Variant Evidence and Cataloguing

Variant evidence and cataloguing is the discipline of turning observed differences between collectibles into records that another researcher can inspect, test and revise. It begins with the object, but it does not end with noticing that two copies look different. The collector must determine whether the difference belongs to manufacture, issue, distribution or authorised presentation—or whether it is merely tolerance, deterioration, damage, replacement, repair or later ownership history.

Cataloguing then converts that research into a stable structure: an identity, a parent relationship, diagnostic features, supporting evidence, confidence, unresolved questions and a history of later correction. A trustworthy catalogue is therefore not a list of confident names. It is a transparent and maintainable evidence system.

Governing principle

Do not create a variant merely because a difference can be seen. Create or propose a variant when the difference can be shown to belong to the object's production, issue, distribution or authorised presentation history.

Orientation

The catalogue entry is already an argument

Collectors often treat cataloguing as administration: choose a title, assign a type, add a note and move on. Variant cataloguing is more consequential. The moment a record names a variant, it tells other collectors that a visible difference has been judged meaningful. That judgement can influence searches, want lists, prices, insurance descriptions, auction claims and later research.

A neat label can therefore manufacture authority. Once a catalogue says “rare early export variant”, later sellers may repeat the phrase without checking the object or the original evidence. The first record must expose the reasoning behind the name, not merely provide language that others can copy.

Collector scenario

The white-label “rare early variant”

Three apparently identical boxed items are compared. Two have blue manufacturer labels. One has a white label with different wording. A seller calls the white-label copy a rare early variant, and a forum post repeats the claim.

The careful record does not begin by creating a settled “white-label variant”. It records the label dimensions, exact text, placement, whether the label covers printed information, the source of each comparison image and whether the three examples are genuinely independent. It asks whether the label is manufacturer-applied, distributor-applied, retailer-applied or later. Until that is resolved, “white-label example under review” is a stronger catalogue entry than an unsupported rarity claim.

Foundational distinction

Observation, interpretation and catalogue conclusion

Layer 1

Observation

What can be directly seen, measured, photographed or transcribed.

Example: the logo is blue; the rear panel carries product code 1017-UK; no barcode is visible.

Layer 2

Interpretation

What the researcher believes those facts may mean.

Example: the code and distributor address may indicate a later United Kingdom distribution configuration.

Layer 3

Catalogue conclusion

The present classification adopted after weighing evidence and alternatives.

Example: catalogued provisionally as a UK distributor variant, pending additional independent examples.

These layers should remain separately visible because later researchers may agree with the observation while rejecting the interpretation. A record that collapses all three into “rare later UK variant” makes correction unnecessarily difficult.

Evidence hierarchy

What evidence can support a variant classification

No single evidence source is automatically decisive. Physical evidence may be altered; documents may show prototypes; provenance may date ownership rather than production; published references may repeat old errors. Strong classifications normally emerge when different forms of evidence reinforce one another.

The object is the starting point

Primary physical evidence

Records features that can be inspected, measured, photographed or transcribed directly from the collectible.

product, printer, mould, matrix, batch or serial codes
copyright lines, addresses, legal notices and country statements
materials, construction, dimensions, seams and fasteners
package structure, component combinations and artwork revisions

Collector caution

A visible feature is not automatically original, systematic or significant. Damage, repair, fading, replacement and normal manufacturing tolerance can imitate production differences.

Explains authorised release and production context

Contemporary documentary evidence

Connects physical differences to a dated release, specification, market, distributor, factory or documented change.

manufacturer and retailer catalogues
price lists, advertisements and product sheets
instructions, replacement-part lists and distributor notices
production records, drawings, invoices and shipping documents

Collector caution

Promotional material may show prototypes, reused photographs, planned features or stock imagery that never matched every released object.

Turns an isolated difference into a repeatable pattern

Comparative specimen evidence

Shows whether features recur across independently sourced examples and whether they appear in consistent combinations.

unrelated collections and independent provenance
sealed or long-held examples
matching feature clusters across regions
serial, batch or component patterns

Collector caution

Ten online listings may represent one repeatedly resold object. Count independent specimens, not repeated appearances or copied descriptions.

Supports dating, distribution and long-term configuration

Provenance evidence

Provides context through receipts, labels, correspondence, early photographs, acquisition records or documented ownership.

dated receipts and gift inscriptions
retailer, postal and customs labels
collection inventories and warehouse records
period photographs and owner correspondence

Collector caution

Provenance can date possession without proving that every component remained untouched or original to the package.

Extends comparison and identifies prior scholarship

Secondary published evidence

Provides established terminology, specimen studies, historical arguments and references to otherwise inaccessible objects or documents.

specialist books and collector catalogues
museum and institutional databases
auction catalogues and grading classifications
scholarly articles, club publications and documented forum studies

Collector caution

Published authority is not the same as verified authority. Later catalogues can repeat earlier errors without returning to the evidence.

