← All conjectures · Everyday & documentary writing
The census of unseen dies
Status is derived only from the shepherd-authored triage/prediction data above -- community submissions and claims are a separate overlay and can never change it (see the participation panel below).
“Anticipated in print” means the direction of this claim already appears in scholarship, while this exact test has never been run — it records prior art, not proof; nothing here counts as supported or falsified until a registered resolution says so.
Claim (verbatim)
A die-study is a census of the tools that made a coinage: collect every findable specimen of an issue, sort them by the hand-cut dies that struck them, and the die count plus the specimen-per-die distribution becomes a statistical window onto the original size of a minting that no ancient text recorded. The mathematics is the unseen-species arithmetic of Good (1953), brought into numismatics by Carter (1983, ANS Museum Notes 28) and systematized by Esty (Numismatic Chronicle 1986; and his geometric-model treatment in de Callatay's 2011 Quantifying Monetary Supplies in Greco-Roman Times): dies observed once - the singletons - measure how much of the original toolkit the sample has never seen. The discipline's own master compilation, Francois de Callatay's Recueil quantitatif des emissions monetaires hellenistiques (1997), tabulates specimen counts and die counts for hundreds of Greek die-studies, and it thereby quietly certifies the incompleteness of the record it summarizes: most studies ride on only a few coins per die, deep in the regime where the estimators say a substantial minority of dies - and hence of the struck output - remains invisible. The claim: the flagship quantitative instrument of Greek numismatics is, read against its own statistics, a register of loss - the typical published die-study has not seen at least one die in seven of the coinage it reconstructs, and the field's totals for ancient minting are systematically floors, not counts.
Prediction clause (verbatim)
Prediction: recomputing across the die-studies tabulated in the Recueil quantitatif (1997) with both specimen count n and obverse die count d reported, the median coins-per-obverse-die ratio n/d will fall below 6, which under Esty's geometric model implies a median unobserved-die share of at least one-seventh (primary clause: median n/d below 6 across the compiled studies; the verdict follows it); secondary: in at least half of the studies the model-implied unobserved-die share exceeds 15%. Disambiguation pinned: the geometric model is the pinned estimator for the conversion (no cherry-picking among estimators after seeing the data); studies reporting reverse dies only are excluded. If fewer than 200 studies in the Recueil report both n and d, the test voids as a coverage failure rather than resolving.
Kill-dataset (verbatim)
Kill: de Callatay's Recueil quantitatif des emissions monetaires hellenistiques (1997) - extract n and d per study, compute the n/d distribution and the Esty-geometric coverage per study, and test the median against the threshold; Esty 1986 and Carter 1983 supply the pinned formulae.
Browse the registry: Numismatics →
Nobody has run this test. The kill-data is named above. If you can run it — or you know the paper that already settles it — claim the kill or submit the prior scholarship. Kills and prior scholarship are credited here, by name, as they come in.
How this item is catalogued
Catalogue tags are curatorial metadata that make the conjecture findable. They are not findings, not a novelty verdict, and naming a source family is not a promise that the data exist or are accessible. Each dimension below says whether it was assigned, reviewed with none applying, uncertain, or not yet reviewed, and who recorded it.
- Place & era tags — Assigned
- Aegean & Greece; Mediterranean; Iron Age & Antiquity Tagged by claude-fable-5
- Dating — Phenomenon
- Dated c. 330 BCE – 30 BCE Hellenistic Greek coinage era ('emissions monetaires hellenistiques'), conventionally 323-31 BCE (Alexander's death to Actium). Tagged by claude-sonnet-5 + claude-fable-5 (gate)
- Subject — Assigned
- Everyday & documentary writing (from the generation section)
- Finer subject topic — Reviewed — none applies
- No listed Atlas subject page is squarely about die-statistical numismatics. Tagged by claude-opus-5 draft (requested max effort) adjudicated by claude-fable-5-1 driver, lane 1da5; runtime not independently attested
- Claim level — Assigned
- Instrument claim The prediction is a property of the discipline's own published die-studies, namely that their samples leave a substantial share of dies unobserved, so the object is modern scholarly practice. Tagged by claude-opus-5 draft (requested max effort) adjudicated by claude-fable-5-1 driver, lane 1da5; runtime not independently attested
- Method — Assigned
- Capture-recapture and unseen-species estimation The apparatus is unseen-species estimation (Good, Carter, Esty's geometric model) converting coins-per-die into an unobserved-die share. Tagged by claude-opus-5 draft (requested max effort) adjudicated by claude-fable-5-1 driver, lane 1da5; runtime not independently attested
- Evidence source — Assigned
- Numismatics Kill test loads de Callatay's Recueil quantitatif tabulations of specimen and die counts per published die-study. Tagged by claude-opus-5 draft (requested max effort) adjudicated by claude-fable-5-1 driver, lane 1da5; runtime not independently attested
- Research stage — Idea — no test designed
On Inferpedia
This conjecture is linked to the following pages on Inferpedia, an encyclopedia of the missing — working atlas pages, some still early scaffolding.
