2017-18 Chronicles

• G-002 BASE SET SIZE - TWO SOURCES MODEL THIS PRODUCT DIFFERENTLY AND THIS BUILD PICKED ONE. tcdb treats the product as one continuous 400-card run: "Total Cards: 400" at https://www.tcdb.com/ViewSet.cfm/sid/162386/2017-18-Panini-Chronicles, and an independent contiguous enumeration of 1..400 over four Checklist.cfm pages and thirteen Gallery.cfm pages found no hole and no duplicate number. cardboardconnection instead states the base set is "150 cards." and files the remaining 250 numbers as ten separate 25-card rookie sets (Gold Standard, Gala, Titanium, Marquee, Pinnacle, Panini, Majestic, Momentum, Brilliance, Limited) at https://www.cardboardconnection.com/2017-18-panini-chronicles-basketball-nba-cards; its own arithmetic reconciles to the same 400 printed numbers (150 + 10 x 25 = 400). The two sources agree on what was printed and disagree on what to call the base set. base_set_size=400 was written, following the Chronicles anthology convention in CHRONICLES_BLOCK_ADDENDUM.md, CAV-020 and the sibling chronicles-bkb-2018-19 (base_set_size=699 over 21 sub-brand blocks). A reader who prefers cardboardconnection's model would read base_set_size=150 with ten insert sets. Recorded rather than smoothed over.
• G-003 BLOCK ORDER WITHIN 251-325 IS THE WEAKEST CLAIM IN THIS RELEASE. Eight of the eleven blocks are pinned by a tcdb page that renders exactly one block's card numbers (Gold Standard Au 151-175; Gala via the parent gallery rendering SN8 on exactly 176-200; Titanium Draft Pick 201-225; Marquee Tier 2 226-250; Momentum Drive Silver 326-350; Brilliance Starbursts 351-375; Limited Black 376-400; and Chronicles 1-150 as the remainder below 151). The three rookie sub-brands Pinnacle, Panini and Majestic are pinned COLLECTIVELY to 251-325 by tcdb's Silver set (https://www.tcdb.com/ViewSet.cfm/sid/168232/2017-18-Panini-Chronicles-Silver, "Total Cards: 75", rows 251-325), which cardboardconnection lists as a parallel of exactly those three blocks - but their ORDER inside 251-325 does not follow from that. It rests on: (a) one fetched free-text eBay seller listing, https://www.ebay.com/itm/166413124971, title "2017-18 Panini Chronicles Pinnacle #257 Donovan Mitchell /99 Rookie Card RC", carrying no parenthesised tcdb flag string, which places Pinnacle at 251-275; and (b) for Panini 276-300 and Majestic 301-325, only eBay CATALOGUE pages whose titles are in tcdb's own "<Release> - <Set> <Player> #N (RC)" form - https://www.ebay.com/p/23056586103 reads "2017-18 Panini Chronicles - Panini Lonzo Ball #298 (RC)" and https://www.ebay.de/itm/154295222884 reads "2017-18 Panini Chronicles - Majestic Josh Jackson #309 Red /249 (RC)". Per CHASE_CHECKLIST_SPEC those are tcdb-seeded and are NOT independent corroboration; two of them were reached from free-text search snippets and proved to be retitled on the page itself, exactly the documented contamination shape. So Panini and Majestic rest on elimination plus second-hand tcdb. If those two blocks are swapped, base_cards.csv card_tier is wrong on 50 rows; the card numbers, players, teams and print runs are unaffected.
