{ "dataset_version": "2026.10.1", "observations_through": "2026-09-30", "computed_on": "2026-10-03", "count": 8, "note": "Each record takes a claim the collector community makes, measures it against this corpus and states a verdict plus the bias on BOTH sides. None of these checks produced a publishable figure and that is the result rather than a failure.", "checks": [ { "check_id": "RP019-A-CHECK-001", "reviewed_at": "2026-09-10", "claim": "A first quota bag typically requires a ratio between 1.5:1 and 2.5:1.", "claim_source": "Summaries of what public collector discussions say", "tested_against": "157 MyQuotaBag experiences, ratios declared by the collectors themselves", "result": { "first_quota_bag_rows": 19, "median": 1.71, "p25": 1.05, "p75": 2, "inside_the_claimed_band": 8, "share_inside": "42%", "at_offer_rows": 31, "at_offer_median": 1.1, "at_offer_share_inside": "29%" }, "verdict": "PARTIALLY SUPPORTED", "reading": "The median of first-bag reports (1.71) sits inside the quoted band, but fewer than half the reports do. Ratios declared at the moment of the offer are markedly lower, with a median of 1.1. Two honest readings coexist: the community quotes the band it expects and the reports MyQuotaBag holds are lower and far more scattered.", "caveat": "Both sides are biased. The community band comes with no counting method and no link. The MyQuotaBag corpus over-represents threads whose subject is the ratio itself. Neither figure is a benchmark." }, { "check_id": "RP019-A-CHECK-002", "reviewed_at": "2026-09-10", "claim": "Expected ratios have risen compared with previous years.", "claim_source": "Summaries of what public collector discussions say", "tested_against": "68 MyQuotaBag experiences carrying both a numeric ratio and a usable year", "result": { "2024": { "n": 7, "median": 1.2 }, "2025": { "n": 26, "median": 1.33 }, "2026": { "n": 34, "median": 1.45 }, "rows_without_year": 31 }, "verdict": "NOT CONTRADICTED, NOT DEMONSTRATED", "reading": "The median rises from 1.2 to 1.45 across three years, in the direction the community describes. The samples are small, the ranges overlap heavily and a third of the rows carry no usable year. This is a hint, not a trend.", "caveat": "Never publish this as a measured rise." }, { "check_id": "RP019-A-CHECK-003", "reviewed_at": "2026-09-10", "claim": "Vancouver is particularly competitive; Toronto shows a wide spread.", "claim_source": "Summaries of what public collector discussions say", "tested_against": "MyQuotaBag experiences with a resolved city", "result": { "Toronto": { "n": 11, "median": 1.9, "min": 1, "max": 3 }, "Vancouver": { "n": 4, "median": 1.4, "min": 1, "max": 3 }, "Montreal": { "n": 2 }, "New York": { "n": 7, "median": 1, "min": 0.5, "max": 2.2 } }, "verdict": "NOT SUPPORTED BY THE CURRENT SAMPLE", "reading": "On the reports MyQuotaBag holds, Toronto shows the higher median and the wide spread the community describes, while Vancouver sits lower on four reports only. Four reports cannot rank a city.", "caveat": "This is a reason to collect Vancouver, not a reason to publish a ranking." }, { "check_id": "RP019-A-CHECK-004", "reviewed_at": "2026-09-10", "claim": "The band quoted for a first quota bag has risen steadily: 1:1 to 2:1, then 1.3:1 to 2:1, then 1.5:1 to 2.5:1 in the two most recent windows.", "claim_source": "Four readings of public collector discussions, one per six-month window", "tested_against": "MyQuotaBag experiences carrying both a numeric ratio and a date, split into the same four windows", "result": { "2024-09/2025-03": { "n": 3, "median": 1.3, "quoted_low": 1 }, "2025-03/2025-09": { "n": 20, "median": 1, "quoted_low": 1.3 }, "2025-09/2026-03": { "n": 7, "median": 2, "quoted_low": 1.5 }, "2026-03/2026-09": { "n": 31, "median": 1.5, "quoted_low": 1.5 }, "by_calendar_year": { "2024": 1.2, "2025": 1.33, "2026": 1.45 } }, "verdict": "NOT SUPPORTED AT THIS RESOLUTION", "reading": "Split by calendar year, the MyQuotaBag median rises from 1.20 to 1.45 and appears to follow the quoted rise. Split into the same six-month