2016-17 Prestige

• G-001 NO FLOOR IS CLAIMED FOR THIS LINE AND NONE MAY BE INFERRED FROM THIS RELEASE. specs/2016-17_SCOPE_MEMO.md section 8a decision 3 extends the prestige admission from "2017-18 only" to 2016-17 and states "No floor is claimed for either, and none should be inferred from the seasons present." That memo's section 3 records, from tcdb's own season indexes, a 2015-16 Panini Prestige at sid 119634 and a 2014-15 at sid 100261, and notes that 2014-15 carried THREE concurrent Prestige SKUs - Prestige, Prestige Plus (100287) and Prestige Premium (100324). This build did not chase depth and learned nothing new about it; it opened no 2015-16 or 2014-15 page. What it did learn incidentally is that cardboardconnection's own copy for THIS release describes the line as continuous through the preceding years under a different channel - "Exclusive to Target over the last few years" (https://www.cardboardconnection.com/2016-17-panini-prestige-basketball) - so the line ran without a break below this season and the depth question is a channel question as much as a season count. Per the Studio lesson in section 3b of that memo, a floor needs a positive statement and there is none here.
• G-002 RELEASE DATE 2017-03-08 IS DOUBLE-SOURCED AND THE OBVIOUS SHORTCUT WOULD HAVE WRITTEN A DIFFERENT DATE. Both values were taken off PRODUCT pages, not off a season index. tcdb's parent renders the unqualified line "Release Date: March 8, 2017" (https://www.tcdb.com/ViewSet.cfm/sid/146982/2016-17-Panini-Prestige); cardboardconnection's product page renders "Release Date: 3/8/2017" (https://www.cardboardconnection.com/2016-17-panini-prestige-basketball). A third rendering inside tcdb agrees: Gallery.cfm describes the set as "released in Mar 8, 2017" (https://www.tcdb.com/Gallery.cfm/sid/146982/2016-17-Panini-Prestige?PageIndex=10). THE TRAP, CONFIRMED: specs/2016-17_SCOPE_MEMO.md section 1a records that cardboardconnection's SEASON INDEX row for this product prints "Apr 11, 2017" - a month and three days off its own product page - while the same index is exactly right for Revolution and Impeccable. Taking the date off the season index would have written a wrong date that looked perfectly sourced, and would have done so on the one product of the three where the index is wrong. This build did not open the season index at all; it went to the product page, as that memo instructs. Neither source labels its date an estimate, so EXTRACTION_SPEC 3.2's estimated-date prohibition never had to fire. The two lineages are independent by publication (a hobby database record and an editorial product page) but they agree, so this is corroboration rather than a contest.
• G-003 base_print_run=unnumbered CARRIES basis=assumed, NOT sourced, because no source states in words that the base cards carry no serial numbering. What supports it is a within-product positive discriminative test: tcdb renders no serial token on any of the 200 base rows across Checklist.cfm pages 1 and 2 under two sorts and Gallery.cfm pages 8, 9, 10, 18, 19 and 20, while the SAME endpoint renders SN10, SN5, SN75 and SN1 on every rendered row of four base-set-scope parallels of the same product. EXTRACTION_SPEC 2.1 is explicit that this evidence shape is 'assumed' and that talking oneself into 'sourced' is the error. https://www.tcdb.com/ViewSet.cfm/sid/146982/2016-17-Panini-Prestige
