2019-20 Contenders Optic

• G-001 base_print_run recorded as unnumbered/assumed. No source STATES that the 2019-20 Contenders Optic base cards 1-100 are unserialised. The evidence is indirect but strong: tcdb prints a per-row serial flag on every numbered base parallel ("SN99" on Blue, "SN49" on Orange, "SN10" on Gold, "SN1" on Gold Vinyl and all four Printing Plates) and prints NO FLAG TOKEN ON ANY ROW of the base checklist or of Silver/Red; and both aggregators open the ladder with unfigured "Silver, Red" before "Blue #/99". Per EXTRACTION_SPEC 2.1 that is a rule I applied, not a statement I found, so the basis is `assumed`, not `sourced`.
• G-002 BASE SET SIZED AT 100 AGAINST A PARENT "Total Cards" OF 140, AND THIS DIVERGES FROM AN EXISTING RELEASE IN THIS CATALOGUE. EXTRACTION_SPEC 5.3 requires base_set_size to exclude an autograph block filed inside the parent checklist past the base run, and cards 101-140 are exactly that: tcdb tiles 101-140 carry "(AU, RC)" while tiles 1-100 carry no token; checklistcenter heads a 100-card "Base Season Ticket Set" and a separate 40-card "Rookie Ticket Auto Set" numbered 101-140; the Silver base parallel stops at 100 while every other base parallel runs to 140; and the two blocks carry different print runs under the same tcdb parallel set (Orange is SN49 on 1-100 and SN25 on 101-140). BUT data/contenders-bkb-2019-20 made the OPPOSITE call on the same structural shape: its release note says "All 165 are recorded in base_cards.csv because they share the single Season Ticket numbering; the autographed portions are also recorded as autograph sets in sets.csv", giving base_set_size=165 with cards 101-165 being on-card Rookie Ticket autographs. Same manufacturer, same season, same design family, same evidence shape, opposite base_set_size convention - and nothing in either release row tells a reader which convention it is under. This release follows the rewritten EXTRACTION_SPEC 5.3; a repair pass over contenders-bkb-2019-20 (and its 2020-21..2023-24 siblings) is needed so the two lines agree. Flagged, not worked around.
• G-003 The two largest autograph sets in this release have NO NAME ON tcdb because they have no sid - they sit inside the parent checklist (101-140) and in Errors.cfm (101b-140b). The names written are cardboardconnection's section headings, "### Rookie Ticket Set Checklist" and "### Rookie Ticket Variation Set Checklist". checklistcenter names the same two "Rookie Ticket Auto Set" and "Rookie Variation Season Ticket Auto Set". "Rookie Ticket" was chosen because it is the form tcdb uses for the sibling set it DOES name ("Veteran Ticket"), but no source prints exactly the string written. Unresolved: the manufacturer's own name for these two sets.
• G-004 Veteran Ticket card count disagreement: tcdb states Total Cards "10" and renders 10 rows, VT-ADV through VT-SCY (https://www.tcdb.com/ViewSet.cfm/sid/306030/Veteran-Ticket), and checklistcenter states "10 Cards"; cardboardconnection states 25. sets.csv uses 10. The source of cardboardconnection's 25 was not identified - it is not a parallel-line count, since the Veteran Ticket ladder is Red/Blue/Orange/Gold/Gold Vinyl/Printing Plates.
• G-005 Winning Tickets card count disagreement: tcdb states Total Cards "25" and renders 25 rows, card 1 Kawhi Leonard through card 25 Kevin Garnett (https://www.tcdb.com/ViewSet.cfm/sid/247357/Winning-Tickets), and checklistcenter states "25 Cards"; cardboardconnection states 35. sets.csv uses 25.
• G-005b Winning Tickets parallel-ladder disagreement: cardboardconnection prints "PARALLEL CARDS: Gold #/10, Gold Vinyl 1/1." and no Cracked Ice tiers, while tcdb catalogues "Winning Tickets Blue Cracked Ice" (sid 247358) and "Winning Tickets Red Cracked Ice" (sid 247359) and checklistcenter lists Cracked Ice "Blue; Red". The two Cracked Ice rows are written (two sources including tcdb's own sids, and tcdb renders NO FLAG TOKEN ON ANY ROW of sid 247358, consistent with the other eight sets in this family). cardboardconnection's omission is unexplained.
