb renders the subset code ("GR" or "IN") in the flag column of every row of 101-150 and renders NO "RC" anywhere in that range - including on Ben Simmons (117), Buddy Hield (118), Dragan Bender (119), Juan Hernangomez (120) and Thon Maker (135), all of them 2016-17 NBA rookies whose cards in 1-100 ARE flagged RC by the same page. Confirmed on two endpoints (Checklist.cfm?PageIndex=2 and Gallery.cfm?PageIndex=11 and 12). So the column is not readable for RC in that block and an absence there is evidence of nothing; UNKNOWN is written rather than "no". In 1-100 the test runs the other way - 20 rows carry RC and every one is a genuine 2016-17 rookie - so yes/no is written there. Resolving the 50 needs a source that prints RC designations for the Global Reach and Inducted subsets.
• G-004 THE NAME OF THE 101-140 SUBSET IS CONTESTED AND WAS NOT MERGED. Three strings are in play. (1) Panini's OWN first-party copy says "Global Icons": "A 150-card base set featuring the Global Icons and Inducted subsets." (https://blog.paniniamerica.net/panini-america-provides-detailed-first-look-at-2016-17-aficionado-basketball-gallery/, published February 02, 2017), re-fetched under a second parameterisation and identical both times. (2) cardboardconnection's prose copies that - "subsets include Global Icons and Inducted" - while cardboardconnection's OWN checklist heading on the same page reads "Base Global Reach / Inducted Set Checklist" (https://www.cardboardconnection.com/2016-17-panini-aficionado-basketball), so that single page carries both strings. (3) tcdb renders the two-character token "GR". tier_name is written as "Global Reach" because the shipped-card evidence agrees with it: three independent marketplace listing titles name it and each carries a card number matching this build's own enumeration - "2016-17 Panini Aficionado Basketball Global Reach #103 Dirk Nowitzki Mavericks", "...Global Reach #138 Omri Casspi Sacramento Kings" and "2016-17 Panini Aficionado - Global Reach Nik Stauskas #127" - none of which contains a tcdb flag string, so the contamination test passes; and the 1996 Pinnacle Aficionado BASEBALL ancestor carried a "Global Reach" subset of its own. The most likely reading is that "Global Icons" is a pre-release sell-sheet name and "Global Reach" is what shipped, but NO source states that, so it is recorded as a contest rather than resolved. A consumer searching for "Global Icons" will not find this tier.
• G-005 TWO PARALLELS COVER PART OF THE BASE RUN AND THE SCHEMA CANNOT SAY HOW MUCH. Tip-Off (sid 149132, "Total Cards: 100") covers exactly the Base tier 1-100, which block scope expresses cleanly. Opening Night Preview Edition (sid 149136, "Total Cards: 80") covers EIGHTY of those hundred cards on a non-contiguous selection, which block scope does NOT express - the row says "Base" and a consumer will read it as all 100. The full 80-number list was enumerated and closes exactly against the header (75 rows from ViewSet.cfm spanning 1-93, plus 94, 95, 97, 98 and 100 from the last Gallery.cfm page), so the information exists and only the schema cannot hold it. base_card_tiers cannot be used to carve an 80-card tier either, because the tiers must tile base_set_size exactly and these 80 cards overlap the Base tier. Recorded here in full rather than written as a false whole-tier claim.
• G-006 CARD COUNT CONTESTED ON "Signatures": tcdb says 14, cardboardconnection says 13. tcdb's header reads "Total Cards: 14" and its own table renders exactly 14 rows (cards 2, 3, 4, 6, 7, 10, 11, 13, 14, 15, 16, 17, 18, 20), so tcdb is internally consistent; cardboardconnection's checklist heading reads "Signatures Set Checklist" over "13 cards." (https://www.cardboardconnection.com/2016-17-panini-aficionado-basketball). 14 is written because it is the figure that survives a rendered-row count, and EXTRACTION_SPEC 3.3 records aggregator headers disagreeing with their own lists. The cardboardconnection list itself was not counted row by row, so the contest is narrowed rather than settled.
• G-007 SEVEN SETS CARRY A PER-CARD VARIABLE PRINT RUN AND sets.csv HAS NOWHERE TO PUT IT. sets.csv has no print_run column (EXTRACTION_SPEC 2.2) and these sets have no single figure anyway, so each full enumeration produced a fact the schema cannot hold. The figures, every one of them my own count of the enumerated rows: Authentics (64 rows) MEM+SN175 x62, MEM+SN131 on card 12, MEM+SN93 on card 61. Authentics Prime (61 rows) SN25 x60, SN15 on card 24. Dual Authentics (32 rows) MEM+SN299 x30, MEM+SN80 on card 31, MEM+SN5 on card 34. Dual Authentics Prime (31 rows) MEM+SN25 x28, MEM+SN1 on card 31, MEM+SN5 on card 32, MEM+SN10 on card 34. Endorsements Autographs (50 rows) eleven distinct tokens - SN199 x18, SN149 x7, SN99 x7, SN60 x7, SN65 x5, and SN175, SN53, SN85, SN90, SN145, SN70 once each, plus EXCH on card 25. First Impressions Autographs (25 rows) SN199 x6, SN249 x19. International Ink Autographs (25 rows) SN199 x11, SN249 x7, SN60 x3, and SN59, SN70, SN149, SN99 once each. The eighth set with a single figure, Signatures, is uniformly AU+SN125 over all 14 rows and that figure is likewise recorded only in prose. This is the same expressiveness gap CAV-041 and CAV-028 describe, arriving on sets rather than on parallels.