• G-004 THE HEADLINE FINDING, RE-MEASURED FROM SCRATCH, WITH A DENOMINATOR. Every tier in this release was enumerated for its DISTINCT serial-token set over EVERY row; nothing was sampled. SIX tiers carry a per-card variable run. FIVE are autograph base tiers, all filed in sets.csv because sets.csv cannot hold a print run: Distinctive Ink 25 rows {SN199, SN149, SN75}; NBA Passport Signatures 25 rows {SN99, SN149}; Preeminent Ink 40 rows {SN199, SN99, SN49}; Old School Signatures 20 rows {SN199, SN99, SN49}; Stars of the NBA Signatures 50 rows {SN199, SN99, SN49, SN10}. Each distinct token set was reproduced on TWO endpoints, Checklist.cfm and ViewSet.cfm, per CAV-042. The SIXTH is a parallels.csv row and is written UNKNOWN/UNKNOWN per EXTRACTION_SPEC 2.3: True Colors Materials Prime, 31 rows, {SN25, SN6}, with the outlier quoted in full from ViewSet.cfm as "2 | Al Horford | MEM, SN6 | Boston Celtics". DENOMINATORS, stated rather than implied: 6 of 67 enumerated tiers (40 parallels.csv rows plus 26 insert/auto sets plus the base run); 1 of 40 parallels.csv rows; 0 of 8 base-set-scope parallels; 5 of 7 autograph base tiers, and 5 of 5 among those autograph base tiers that carry any serial token at all (Bonus Shots Signatures and Prestigious Premieres carry AU and no serial token on any row on two endpoints). PHASE 0's 4-OF-4 SURVIVES: all four tiers it probed were re-measured here on fresh fetches and the distinct token set matched its reading exactly on all four. A fifth variable autograph base tier, Stars of the NBA Signatures, was found that Phase 0 never opened. THE SCHEMA COST: five of the six findings can only be recorded in prose, because sets.csv has no print_run column, and the sixth degrades to UNKNOWN/UNKNOWN. This is the same expressiveness gap CAV-041 records; note that all six sit at set scope, where UNKNOWN is at least legal - had any been base-set scope there would have been no legal row at all. https://www.tcdb.com/Checklist.cfm/sid/148274/2016-17-Panini-Prestige-Distinctive-Ink https://www.tcdb.com/ViewSet.cfm/sid/148289/stars-of-the-nba-signatures-second-probe https://www.tcdb.com/ViewSet.cfm/sid/148293/true-colors-materials-prime-second-probe
• G-005 TWO CORRECTIONS TO THE ENUMERATION AS THE TASK BRIEF DESCRIBED IT, BOTH MEASURED. (1) Inserts.cfm ON THIS PARENT SILENTLY DROPPED TWO ROWS. Its own heading reads "Insert Sets (58)" and the rows it rendered across four differently-parameterised fetches summed to 56 by my own count. The two missing rows are "Reminiscent" (sid 148283) and "Reminiscent Rainbow" (sid 148284), both of which fall alphabetically between Prestigious Pros and Rookie Class and neither of which appeared in any fetch. They were recovered by probing the numeric sid gap and confirmed against each page's own heading - "2016-17 Panini Prestige - Reminiscent" Total Cards 15, and "2016-17 Panini Prestige - Reminiscent Rainbow" Total Cards 15. Comparing my own row count against the section heading is what caught this; the page otherwise looked complete. https://www.tcdb.com/Inserts.cfm/sid/146982/2016-17-Panini-Prestige (2) THE CHILD SID RANGE IS NOT CONTIGUOUS. The task brief states children occupy 148214-148293. They do not occupy it continuously: sids 148246-148259 are a FOREIGN ISLAND of fourteen sids inside the range, proved at both its ends - 148246 renders "1994 SunSet II - Special" Total Cards 15 and 148259 renders "1994 SunSet II - Kaantokortit" Total Cards 10. 148214-148245 plus 148260-148293 is 66 sids, which is exactly 58 insert rows plus 8 parallel rows, and the arithmetic closes with nothing hiding. (3) BOTH OUTER BOUNDARIES PROVED, including the one the brief flagged: 148213 renders "1999 Fleer Tradition Spectra Star Kites Cards" Total Cards 6 (foreign), and 148294 renders "2016 Panini Encased - Gold" Total Cards 140, sport FOOTBALL - a set carrying a brands.csv line name (encased) that a sweep upward would read as a plausible Panini release. It is not in anything here. (4) PROOF THE LISTING WAS REACHED TO THE END: the page renders Insert Sets FIRST and Parallel Sets SECOND, and all 8 Parallel Sets rows plus the footer were rendered, so the tail of the page was reached. It did not truncate here, unlike the National Treasures parent this season.