• G-004 BLOCK PARALLEL NAMES ARE CONSTRUCTED AS BLOCK+COLOUR AND NO SOURCE PRINTS THAT EXACT STRING. tcdb files the colour parallels as ONE sid per colour spanning several blocks at different print runs - its "Blue" sid (https://www.tcdb.com/ViewSet.cfm/sid/167088/2017-18-Panini-Chronicles-Blue, "Total Cards: 275", Notes "Serial numbered to 199") actually renders SN199 on blocks Chronicles and Pinnacle/Panini/Majestic/Momentum and SN149 on the Limited block. cardboardconnection prints the same tiers as bare colours under each sub-brand heading ("Blue #/199" under Momentum Rookies, "Blue #/149" under Limited Rookies). Because migrations/003_parallels.sql gives every scope_type=block row scope_set_id NULL, all block rows in a release share one uniqueness key and two may never share a name, so bare "Blue" cannot be written five times. The names here (Chronicles Blue, Pinnacle Blue, ... , Limited Blue) prefix the colour with the block name, which is itself sourced; no colour word was invented and nothing was renamed away from a source. EXTRACTION_SPEC 2.3 prefers a source's longer name over an invented disambiguator; here no longer name exists, so this is the minimal construction. A resolver matching tcdb's bare "Blue" will not hit these rows without the block.
• G-005 SERIAL NUMBERED WITH AN UNRESOLVABLE RUN - THE SCHEMA CANNOT HOLD IT, ON THREE TIERS. (1) Chronicles Autographs Blue, https://www.tcdb.com/ViewSet.cfm/sid/167098/2017-18-Panini-Chronicles-Chronicles-Autographs-Blue, "Total Cards: 46": AU on all 46 rows, SN99 on 42 and SN75 on 4 - a per-card variable run. (2) Titanium Draft Pick and (3) Titanium Draft Year - see G-006. All three are PROVABLY serial numbered and none has one print run, and migrations/003_parallels.sql's run_and_serial_agree constraint forces print_run=UNKNOWN to pair with serial_numbered=UNKNOWN, which is indistinguishable from "we found nothing". The distinct token sets were enumerated, not sampled, exactly so this is recorded rather than collapsed to the modal token. This is the EXTRACTION_SPEC 2.3 expressiveness gap recurring on a fourth line.
• G-006 THE TITANIUM BLOCK NUMBERS EVERY CARD TO A PER-CARD VALUE - CAV-023, REPRODUCED ONE SEASON EARLIER THAN RECORDED. The Titanium block (201-225) renders a different serial token on nearly every row in the parent checklist: https://www.tcdb.com/Gallery.cfm/sid/162386/2017-18-Panini-Chronicles?PageIndex=23&sort=0 renders "221 - Frank Ntilikina RC, SN11", "225 - Malik Monk RC, SN1", and https://www.tcdb.com/Checklist.cfm/sid/162386/2017-18-Panini-Chronicles?PageIndex=3&sort=0 renders SN5, SN13, SN24, SN23, SN20, SN7, SN16, SN9, SN32, SN10, SN2 across 201-225 - consistent with each card being numbered to that player's JERSEY number. Its Draft Pick parallel (https://www.tcdb.com/ViewSet.cfm/sid/167121/2017-18-Panini-Chronicles-Titanium-Draft-Pick) renders a different per-card set - SN5 De'Aaron Fox, SN3 Jayson Tatum, SN1 Markelle Fultz, SN8 Frank Ntilikina, SN47 Ike Anigbogu - consistent with draft position; and Draft Year (https://www.tcdb.com/ViewSet.cfm/sid/167122/2017-18-Panini-Chronicles-Titanium-Draft-Year) renders only SN17, SN16 and SN15. The jersey-number / draft-position / draft-year readings are PATTERN RECOGNITION on the tokens, not a source statement, and are recorded here rather than written into any field. print_run and serial_numbered are UNKNOWN on the Titanium tier row and on both parallels. CAV-023 already documents this shape for Chronicles and names four agents who hit it; this is the fifth, and it moves the earliest known occurrence back to 2017-18.