• G-008 CHANNEL DELIBERATELY DECLINED ON "Opening Night Preview Edition" RATHER THAN INFERRED FROM ITS NAME. The name invites a launch-event, promotional or preview-distribution reading and that reading may well be right, but no source consulted says anything about how the tier was distributed - cardboardconnection files it as a base checklist ("Base Opening Night Preview Edition Set Checklist", "80 cards.") with no distribution note, tcdb's Inserts.cfm Notes cell for it is EMPTY, and Panini's first-look post does not mention it. channel=UNKNOWN with an empty channel_basis, per EXTRACTION_SPEC 2.3's rule that "unsupported" describes a KNOWN channel you could not source and is not a way to say the channel itself is unknown.
• G-009 TWO ROWS OF "Artist's Proof Bronze" SCOPED TO Endorsements Autographs WERE NEVER READ. tcdb sid 149145 states "Total Cards: 44" and cardboardconnection independently states "44 cards.", but tcdb renders only 42 rows, missing card numbers 10, 11, 12, 13, 15, 20, 25 and 49 - and the SAME 42 rows with the SAME eight absent numbers came back from a second differently-parameterised endpoint (Checklist.cfm?sort=0), so this is not a truncation a re-fetch cleared. All 42 rendered rows read "AU, SN49" and 49 is written on that basis. If either unread row carries a different token the tier is per-card variable and the written 49 is wrong - which is the one place in this release where a sampling risk of the CAV-050 shape survives. Every other tier in this release was enumerated over every row.
• G-010 tcdb's HEADER AND ITS OWN TABLE DISAGREE ON "International Ink Autographs", AND THE TABLE IS RIGHT. The header reads "Total Cards: 23"; the table renders 25 rows numbered contiguously 1-25. The first fetch stopped at 23 and said so in its own words - "Card 24 and 25 appear in the table but exceed the stated total of 23 cards" - so a reader trusting the header would have lost two cards. A second fetch instructed to ignore the header returned all 25, ending 24 Ivica Zubac and 25 Paul Zipser. cardboardconnection states "25 cards." and this product's own International Ink Artist's Proof Gold parallel (sid 149152) states "Total Cards: 25" and renders 25 rows, so three further readings agree with the table. 25 is written. Whether tcdb's 23 is stale or counts a different unit is not established.
• G-011 DERIVED parallel_family IS WEAK ON THIS LINE AND I DID NOT EDIT THE LIST. pipeline/fix_schema.py's FAMILIES carries the pattern "Artist Proof" WITHOUT the apostrophe, while every source for this product prints "Artist's Proof" with one, so family_of() misses it and 16 of this release's 20 parallel rows derive to "other". The remaining four are inconsistent rather than right: "Artist's Proof Bronze" matches the "Bronze" pattern and derives to Bronze on 3 rows, while its siblings "Artist's Proof Red", "Artist's Proof Gold" and "Artist's Proof Purple" have no colour pattern to match and derive to "other" - so one member of one visual system is filed apart from the rest. "Opening Night Preview Edition" derives to "Opening Night", which is correct. The brief instructed that no FAMILIES entry be added (CAV-039) and none was; the derivation was lifted by AST and applied as it stands. pipeline/check_families.py was run read-only after this file was written and reports zero defect candidates, so nothing here matches inside an unrelated word - the problem is a MISS, which that checker is not designed to catch. Whether an "Artist's Proof" family should exist is a taxonomy decision of the CAV-021 kind and is not mine.
• G-012 THIS RELEASE RESTS ON THREE SOURCES AND ONLY TWO LINEAGES, AND IT HAS NO SIBLING SEASON ANYWHERE. tcdb carries the structure; cardboardconnection is the ONLY editorial aggregator consulted; blog.paniniamerica.net is first-party and carried the debut statement and the base-set size. beckett.com was not attempted - CAV-033 records it 302ing to a maintenance page since at least 2026-09-20. checklistcenter, checklistinsider, sportscardchecklist and waxstat were NOT swept, so no independence test of the CAV-037 kind was run on any field of this product. EXTRACTION_SPEC 6 also requires every structural call to be compared against how the same line or a sibling product was built in an adjacent season: THAT CHECK CANNOT BE RUN AT ALL HERE. aficionado-bkb-2016-17 is the first and only release of this line in data/, 2016-17 is the catalogue floor, and the line has no basketball predecessor by its own first-party account. Every structural call in this release is therefore uncompared.