• G-006 THE ONE SET FAMILY IN THIS RELEASE THAT WAS NOT ENUMERATED OVER EVERY ROW, named rather than glossed. Bonus Shots Signatures states Total Cards 100. Checklist.cfm rendered 50 rows under the default sort (cards 1-50) and 50 again under ?sort=0 (spanning 1-77, so interior rows were dropped rather than a tail cut), and ChecklistByLastName.cfm rendered 83 rows in a surname ordering. 83 of 100 rows were reached; all 83 carry AU and none carries a serial token. Its Gold parallel states Total Cards 99 and rendered 49 rows spanning 1-73, all SN10; 50 of its 99 rows were not reached. WHAT IS STILL POSSIBLE: a second distinct serial token on one of the unreached rows of either tier, which would make one or both variable and would change the rate in G-004. The Gold row is written as a flat 10 because tcdb's Inserts.cfm Notes cell states "Serial numbered to 10" for the tier as a whole and 49 of 49 reached rows agree; that is the weakest print run in this file. The base tier is a sets.csv row and carries no print run either way. https://www.tcdb.com/ChecklistByLastName.cfm/sid/148272/2016-17-Panini-Prestige-Bonus-Shots-Signatures https://www.tcdb.com/Checklist.cfm/sid/148273/2016-17-Panini-Prestige-Bonus-Shots-Signatures-Gold
• G-007 THREE COUNT QUESTIONS, EACH TREATED AS A QUESTION ABOUT THE UNIT RATHER THAN AS PROOF A HEADER IS STALE. (1) True Colors Materials: its own page states "Total Cards: 30" and renders 31 rows over contiguous numbers 1-31, and its True Colors Materials Prime parallel states "Total Cards: 31" and renders 31. WROTE 31 - two renderings and one header against one header. This is the CAV-044 shape with the parallel page opened, as CAV-044 requires. (2) Freshman Fabrics Jumbo Prime: states "Total Cards: 37" and renders numbers 1-38 with 9 absent, i.e. 37 issued numbers. The header counts distinct issued numbers and RECONCILES EXACTLY; it is not wrong. No unissued number is recorded from this, because CAV-044 forbids reading a hole in one page as an unissued number. (3) Freshman Fabrics base renders 39 rows while its Jumbo Prime parallel holds 37 issued numbers, so a parallel's roster is not its base tier's roster in this release either - the same shape data/prestige-bkb-2017-18 recorded in its own G-008 for a different season, found again here independently. Every count in sets.csv came from that set's own page; none was carried across a tier. https://www.tcdb.com/Checklist.cfm/sid/148292/2016-17-Panini-Prestige-True-Colors-Materials https://www.tcdb.com/Checklist.cfm/sid/148277/2016-17-Panini-Prestige-Freshman-Fabrics-Jumbo-Prime
• G-008 THE CHANNEL QUESTION, ANSWERED AT RELEASE LEVEL AND DECLINED AT ROW LEVEL. WHAT THE EVIDENCE SAYS: this release shipped to BOTH hobby and a non-hobby channel. The sentence relied on is cardboardconnection's opening line, "Exclusive to Target over the last few years, 2016-17 Panini Prestige Basketball sees the NBA brand finally return to the hobby lineup." (https://www.cardboardconnection.com/2016-17-panini-prestige-basketball) Read carefully, its Target clause is about the LINE'S HISTORY - the years BEFORE this one - and its second clause is what it says about THIS release: that hobby exists for it. That hobby is not the whole story is proved from the same page, which qualifies six sets as hobby-only: "Hobby-only autographs include Distinctive Ink along with rookie-focused NBA Passport and Prestigious Premieres autographs", "Featuring holo board, hobby-only Playmakers and Prestigious Pros inserts are each a case hit", and "Prestigious Picks is a hobby-only case hit showcasing 2016 rookies." A hobby-only qualifier is vacuous unless a non-hobby channel exists for the same release, so a retail channel exists. No source reached names it, and Target is NOT written for this season - the quote places Target in the preceding years. WHAT WAS WRITTEN: channel=UNKNOWN with channel_basis EMPTY on all 40 parallels.csv rows, and channel_confidence empty. WHY: per CAV-010 an exclusivity claim earns sourced-exclusive only when it is about the thing being filed, and not one of these sentences is about a PARALLEL. The hobby-only sentences name insert and autograph SETS, and sets.csv has no channel column, so those claims are recorded in those sets' notes in words. Promoting a set-level hobby-only claim onto that set's colour tiers would be inheritance the quote does not cover, which EXTRACTION_SPEC 2.3 names as the error it exists to prevent. THE INFORMATION LOSS IS REAL AND IS RECORDED HERE: six sets carry a genuine sourced exclusivity claim that the schema has nowhere to put.