• G-007 CARD COUNT DISAGREEMENTS ON TWO CHASE SETS, WITH A LIKELY MECHANISM. Chronicles Autographs: tcdb renders "Total Cards: 46" with 46 rows counted (https://www.tcdb.com/ViewSet.cfm/sid/167097/2017-18-Panini-Chronicles-Chronicles-Autographs) while cardboardconnection states "47 cards" (https://www.cardboardconnection.com/2017-18-panini-chronicles-basketball-nba-cards). Chronicles Signature Swatches: tcdb renders "Total Cards: 35" with 35 rows while cardboardconnection states "37 cards". tcdb's rendered row counts were written. The likely mechanism is EXTRACTION_SPEC 5.4 unit 4, a PARALLEL'S ROSTER counted as the set: the Chronicles Autographs Gold, Holo Gold and Pink parallels each render exactly 47 rows, and the Chronicles Signature Swatches Gold parallel renders exactly 37 - the two cardboardconnection figures. That is a hypothesis fitted to the numbers, not a sourced explanation, so both sets keep the tcdb count and this gap keeps the alternative. Note also that the parallel pages here are LARGER than their base pages (47 vs 46, 37 vs 35), the CHASE_CHECKLIST_SPEC 1c-ter shape, which the insert-checklist pass should resolve by roster.
• G-008 SOURCE COVERAGE IS THINNER THAN THE CITATION COUNT SUGGESTS. beckett.com was not consulted: CAV-033 records it unreachable and the 2017-18 scope memo re-confirmed the 302 to maintenance on a 2017-18 URL. checklistcenter.com has no 2017-18 Chronicles page in any search result returned here; per CAV-037 no URL was constructed and no absence is claimed - the correct statement is "not surfaced by search", not "does not exist". sportscardradio.com/2017-18-panini-chronicles-basketball-checklist/ returned ROBOTS_DISALLOWED and breakninja.com's page is organised by team with no sub-brand headings, so neither could supply block ranges. gogts.net carries the product page but publishes its checklist only as PDF/Excel downloads and no card numbers in the page text. The structural spine of this release is therefore tcdb, with cardboardconnection as the one independent editorial source for block names and print runs - a two-source release, not a three-source one.
• G-009 NO PARALLEL COVERS THE WHOLE BASE SET, SO parallels.csv HAS ZERO scope_type=base-set ROWS. Every colour tier in this product is block-scoped: the widest is Blue at 275 of 400 numbers (https://www.tcdb.com/ViewSet.cfm/sid/167088/2017-18-Panini-Chronicles-Blue, "Total Cards: 275"), then Red and Gold at 250, Holo Gold at 225, Purple and Pink at 150. No colour reaches blocks Gold Standard (151-175), Gala (176-200), Titanium (201-225), Marquee (226-250) or Brilliance (351-375), each of which has its own differently-named tiers instead. This is recorded because CAV-026 tells consumers to filter scope_type=base-set to get "the parallel applies to the whole base set", and on this release that filter correctly returns nothing. It is also why the v_card_parallels view will produce no rows for this release.
• G-010 NO VARIATION-LAYER CLAIM IS MADE IN EITHER DIRECTION. Errors.cfm was not exercised on sid 162386 in this build, and per CAV-020 its silence would have been worthless anyway. What IS established is independent of it: the base run enumerates contiguously 1..400 with no hole, no duplicate number and no letter-suffixed row seen on any endpoint, across https://www.tcdb.com/Checklist.cfm/sid/162386/2017-18-Panini-Chronicles?PageIndex=1&sort=0 through PageIndex=4 and thirteen Gallery.cfm pages, agreeing at three page overlaps. Since Gallery.cfm strips a trailing b suffix (2017-18 scope memo 7.7) a variation layer would not surface in a base-run enumeration, so this release may under-report one. An Errors.cfm pass on the parent sid is the cheap next step.