windows the synthesis uses, it does not: 1.3, then 1.0, then 2.0, then 1.5. The apparent trend survives one way of cutting the data and not another, on samples of 3 to 31 reports. A signal that depends on the cut is not a signal.", "caveat": "This check exists to stop a plausible rise from being published. The quoted bands carry no counting method and no link and four readings of the same discussions by the same method are not four independent measurements. Treat both the rise and its absence as unproven." }, { "check_id": "RP019-A-CHECK-005", "reviewed_at": "2026-09-10", "claim": "Low ratios (around 1:1 to 1.3:1) become less frequent over time and the concrete first-person data confirms this.", "claim_source": "A comparative summary of public collector discussions", "tested_against": "Share of MyQuotaBag numeric ratios at or below 1.3, by six-month window and by calendar year", "result": { "by_window": { "2024-09/2025-03": { "n": 3, "share_low": "66%" }, "2025-03/2025-09": { "n": 20, "share_low": "55%" }, "2025-09/2026-03": { "n": 7, "share_low": "42%" }, "2026-03/2026-09": { "n": 31, "share_low": "45%" } }, "by_year": { "2024": { "n": 7, "share_low": "57%" }, "2025": { "n": 26, "share_low": "50%" }, "2026": { "n": 34, "share_low": "47%" } } }, "verdict": "WEAKLY CONSISTENT, NOT CONCLUSIVE", "reading": "Unlike the median, this measure points the same way under both cuts: the share of low ratios falls from 66 to 45 percent across the windows and from 57 to 47 percent across the years. The direction is consistent. The size is small, the last window rises again and the yearly samples are 7, 26 and 34 reports. Nothing here separates a real drift from ordinary noise.", "caveat": "Circularity warning. The summary states that first-person reports confirm its trend, but it was built with the same reports MyQuotaBag holds, so that agreement adds no independent confirmation. Its only independent element is its reading of the discussions, which has no link and no counting method.", "what_would_settle_it": "Fifty or more dated first-person ratios per six-month window, collected without regard to the ratio being the thread subject. Until then this stays an open question, never a published trend." }, { "check_id": "RP019-A-CHECK-006", "reviewed_at": "2026-09-10", "claim": "Kelly leads Birkin in discussion volume, Constance is a distant third and Kelly pulls further ahead through 2026.", "claim_source": "Summaries of what public collector discussions say about each model", "tested_against": "186 captured threads, counted by presence of each model, plus the 90 structured experiences that name a model", "result": { "threads_mentioning": { "Birkin": { "threads": 111, "share": "59%", "mentions": 482 }, "Kelly": { "threads": 106, "share": "56%", "mentions": 484 }, "Mini Kelly": { "threads": 47, "share": "25%", "mentions": 150 }, "Constance": { "threads": 30, "share": "16%", "mentions": 93 } }, "structured_experiences": { "Birkin": 52, "Kelly": 51, "Mini Kelly": 10, "Constance": 3, "total_with_model": 90 }, "by_year_experiences": { "2024": { "Kelly": 6, "Birkin": 4 }, "2025": { "Kelly": 11, "Birkin": 9 }, "2026": { "Birkin": 15, "Kelly": 14 } } }, "verdict": "SPLIT: third place CONFIRMED, first place NOT CONFIRMED", "reading": "Constance being a distant third is confirmed by counting: it appears in 16 percent of threads against 56 and 59 percent for the two leaders and in 3 of 90 structured experiences. The claim that Kelly leads Birkin is not confirmed. Birkin appears in slightly more threads (111 against 106) and in slightly more experiences (52 against 51) and in 2026 Birkin is ahead in the corpus. The two are effectively tied at this sample size and the claimed 2026 lead for Kelly runs the other way here.", "caveat": "Counting a model name is not counting interest in it: a thread about a Birkin can mention Kelly to compare. Both counts measure co-occurrence, not intent. The 186 threads were also gathered by ratio keywords, not sampled across the forum.", "what_would_settle_it": "Counting by thread subject