• G-009 THE FOUR UNNUMBERED BASE-SET PARALLELS WERE NOT SEEN OVER ALL 200 ROWS EACH, AND THE EXACT COVERAGE IS RECORDED HERE RATHER THAN IMPLIED. print_run=UNKNOWN is forbidden at base-set scope by migrations/003_parallels.sql, so the only alternatives were unnumbered or deleting four provably-real tiers; the discriminative test is available and strong, so unnumbered was written. Coverage, counted by me from each render: Crystal 81 rows (1-81) on Checklist.cfm page 1, 69 (1-69) under ?sort=0, 50 (101-150) on page 2, and 90 rows on ChecklistByLastName.cfm. Horizon 71, 76, 50 and 86. Metallized 80, 50, 86, plus 75 rows on a third endpoint ViewSet.cfm. Rain 87, 57 (spanning 101-177) and 89. Zero serial tokens on any row of any of those renders, across two or three endpoints and up to three sort orders per tier. The positive half of the test is inside the same renderer and the same product: Checklist.cfm renders SN10, SN5, SN75 and SN1 on every rendered row of the four annotated base-set siblings. WHAT REMAINS POSSIBLE: a serial token on one of the roughly 40 to 70 rows per tier that no render reached. That is this release's weakest surviving claim after G-006. https://www.tcdb.com/ChecklistByLastName.cfm/sid/148260/2016-17-Panini-Prestige-Crystal
• G-010 FOUR NAME DIVERGENCES RECORDED RATHER THAN TIDIED. (1) tcdb's base-set parallel is headed bare "Platinum" while cardboardconnection prints "Platinum Bonus Shots 1/1"; tcdb's page heading was used. Likewise tcdb's "Bonus Shots Gold/Green/Red" against cardboardconnection's "Gold/Green/Red Bonus Shots" - word order differs and no row was renamed. (2) tcdb heads sid 148281 "2016-17 Panini Prestige - Prestigious Premieres" while heading its parallel sid 148282 "... Prestigious Premieres Signatures Rainbow", so the word Signatures appears in the tier name and not in the base name; both are reproduced verbatim. (3) cardboardconnection's parallel list names only Metallized and the four Bonus Shots tiers, and does not name Crystal, Horizon or Rain at all, though tcdb gives each its own sid and a 200-card roster - an editorial omission, not a contradiction. (4) "Prestige" names two different objects in this catalogue: CAV-021 records that Panini Chronicles carves its base run into sub-brand blocks one of which is called Prestige, and that block is not this release. https://www.cardboardconnection.com/2016-17-panini-prestige-basketball
• G-011 NO CLAIM IS MADE THAT THIS RELEASE HAS NO VARIATION SET. Errors.cfm on the parent returns the page headed "2016-17 Panini Prestige" with the sentence "No errors or variations have been added for this set yet." Per CAV-020 that silence discriminates nothing - the identical sentence is returned for sets that provably carry UER rows - so it is cited here as having produced no rows and nothing more, and the 2016-17 scope memo section 7.3 records the same empty result on this same parent. What the clean base run actually rests on is the independent contiguous enumeration of 1-200 with cross-endpoint overlaps agreeing row for row, plus Gallery.cfm ending at exactly 200 tiles over 20 pages with no page 21. A b-suffixed variation layer would not surface in a base-run enumeration and Gallery.cfm strips the b suffix, so absence of one is UNSUPPORTED rather than disproved. https://www.tcdb.com/Errors.cfm/sid/146982/2016-17-Panini-Prestige