Can preserve knowledge absent from formal records

Oral and community evidence

Captures first-hand testimony from employees, designers, printers, retailers, distributors and original purchasers.

former employee recollections
designer or printer interviews
original purchaser testimony
collector observations with documented specimens

Collector caution

Memory can be incomplete, reconstructed or influenced by later collector language. Record who spoke, their role, date, exact claim and whether corroboration exists.

Evidence form

Different kinds of proof answer different questions

Direct

Positive evidence

Shows that a feature is present: a revised code, new mould number, added distributor address, different component or printed barcode.

It is usually straightforward to photograph or transcribe, but still needs interpretation before it becomes a variant conclusion.

Absence

Negative evidence

Records the meaningful absence of an expected feature, such as no barcode, no country statement, no edition number or no later warning text.

Phrase absence narrowly. 'No barcode visible on the examined rear panel' is safer than 'pre-barcode edition' until chronology is independently established.

Comparison

Relational evidence

Becomes meaningful only against another version: a box is 5 mm deeper, a rulebook has four more pages or a label moves legal text.

Always state the comparison object or reference group rather than writing vague descriptions such as 'larger version' or 'early style'.

Pattern

Clustered evidence

Combines recurring features that travel together, such as a revised logo, new product code, changed address, altered insert and different barcode.

A stable cluster is usually more persuasive than a single isolated feature because it describes a production configuration rather than an incidental difference.

Order

Sequence evidence

Supports claims that one form preceded, followed or transitioned into another through dated records, serial ranges, copyright changes, advertisements or overlapping component combinations.

Do not infer order from visual style alone. An older-looking design may have remained in use in one market after a newer form appeared elsewhere.

Classification boundary

Object-level history is not variant-level identity

Individual specimen

Object-level evidence

Describes one physical copy and its particular history.

serial number, damage, fading or inscription
missing, replaced or repaired component
handwritten or retailer price label
ownership mark or later storage history

Production group

Variant-level evidence

Describes a recurring configuration shared by a recognisable group of objects.

standard product code or logo form
repeated text, address or market label
consistent package dimensions or component set
shared manufacturing mark or construction method

Collector judgement

Does the difference deserve a separate record?

A difference generally deserves a separate variant record when it materially changes the identity of the manufactured or issued object and can be recognised consistently. The threshold is category-dependent: coin catalogues may split die states that a toy catalogue would treat as ordinary production noise. Every catalogue should state its intended granularity.

Stronger candidates

Differences likely to justify separation

These distinctions usually connect to release, manufacture, authorised presentation or a repeatable production state.

formally designated editions, issues or printings
regional, language or retailer-exclusive releases
revised specifications, text, artwork or packaging
distinct factory, mould, material or component configurations
documented promotional or transitional forms
product-code changes tied to a new release

Usually insufficient

Differences that normally remain specimen observations

These may still be worth recording, but they do not automatically define a production group.

random colour fluctuation or registration error
normal mould wear, trimming or assembly tolerance
ageing, fading, damage or accidental omission
later repair, owner modification or component swapping
retailer-added pricing or an isolated manufacturing flaw

Granularity

Do not flatten product, release, variant and specimen into one record

A serious catalogue distinguishes different levels of identity. Shared facts belong at the highest appropriate level; distinguishing evidence belongs lower down. This avoids duplicated descriptions and stops one owned copy's condition or provenance from being mistaken for the definition of the variant.

Work or product concept

The underlying designed product or object type.

Edition, issue or release

A formally or practically distinct market release.

Variant

A systematic difference within or between releases.

State or transitional form

A production-stage change or intermediate configuration.

Individual specimen

One physical copy with its own condition, provenance and interventions.

Example hierarchy

Fantasy Board Game → 1984 English release → second printing → UK packaging variant → transitional insert state → collector's individual copy

Identification

Separate diagnostic evidence from merely descriptive detail

A catalogue becomes practical when it tells collectors which features carry the classification and which do not. Recording everything in one undifferentiated notes box encourages weak identification and accidental over-splitting.

Decisive feature

Separates the variant from its nearest alternatives and should be present for the classification to apply.

Example

A product code ending '-UK' combined with a revised distributor address.

Supporting feature

Usually correlates with the variant but cannot identify it safely on its own.

Example

A four-panel insert that also occurs in a transitional configuration.

Descriptive feature

Worth recording for completeness but not relevant to the classification boundary.

Example

A routine internal packaging texture shared across several releases.

Unreliable feature

Known to vary through ageing, printing tolerance, camera conditions or ordinary manufacture and should not drive identification.

Example

Small ink-shade differences visible only in uncontrolled photographs.