• G-006 ROOKIE TICKET RED AND BLUE ARE CEILINGS, NOT PRINT RUNS, AND THE TWO AGGREGATORS DISAGREE ON RED. cardboardconnection: "Red #/149 or less, Blue #/99 or less"; checklistcenter: "Blue #/99 or Less; Red #/49 or Less". tcdb adjudicates in cardboardconnection's favour - a fetch of Gallery.cfm/sid/244802/...?PageIndex=11 reported "SN149" on card 101 and "SN75" on 103/104, both of which are impossible under a 49 ceiling. EXTRACTION_SPEC 2.3 forbids writing the ceiling, so both rows are print_run=UNKNOWN / serial_numbered=UNKNOWN with value_contested=yes on Red. NOTE this DIVERGES from earlier catalogue practice, where agents wrote the ceiling with a gap (contenders Ticket Stub 77, elite Aspirations 99); the rewritten spec reverses that and a repair pass over those rows is expected.
• G-007 THE SCHEMA CANNOT EXPRESS "SERIAL NUMBERED, RUN VARIES PER CARD" AND THIS RELEASE HAS TWO SUCH TIERS. Rookie Ticket Blue is PROVABLY serial numbered with a per-card run: tcdb renders "(AU, SN99)" on card 101, "(AU, SN35)" on 102 and "(AU, SN49)" on 103 and 104 (Gallery.cfm/sid/244803/...?PageIndex=11). Rookie Ticket Red is partly numbered and partly not - "SN149"/"SN75" reported on 101/103/104 at PageIndex=11 but plain "(AU)" with no serial on all of 106-110 at PageIndex=12. Both collapse to UNKNOWN/UNKNOWN in parallels.csv, which is indistinguishable from "we found nothing". The held facts are recorded here because the CSV cannot carry them. Also note the two Gallery fetches of the Red set disagreed with each other about whether cards 101/103/104 carry a token at all, so the Red per-card values above are reported once and not confirmed by a second read.
• G-008 Card 81 Darius Garland is a 2019 NBA draftee - a rookie in the 2019-20 season - yet he sits inside the 1-100 veteran block and tcdb renders NO (RC) token on his tile (Gallery.cfm/sid/244800/...?PageIndex=9), while all forty 2019 draftees at 101-140 carry "(AU, RC)". He does not appear a second time in 101-140 (the block runs 101 Nassir Little through 140 Chuma Okeke and contains no Garland). rookie_flag=no is written on tcdb's display alone. Unresolved: whether the card is genuinely not RC-designated or whether tcdb has simply not tokenised it.
• G-009 rookie_flag rests on a single source. tcdb renders no (RC) token on any of cards 1-100 and renders "(AU, RC)" on all of 101-140, so absence across the base run is evidenced rather than unresolved - but no other source consulted publishes RC designations for this base set, and neither cardboardconnection nor checklistcenter marks rookies on the base checklist. All 100 rows written rookie_flag=no on tcdb's display alone.
• G-011 Rookies T-Mall classification is contested and could be a set OR a parallel. tcdb gives it a sid, a distinct name ("2019-20 Panini Contenders Optic - Rookies T-Mall", Total Cards "25"), two parallels of its own and the page note "China-Exclusive retail product. Cards are NOT autographed." - all set-shaped. But its rows reuse the Rookie Ticket card numbers (rendered 101 Nassir Little through 138 Bruno Fernando) and cardboardconnection describes it as "25 base Rookie Ticket versions", which is parallel-shaped. This is EXTRACTION_SPEC 5.4 trap 4 (a parallel's roster sitting in a card_count field) pointing the opposite way. Written as a sets.csv row following tcdb; if it is really an unsigned parallel of Rookie Ticket, this release asserts one set too many and parallels.csv is missing a base-scope row. Not resolved either way. checklistcenter does not list it at all, so the set row is effectively two-source (tcdb + cardboardconnection) with those two describing it differently.
• G-012 Base Gold Wave and Red Wave extent is contested. tcdb states Total Cards "140" for both (sid 250787, sid 250786), which would make them full base parallels covering cards 1-140. cardboardconnection instead frames the Wave finishes as belonging to a 25-card T-Mall rookie product: "T-Mall exclusives include 25 base Rookie Ticket versions with Gold Wave and Red Wave parallels.", and separately "T-Mall boxes have the exclusive Red Wave autographs." Neither aggregator lists Gold Wave or Red Wave in the ordinary base parallel ladder. Written at base-set scope on tcdb's 140-row header; if cardboardconnection is right, these two rows overstate the parallels' extent. Note tcdb ALSO carries separate 25-card "Rookies T-Mall Gold Wave" and "Rookies T-Mall Red Wave" sets, so tcdb is asserting both a 140-card and a 25-card Wave object under nearly the same names.