• G-012 family_of('Metallized') RETURNS 'Meta', AND THAT IS THE CAV-019 ORDERED-SUBSTRING FAILURE FIRING A THIRD TIME. parallel_family was derived by AST-lifting family_of out of pipeline/fix_schema.py per CAV-034 and CAV-039, and nothing was hand-written. FAMILIES carries ('Metallic','Metallic') at index 152 and ('Meta','Meta') at index 153, so the longer pattern is correctly ordered first - but "Metallized" contains neither "Metallic" nor any longer pattern, and falls through to the four-letter "Meta". 'Meta' is a real family minted by the Spectra line (50 rows across five spectra-bkb seasons) and describes a different finish entirely. This is the Ink-inside-Pink and Cosmo-inside-Cosmos shape on a new pattern pair. CORPUS CHECK RUN, per CAV-019's procedural lesson that 'no row in my slice is affected' is not 'no row is affected': ZERO existing rows in the corpus carry a parallel_name containing "Metal" that derives to 'Meta'. So the defect is not latent - it goes live for the first time on this one row. The derived value 'Meta' was written anyway, because CAV-019 states the rule is its own provenance and the task brief forbids adding a FAMILIES entry. The fix belongs in fix_schema.py, scoped to the exact name, and was NOT made here. SEPARATE, AND NOT A DEFECT: family_of('Distinctive Ink Rainbow') returns 'Ink', which is arguably right for an autograph set and arguably wrong for a Rainbow tier; the derivation was left to stand. The line's Horizon / Rain / Rainbow / Metallized ladder has no family in the list and lands in 'other' on 25 of 40 rows. NO FAMILY WAS MINTED.
• G-013 WHICH LINEAGES WERE AND WERE NOT REACHED, AND WHAT THAT MEANS FOR source_confidence. USED: tcdb (Checklist.cfm, ViewSet.cfm, ChecklistByLastName.cfm, Gallery.cfm, Inserts.cfm, Errors.cfm) and cardboardconnection's product page. Those are TWO independent lineages, and the four Bonus Shots print runs are the only values in this release carried by both - they are the only four rows written source_confidence=multi. Everything else is single. NOT REACHED, and no absence is treated as evidence: checklistcenter.com was not surfaced for this product and per CAV-037 no URL was constructed and no 404 interpreted; beckett.com was not budgeted for per CAV-033; sportscardchecklist, checklistinsider and waxstat were not swept; blog.paniniamerica.net was not opened for this product. MARKETPLACE CORROBORATION WAS NOT ATTEMPTED - this release writes no plain integer card number onto any autograph set (insert card numbers are a separate pass under CHASE_CHECKLIST_SPEC), so the required check in EXTRACTION_SPEC 3.4 was not engaged, but it means no independent confirmation of any serial run exists outside the two lineages above. CONSEQUENCE: the CAV-037 per-product independence test between cardboardconnection and checklistcenter could not be run, because only one of the two was reachable.