Specimen-history feature

Belongs to one owned copy rather than to the production group represented by the variant record.

Example

An owner's price sticker, repair, inscription, fading or missing component.

Recording practice

Photographs, transcriptions, measurements and components

Images

Photograph the evidence, not just the object

Capture full views and close-ups of codes, legal text, marks, mould details, component layouts and the exact distinguishing feature.

Record photographer, date, specimen identifier, source, crop or colour correction and whether scale is shown. Colour alone is unsafe under uncontrolled lighting and white balance.

Text

Transcribe without silently normalising the clue away

State whether the transcription is exact, normalised, partial, translated or reconstructed.

Preserve significant capitalisation, punctuation, spelling errors, line breaks, abbreviations and trademark symbols. Use visible conventions for illegible, obscured or uncertain text.

Dimensions

Make measurements reproducible

Record unit, orientation, precision, measuring method and whether packaging is compressed, unfolded or measured across protrusions.

Test small differences against manufacturing tolerance, deformation and instrument precision before treating them as diagnostic.

Contents

Separate expected components from observed contents

Record expected original components, what this specimen contains, what is missing, what may be substituted and what remains uncertain.

A sealed example is strong evidence, but one sealed object does not automatically define every package in the run; transitional or substituted factory contents may still occur.

Catalogue architecture

What a variant record should preserve

Identity and placement

Locates the record in the catalogue hierarchy without forcing uncertain conclusions into permanent identifiers.

persistent catalogue identifier
preferred name and alternative or deprecated names
parent product, release, edition or printing
manufacturer, publisher, brand, model, region and language
stated, release, manufacture and estimated date fields kept distinct

Variant definition and diagnostics

Explains exactly why the record is separate and how collectors can distinguish it from the nearest alternatives.

concise variant definition
decisive, supporting and unreliable features
exact transcriptions, measurements and component expectations
differentiation notes for commonly confused variants
known internal variation that does not justify further splitting

Evidence and reasoning

Makes the conclusion reproducible by showing what was observed, how it was interpreted and what sources were used.

object observations and specimen references
comparison basis and specimen count
documentary, provenance, published and oral sources
alternative explanations and contrary evidence
evidence provenance: who entered, supplied, verified or inferred each claim

Confidence, status and relationships

Separates strength of evidence from editorial workflow and connects the record to neighbouring forms.

classification confidence with written reasoning
workflow status such as proposed, under review, accepted or deprecated
preceded by, followed by, corrected by or transitional between
regional or language equivalent, reissue or shared-component relationship
commonly confused with and supersedes links

Maintenance and accountability

Preserves the catalogue as a research history rather than an unexplained set of current assertions.

editor and review date
previous and new values
reason and evidence for change
merge, split and redirect history
outstanding questions and next review trigger

Naming rule

Stable, neutral and diagnostic beats dramatic

Prefer names such as “blue-logo box variant”, “revised copyright-text printing” or “bilingual Canadian package”. Avoid “rare version”, “old one”, “prototype” without production evidence, or “European variant” when only one country has been confirmed.

Use a persistent identifier that survives later renaming, merging or reclassification. Do not embed uncertain conclusions such as FIRST-EDITION-UK-1984 into an identifier that should remain stable if the research changes.

Confidence

Show how strong the classification is—and why

Confidence describes the evidential strength of the classification, not the researcher's enthusiasm. Keep it separate from workflow status: a strongly supported proposal may still be awaiting editorial review, while an accepted legacy record may later become disputed or obsolete.

Confirmed

Strong physical evidence and independent corroboration support the classification, normally through repeated specimens, authoritative documentation or both.

Catalogue language

Confirmed from seven independently sourced specimens, including two sealed examples and a contemporary distributor catalogue.

Strongly supported

The evidence is consistent and persuasive, but one significant element—such as exact chronology or complete production context—remains incomplete.

Catalogue language

Strongly supported as a UK packaging variant; exact introduction date remains unresolved.

Probable

The proposed interpretation is more likely than the alternatives but lacks decisive confirmation.

Catalogue language

Probably a later distributor configuration, based on address, code and dated purchase evidence.

Provisional

A meaningful difference has been recorded, but repeatability, cause, boundary or production significance is still under review.

Catalogue language

Provisional white-label cluster; four independent examples recorded, with date sequence unresolved.

Disputed

Qualified researchers or credible sources disagree. The catalogue should preserve the competing interpretations and the evidence for each.

Catalogue language

Disputed classification: treated by some references as a regional issue and by others as retailer relabelling.

Rejected or obsolete

Evidence shows that the proposed variant resulted from alteration, damage, duplication, misassembly or an interpretation no longer supported.

Catalogue language

Former 'Type 2' classification rejected; colour difference is now attributed to fading. Retained as a deprecated search term.