• G-013 Two autograph sets rendered fewer rows than their own header states, through a summarizing fetch: Legendary Contenders Autographs states Total Cards "30" and returned 23 rows (sid 263354); Sophomore Contenders Autographs states "15" and returned 14 rows (sid 263379). Per EXTRACTION_SPEC 3.2 a disagreement between rendered rows and the header is a question, not an answer - but here both aggregators independently state 30 and 15, so the header figures were written and the short renders are treated as fetch truncation. The missing rows were not identified.
• G-014 No pack odds were found for any set in this release. cardboardconnection publishes only box-level averages ("Each Hobby box averages one autograph.", and a separate "T-Mall Box Break Average" heading) and explicitly no per-set odds or stated print run line. stated_odds / odds_evidence_url / odds_evidence_quote are empty on every sets.csv row and scarcity_tier is unset throughout sets.csv and parallels.csv.
• G-015 beckett.com could not be consulted: https://www.beckett.com/news/2019-20-panini-contenders-optic-basketball-cards/ 302-redirects to https://beckett-maintenance-page.s3.amazonaws.com/index.html - the identical failure recorded on the certified-bkb-2019-20 pilot (its G-013), so the outage is persistent, not transient. The usual third structural source is therefore gone. HOWEVER, on this release cardboardconnection and checklistcenter demonstrably are NOT one opinion: they disagree on Veteran Ticket (25 v 10), Winning Tickets (35 v 25), the Winning Tickets Cracked Ice tiers (absent v present) and the Rookie Ticket Red ceiling (149 v 49). tcdb sided with checklistcenter on the first three and with cardboardconnection on the fourth, so neither aggregator is uniformly reliable here and neither is copying the other.
• G-016 FABRICATION CAUGHT: a summarizing fetch of the cardboardconnection page invented the base-set structure. Asked a broad question it returned "Base Set: 100 cards featuring current NBA stars, plus 40 rookie cards. The structure separates established veterans from the draft class." A narrow verbatim re-fetch of the same URL, demanding quotation marks, showed the page's actual base-set sentence is "Using a one-pack-per-box format, 2019-20 Panini Contenders Optic Basketball offers five non-hit cards in every box." - the 100/40 split is nowhere on the page in that form. The fabricated summary happened to agree with the truth established from other sources, which is exactly what makes it dangerous. base_set_size=100 rests on tcdb tokens, the Silver parallel header and checklistcenter, not on that string.
• G-017 ENDPOINT BEHAVIOUR, two findings. (1) CONFIRMED: Gallery.cfm validates its slug where ViewSet.cfm does not - all Gallery fetches here succeeded only under the canonical slug 2019-20-Panini-Contenders-Optic (and the canonical slug of each parallel sid), matching the pilot's G-014. (2) NEW: ViewSet.cfm SILENTLY IGNORES ?PageIndex. ViewSet.cfm/sid/244803/2019-20-Panini-Contenders-Optic-Blue?PageIndex=2 returned page 1 again (rows 1-73), with pagination controls visible but no way to reach rows 101-140. The only route into the tail of a 140-row parallel checklist was Gallery.cfm?PageIndex=N on that parallel's own sid. EXTRACTION_SPEC 3.2 does not mention this.
• G-018 Set-name variant not resolved: tcdb and checklistcenter render this insert singular ("Superstar"), cardboardconnection plural ("Superstars"). tcdb's singular is written because tcdb also names the four parallel sets Superstar Gold, Superstar Blue Cracked Ice, Superstar Red Cracked Ice and Superstar Gold Vinyl. All three sources agree the count is 15, so only the name is in question.
• G-019 The Silver base parallel covers only cards 1-100 while every other base parallel set covers 140: tcdb states Total Cards "100" for Silver (sid 244801) against "140" for Red, Blue, Orange, Gold, Gold Vinyl, Gold Wave, Red Wave and all four Printing Plates. Both aggregators' Rookie Ticket ladders independently omit Silver, so this reads as a real product fact (the autograph block has no Silver tier) rather than an incomplete tcdb entry - but no source STATES that the Rookie Ticket cards have no Silver parallel, so it is inferred from two silences plus one header.
• G-020 Release date 2020-09-23 is stated by two sources - tcdb parent "Release Date: September 23, 2020" and cardboardconnection "**Release Date**: September 23, 2020" - but both could be one figure. It is eight months after the 2019-20 Contenders release (2020-01-03) and seven after Donruss Optic (2020-02-12), i.e. after the 2019-20 season itself ended, which is plausible for a COVID-delayed 2020 release but was not independently corroborated. Beckett, the usual check on a release date, is unreachable (G-015).

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