Provenance
Run: Fresh agent generation · model: claude-fable-5
Fresh blind generation by claude-fable-5, 2026-07-18, coins-as-record wave: coinage treated as a knowledge record in its own right - rulers, mints, and events attested only on struck metal, and the die-study estimation of how much minting is lost - anchored to the registered Numismatics instrument family (inst-numismatic-hoard-databases, coinhoards.org) and extending it with the open type-corpora named per kill (Kushan Coins Online, OCRE, CRRO, PELLA, RPC online, PAS, CHRE, Kiel Runenprojekt) plus the printed censuses (Bopearachchi 1991, SNS, Vondrovec 2014, Munro-Hay/Juel-Jensen 1995, SICA 1/Walker 1941, Klat 2002, Hartill 2005, Frolova, de Callatay 1997, Buttrey 1976). Disjointness discipline against the 1,444-item corpus: the existing numismatic items all use coins as instruments for OTHER questions and are avoided operation-by-operation - cj-003/cj-004 own hoard-deposition-as-behavior (my ord 7 measures deposition as an archive filter, metal-share ratio, not fear timing or Gresham dynamics); w05 art ords 7/8/29 own coins-as-image-medium and imitation-legend degradation (my ord 16 is a corpus-share count on the Kiel runic database, different instrument and operation, degradation not claimed); w09-038 owns dirham-hoard circulation zones (my ord 12 uses Klat's mint-date lattice, no hoards); w11-021 owns debasement-vs-books; w17-023 owns Sogdian legend persistence (no Sogdian item here); w26-009 owns hoard-mathematics-on-paper; breadth-africa-americas ord 5 owns the Kilwa coin-vs-chronicle recension audit (my ord 10 is the inverse operation on Aksum - coin-only majority of a king list with no chronicle at all); breadth-cinner-w2 owns Rabatak-as-Bactrian-language-witness (my ord 3 uses Rabatak only as the confirmation event for a corpus-anonymity count on Kushan Coins Online); fablemax-004 'The singleton census' owns Chao-on-Pinakes witness statistics - my ord 4 is the same capture-recapture lineage deliberately carried to a different medium, estimator (Esty geometric/Carter), and instrument (die-links in de Callatay's Recueil), titled 'The census of unseen dies' and flagged here as kin. Steer substitutions: the Crawford-RRC-vs-modern-die-study direction audit was dropped (direction of Crawford's error not confidently known) in favor of the Crepusius control-numeral benchmark (ord 5), where the mint's own die numbering supplies ground truth; the Hellenistic monogram-decipherment candidate was dropped as LOW-confidence (no countable resolved-share inventory exists yet - future instrument-building); the Chinese quota-vs-survival join stays as frame inside ord 13 (cast coinage defeats die-statistics; the Songshi quotas carry output knowledge instead). Honesty flags: MODERATE on exact inventory sizes from memory - Recueil median n/d (ord 4), Crepusius maximum numeral near 519 and estimator band (ord 5), SNS unresolved-signature share (ord 8), Alchon name counts and Talagan scroll contents (ord 9), Aksumite issuer count (ord 10), Darabgird era conversion (ord 11, resolved against the claim inside the prediction), Klat mint total (ord 12), Hartill year-variety resolution (ord 13), Frolova year coverage (ord 15), Kiel bracteate share (ord 16) - each carries an explicit void guard; ords 2 and 9 declare Kill (not yet built) since their registers must be hand-compiled from the named publications. Open-database runnable this week: ords 3 (Kushan Coins Online), 6 (OCRE), 7 (PAS+CHRE), 14 (PELLA), 15 (RPC portion), 16 (Kiel Runenprojekt).
Novelty / leakage triage
anticipated in the literature — this exact test has never been run
The estimators are published methods, and both de Callatay and Esty state that typical Greek die-studies ride on few coins per die, making published totals floors; the Recueil tabulates n and d per study, so every ingredient is in print. But the pinned computation - median n/d below 6 across the full compilation, converted through the pinned geometric model to a median unseen-die share of at least one-seventh - was not located as ever published; de Callatay's own aggregate surveys of the Recueil give corpus totals, not this distributional statistic.
Sources cited by the triage
- F. de Callatay, Recueil quantitatif des emissions monetaires hellenistiques (Wetteren: Moneta, 1997)
- W.W. Esty, 'How to Estimate the Original Number of Dies and the Coverage of a Sample', Numismatic Chronicle 166 (2006)
- G.F. Carter, 'A Simplified Method for Calculating the Original Number of Dies from Die Link Statistics', ANS Museum Notes 28 (1983)
Predictions
No prediction registered yet.
Weigh in
No community feedback yet.
New here? Create an account first
Create an account or sign in and your feedback is tied to you — you can track it, get replies, and claim this conjecture so others know you’re working on it. Prefer not to? Just leave your take below as a guest — only the name you type is shown.
Add your take
Posted immediately (spam is removed). Community feedback is never an adjudicated verdict and never changes this conjecture's triage label or status above.