• G-014 THE SAME NOTES CELLS CAME BACK EMPTY ON ONE FETCH AND AS PERCENTAGES ON ANOTHER, AND ONLY THE SECOND IS TRUE. The first fetch of this parent's Inserts.cfm reported the Notes cell as EMPTY for Crystal, Horizon, Metallized and Rain. A second, differently-slugged and differently-parameterised fetch asked for the same four cells character for character and returned "6.8% of images added", "43.3% of images added", "23.5% of images added" and "12.0% of images added". specs/2016-17_SCOPE_MEMO.md section 7.3 independently reports percentages in those cells, so the second reading is the right one and the first fetch invented an absence. WHY IT MATTERS: on the first reading the Parallel Sets section splits 4 annotated against 4 blank and passes the both-directions blank test, which would license writing unnumbered off blankness. On the true reading there are NO blank cells in that section at all, only serial statements and image-completion percentages, so the split test is not merely failed - it is UNRUNNABLE, exactly as CAV-036's eleven-parent tally predicts. Every print run in this release therefore came from a per-sid row-level token sweep, and the Notes column was used only as corroboration for the four cells that carry a real serial statement. FORMAT, per CAV-036: this parent renders PROSE in Parallel Sets ("Serial numbered to 10", "to 5", "to 75", "to 1") and prose plus percentages in Insert Sets. That is the OPPOSITE format from 2017-18 Prestige, which CAV-036 records as bare tokens (SN199, SN10, SN1) in the same section of the same line one season up. An agent who calibrated on 2017-18 Prestige and parsed for bare tokens here would have extracted nothing from four provably annotated cells and written unnumbered on a /1, a /5, a /10 and a /75. https://www.tcdb.com/Inserts.cfm/sid/146982/2016-17-Panini-Prestige
• G-015 THREE DISAGREEMENTS WITH specs/2016-17_SCOPE_MEMO.md SECTION 7.5, STATED PLAINLY. All three are in that memo's favour on the structural claim and against it on detail, and its own weakest-claim note predicted exactly this. (1) It records Old School Signatures as carrying "1 row carrying AU and no serial token". Two endpoints here both report 0 such rows and both report per-token counts summing exactly to 20 (Checklist.cfm 10+8+2, ViewSet.cfm 8+9+3). The DISTINCT token set {SN199, SN99, SN49} matches the memo exactly; only the tokenless row does not reproduce. (2) Its Distinctive Ink probe returned per-token counts summing to 23 against 25 rows; my Checklist.cfm render summed to 24 and my ViewSet.cfm render summed to 25. Three renders, three different arithmetics, one identical distinct token set - which is why only the distinct set is relied on anywhere in this release. (3) PER-TOKEN COUNTS DISAGREE BETWEEN ENDPOINTS ON A SET WHERE BOTH SUM CORRECTLY: Stars of the NBA Signatures returns SN199x29 SN49x9 SN99x3 SN10x9 on Checklist.cfm and SN199x20 SN49x20 SN99x5 SN10x5 on ViewSet.cfm. Both total 50 and both give the same four distinct tokens. No per-token count is written anywhere in this release. It also records Prestige Platinum as "79 rendered of 200"; 150 rows were reached here, all SN1.
• G-016 NO PACK ODDS ARE RECORDED FOR ANY OF THE 27 SETS, because no reachable source states any. cardboardconnection carries box-level copy only ("Every hobby box promises a trio of hits with two being autographs") and no odds line; beckett is unreachable per CAV-033. Per CAV-015 a blank stated_odds on a Panini row means unpublished, not unresearched. THREE scarcity_tier VALUES WERE WRITTEN, each off a quote about that specific set: Playmakers and Prestigious Pros from "Featuring holo board, hobby-only Playmakers and Prestigious Pros inserts are each a case hit", and Prestigious Picks from "Prestigious Picks is a hobby-only case hit showcasing 2016 rookies." Every other set's scarcity_tier is left EMPTY rather than UNKNOWN, matching the corpus convention that empty means the scarcity pass has not covered the row. All 40 parallels.csv rows carry an empty scarcity_tier for the same reason. https://www.cardboardconnection.com/2016-17-panini-prestige-basketball

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