• G-011 THE GALA BLOCK (176-200) HAS NO PARALLEL TIER. tcdb's Inserts.cfm listing for the parent (https://www.tcdb.com/Inserts.cfm/sid/162386/2017-18-Panini-Chronicles) returned all 36 rows in one fetch - 29 under "Insert Sets (29)" and 7 under "Parallel Sets (7)" - and none of them renders Gala card numbers; cardboardconnection states only "Gala Rookies: 25 cards. Serial numbered #/8" with no parallel list, where it gives one for every other sub-brand block. The base Gala cards are themselves #/8, so a parallel ladder above them would have little room. This is an absence recorded from two sources that both list parallels for the other ten blocks, which is stronger than a bare silence, but it is still an absence and no negative claim is written into any field.
• G-012 BLOCKS WRITTEN unnumbered/no REST ON A WITHIN-SID DISCRIMINATIVE TEST, NOT ON A SOURCE STATEMENT. Six blocks carry print_run=unnumbered: Chronicles (1-150), Pinnacle, Panini, Majestic (251-325) and Momentum (326-350). No source says in words that these cards carry no serial number. What licenses the value is the EXTRACTION_SPEC 6 positive discriminative test in its strongest, within-sid form: the SAME parent renderer prints serial tokens on the other rows of the same page and none on these - https://www.tcdb.com/Gallery.cfm/sid/162386/2017-18-Panini-Chronicles?PageIndex=23&sort=0 renders "225 - Malik Monk RC, SN1" and "226 - T.J. Leaf RC, SN99" while https://www.tcdb.com/Gallery.cfm/sid/162386/2017-18-Panini-Chronicles?PageIndex=29&sort=0 renders "281 - Mike James RC - Phoenix Suns" with no serial token at all, and the contrast reproduces on a second endpoint, Checklist.cfm?PageIndex=3&sort=0. Confirmed in both directions per the 2026-09-20 direction-neutral correction. Had only the absence been available this would have been UNKNOWN.
• G-013 insert_cards.csv WAS NOT WRITTEN - that is the separate CHASE_CHECKLIST_SPEC pass and is out of scope for this build. Recorded here so the shortfall is visible: the three chase sets carry letter-coded card numbers and the prefixes were observed on tcdb - Chronicles Autographs CA- (CA-ADV Anthony Davis, CA-APT Alec Peters, CA-BAM Bam Adebayo, CA-DAF De'Aaron Fox, CA-DBC Dwayne Bacon ... CA-ZCL Zach Collins), Chronicles Signature Swatches CSS- (CSS-BAM, CSS-CSW, CSS-DAF, CSS-DJW, CSS-DML), Chronicles Swatches CS- (CS-ADV, CS-AWG, CS-BAM, CS-BGF, CS-BLP). Note the CS- prefix collides in shape with other Panini 2017-18 products; key on (set_name, card_number).
• G-014 THE Inserts.cfm NOTES COLUMN ON THIS PARENT IS SECTION-SPLIT AND ITS PRINT RUNS ARE INCOMPLETE - CAV-036 CALIBRATION RESULT. https://www.tcdb.com/Inserts.cfm/sid/162386/2017-18-Panini-Chronicles renders PROSE, not bare tokens: its "Parallel Sets (7)" section carries "Serial numbered to 199", "Serial numbered to 10", "Serial numbered to 1", "Serial numbered to 99", "Serial numbered to 149", "Serial numbered to 299" and a second "Serial numbered to 99" on all seven rows, while its "Insert Sets (29)" section on the same page carries image-completion percentages ("4.0% of images added") and empty cells, which are not serial statements at all. So the format here is the Hoops prose shape, not the Select bare-token shape, and the blank-annotation test cannot be run on the insert section. MORE IMPORTANTLY the prose runs in the parallel section are INCOMPLETE per sid: the Blue row's "Serial numbered to 199" is true of 250 of its 275 cards and FALSE of the other 25, which render SN149, and the Red row's "Serial numbered to 299" is true of 150 of its 250 cards and false of the other 100, which render SN249. An agent trusting that column would have written two wrong print runs. Every parallel row in this release was instead enumerated for its distinct tokens against Gallery.cfm or ViewSet.cfm.

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