rather than by any mention, on threads sampled by publication date across the whole subreddit." }, { "check_id": "RP019-A-CHECK-007", "reviewed_at": "2026-09-10", "claim": "Median reported spend rose from about 12,500 USD in the 2024-09 window to about 21,500 and 22,000 USD in the two most recent windows.", "claim_source": "Public collector discussions: city threads and ordinary discussion threads, three six-month US windows", "tested_against": "The source threads behind each window", "result": { "2024-09/2025-03": { "rows": 7, "threads": 5, "third_party_lines": 0, "first_person": 7, "rows_with_amount": 2, "median_usd": 12500 }, "2025-09/2026-03": { "rows": 9, "threads": 4, "third_party_lines": 8, "first_person": 1, "rows_with_amount": 8, "median_usd": 21500 }, "2026-03/2026-09": { "rows": 11, "threads": 5, "third_party_lines": 8, "first_person": 3, "rows_with_amount": 9, "median_usd": 22000 }, "shared_threads_between_the_two_recent_windows": 3, "shared_threads_with_the_oldest_window": [], "confirmed_by_a_fourth_window": { "window": "2023-09/2024-03", "rows": 7, "threads": 2, "third_party_lines": 0, "rows_with_amount": 3 } }, "verdict": "ARTEFACT OF SOURCE, NOT A RISE", "reading": "The oldest window and the two recent ones do not come from the same kind of thread. The two recent windows are drawn from the same three city threads titled « offers and prespends », where a member lists offers with their amounts: eight of nine and eight of eleven lines are third-party compilations and almost every line carries a figure. The oldest window comes from five ordinary discussion threads, entirely first-person, where only two lines out of seven carry an amount at all. The jump from 12,500 to 21,500 measures the change of source, not a change in what collectors spend. Between the two windows that ARE comparable, sharing the same three threads, the median moves from 21,500 to 22,000: effectively flat. A fourth window, 2023-09 to 2024-03, confirms the split: also discussion threads, also entirely first-person, 3 of 7 rows carrying an amount. Grouped by kind of source rather than by date, the two families read very differently: 14 rows from discussion threads carry 5 amounts with a median of 9,000 USD, while 20 rows from city threads carry 17 amounts with a median of 22,000 USD. The gap is the source, not the year.", "caveat": "This check exists because the three windows look like a time series and are not one. Publishing that jump would have been the most convincing wrong number this dataset could produce.", "what_would_settle_it": "City threads of the same kind covering 2024 and 2025, or amounts drawn from discussion threads across all three windows. Until then, only the two recent windows may be compared with each other." }, { "check_id": "RP019-A-CHECK-008", "reviewed_at": "2026-09-10", "claim": "Spend 16,122 USD, bag 12,600 USD, ratio 1.3 to 1 all included and 1.9 to 1 once non-quota bags are excluded.", "claim_source": "A published collector video (YouTube), Kelly 25 in Waikiki", "tested_against": "Arithmetic on the figures given in the same source", "result": { "stated_spend": 16122, "stated_bag_price": 12600, "computed_ratio": 1.28, "stated_ratio": 1.3, "second_stated_ratio": 1.9, "spend_required_for_1_9": 23940 }, "verdict": "FIRST RATIO CONFIRMED, SECOND RATIO IMPOSSIBLE", "reading": "The headline ratio checks out: 16,122 divided by 12,600 is 1.28, which rounds to the 1.3 stated. That makes this the second row in the dataset whose ratio can be recomputed rather than merely believed. The second figure cannot be right: removing non-quota purchases from the numerator lowers a ratio and reaching 1.9 against a 12,600 bag would require a spend of 23,940. It is not retained.", "caveat": "The check is arithmetic, not a claim that the video is wrong: the error may sit in the transcript, in the fiche, or in what was said aloud. What is certain is that the two figures cannot both describe this spend.", "what_would_settle_it": "Reading the passage directly, which this surface could not do for this video." } ] }