• G-013 "Authentics Prime" IS THE ONE SET IN THIS RELEASE WITH A SINGLE-SOURCE CARD COUNT. tcdb states "Total Cards: 61" and renders 61 rows; cardboardconnection has NO checklist section for the set at all, though it does carry one for Dual Authentics Prime at "31 cards.". Every other set here has two counts. Noted because an absence in one source is not a statement about the set, and because a later pass should not read this count as corroborated.
• G-014 Errors.cfm RETURNED NOTHING AND I CONCLUDE NOTHING FROM IT. https://www.tcdb.com/Errors.cfm/sid/146983/2016-17-Panini-Aficionado renders "No errors or variations have been added for this set yet." CAV-020 records that this endpoint returns that identical sentence on sets that provably DO carry UER rows, so it discriminates nothing. The row is quoted as evidence of what the endpoint said and is NOT read as evidence that this product has no variation set. What this release's clean base run does rest on is an independent contiguous enumeration: 150 numbers, 1-150, no repeat, confirmed by Gallery.cfm rendering exactly fifteen full pages of ten tiles with no page 16. A b-suffixed variation layer would not be surfaced by that enumeration and none was looked for.
• G-015 NO PRIOR-SEASON OR CROSS-PRODUCT SUBSET IS FILED UNDER THIS PARENT, AND THAT IS A POSITIVE FINDING RATHER THAN AN UNCHECKED ASSUMPTION. EXTRACTION_SPEC 5.5 requires prior-season blocks to be FILED in sets.csv rather than gapped; there are none to file. Inserts.cfm lists 34 rows in two sections, "Parallel Sets (5)" and "Insert Sets (29)", and every one of those 34 sids was opened individually in this build - 149132 through 149165 with no number skipped - and every one rendered a page heading beginning "2016-17 Panini Aficionado - ". No row carries another season's prefix and no row carries another product's name. The 34 sids also close the block exactly against its proved boundaries. THE MIRROR WAS NOT CHECKED: EXTRACTION_SPEC 5.5 also warns that THIS season's Update blocks can be filed under the NEXT season's parent, and no 2017-18 parent was searched for a 2016-17 Aficionado block. Since this build found no unissued number, nothing in this release currently depends on that check.
• G-016 A LOOSE SEARCH ON THIS LINE RETURNS TWO FOREIGN PRODUCT FAMILIES AND BOTH MUST BE KEPT OUT. First, the ancestor: Panini's own blog states "A baseball-only release that burst onto the scene with a one-season run in 1996, Aficionado stood out on the strength of its base-card aesthetics." - a 1996 Pinnacle Aficionado BASEBALL product this catalogue does not carry, and nothing was merged into it. Second, and not anticipated by the task brief: a 2016-17 Panini Aficionado SOCCER product exists and shares this product's season string, its "Global Reach" subset name and its "Artist's Proof" and "Artist's Proof Bronze" parallel names. cardboardconnection carries a "2017 Panini Aficionado Soccer Cards" page, and marketplace titles for it name Claudio Bravo #167, Christian Pulisic #184 and Cristiano Ronaldo #190 - card numbers that do not exist in this 150-card basketball set, which is the cheap discriminator. A name-keyed search on this line will pull all three families; brands.csv keys on (product_name, sport) and so is safe, but any later text-matching resolver work is not.
• G-017 PRINT RUNS ON THE 15 SET-SCOPE PARALLEL ROWS REST ON tcdb ALONE. cardboardconnection names only the four base-set tiers ("PARALLEL CARDS: Tip-Off, Artist's Proof, Artist's Proof Red #/99, Artist's Proof Gold 1/1.") and prints no figure for any insert-scope Artist's Proof tier, and Panini's first-look post names no parallel at all. Every set-scope row is therefore source_confidence=single even though each was enumerated over every rendered row. The tokens themselves were read off ViewSet.cfm for those 15 sids; CAV-042 records that ViewSet's flag column is lossy, and this build measured that loss directly - ViewSet on sid 149138 reported the flag column as "SN175" where Checklist.cfm on the same sid renders "MEM, SN175", so ViewSet dropped the MEM token on that page. A dropped SERIAL token on any of the 15 would be the dangerous direction and was not separately tested on each; the tokens that were SEEN are evidence, and no absence was written off ViewSet alone.
• G-018 A STRUCTURAL CALL THAT COULD HAVE GONE THE OTHER WAY: "Authentics Prime" AND "Dual Authentics Prime" ARE FILED AS SETS, NOT AS PARALLELS OF "Authentics" AND "Dual Authentics". The case for sets, which is what was written: tcdb lists both under its "Insert Sets (29)" heading and NOT under "Parallel Sets (5)"; each has a different card count from its parent (61 against 64, 31 against 32) and a different roster, so neither is a full parallel; and cardboardconnection carries "Dual Authentics Prime Set Checklist" over "31 cards." as a checklist section of its own. The case for parallels: Prime is a patch-quality step on the same physical card design, which is exactly what CAV-021 describes as a genuine Prime/Super Prime ladder the catalogue currently collapses, and a consumer asking "what parallels does Authentics have?" gets nothing from this release. Recorded rather than resolved, because resolving it is a corpus-wide taxonomy decision and not one release's to make.

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