Proposal threshold

Minimum evidence for creating a provisional variant

A provisional record is useful when it preserves a real question without disguising that question as a settled answer. A proposal should normally include:

at least one clear distinguishing feature and photographs of the relevant area;
a named comparison form and explanation of the difference;
an account of why the feature appears original rather than condition-related or later;
specimen count, source independence and provenance where available;
known alternative explanations, submitter identity and confidence;
unresolved questions and the evidence needed for review.

There is no universal specimen count. One uniquely documented prototype may be real; ten objects assembled from swapped components may not establish any legitimate production variant.

Research method

A reproducible workflow from question to catalogue record

Define the research question

Ask what must be resolved: separate variants, chronology, regional issue, original component, printing difference or unexplained anomaly.

Establish the comparison group

Gather the suspected form, standard form, neighbouring releases, regional equivalents and known transitional examples.

Record observations neutrally

Transcribe exact text, measure dimensions, list codes, materials, marks and component combinations before applying names such as 'first edition' or 'error'.

Test alternative explanations

Actively test wear, fading, repair, component swapping, tolerance, retailer alteration, later repackaging, restoration and counterfeit possibilities.

Seek independent repetition

Find unrelated specimens carrying the same feature or feature cluster. Record source independence and contradictory examples.

Gather documentary and provenance evidence

Search contemporary catalogues, instructions, advertisements, records, receipts, dated labels and early photographs for production and distribution context.

Build a feature matrix

Compare decisive, supporting, variable, unknown and unreliable features across the specimen group.

Form a cautious hypothesis

State the proposed classification, supporting evidence, alternatives, unresolved problems and current confidence.

Create a provisional record

Do not wait for perfect knowledge. Give uncertainty a visible, searchable place without presenting the proposal as settled fact.

Review, challenge and revise

Add specimens, invite corrections, preserve changes, merge or split records when warranted and retain old names as redirects or deprecated terms.

Comparison

A compact feature matrix can expose the real pattern

Use a table only for compact, repeatable data. Mark decisive, supporting, variable and unknown features elsewhere in the record so the table does not imply that every row is equally diagnostic.

Feature	Variant A	Variant B	Transitional form
Logo	Black	Blue
Copyright	1984	1987
Barcode	None	Printed	Applied label
Insert	Early	Revised
Product code	1017	1017-UK
Status	Confirmed	Provisional

Complex cases

Transitions, anomalies and disputed classifications need somewhere to live

Asynchronous change

Transitional configurations

New packaging may appear with old instructions, revised parts with earlier accessories, or a barcode label on remaining old stock.

Repeated matching examples, sealed or well-provenanced copies and chronological plausibility can show a genuine transition. Do not force every intermediate object into one of two 'pure' versions.

Unresolved observation

Errors, anomalies and one-off differences

Use categories such as production error, printing anomaly, suspected alteration, experimental piece or unresolved specimen when normal variant status is not justified.

This preserves the observation without prematurely elevating it. Later evidence may confirm a variant, demonstrate an alteration or identify a prototype or counterfeit.

Catalogue inflation

Over-splitting

Creates separate records for random, trivial, condition-related or unsupported differences.

The result is false rarity, confused searching and a catalogue whose distinctions cannot be reproduced.

Hidden structure

Over-lumping

Combines materially different printings, regional releases, revisions, factory forms or component configurations under one broad record.

Sometimes the solution is not another flat variant but a parent record with subordinate states or feature groups.

Evidence discipline

Correlation, absence and rarity are easy to overstate

Correlation

Features travelling together do not prove causation or exact sequence

If every recorded barcode example has a revised logo, the sample shows correlation.

It does not prove both changes happened simultaneously, that no intermediate exists or that every barcode example must share the logo.

Absence

Not recorded is not the same as impossible

Prefer 'not observed among the 23 specimens examined' or 'no confirmed example has yet been recorded'.

Avoid absolute claims when the census, documentary search or photographic coverage is incomplete.

Scarcity

Identity comes before rarity

Keep production quantity, distribution, survival, recorded examples, sale frequency and condition rarity separate.

A census of four known examples proves only that four examples are currently documented.

Census and sources

Build evidence trails that survive beyond the original researcher

Specimen census

Record known examples without pretending they are the whole population

Useful fields include anonymised specimen ID, source, provenance, serial or batch number, diagnostic features, component completeness, seal status, photographs and verification date.

A census can reveal feature clusters, serial ranges, geographic patterns, false reports and transitional sequences, but it is not automatically a production total or rarity estimate.

Source citation

Attach important claims to the exact evidence used

Record author or organisation, title, edition, date, page or entry, archive reference, URL and access date where relevant.

For auction evidence, preserve lot number, sale, date, images and the distinction between the seller's description and features independently verified from photographs or inspection.

Authority control

Keep repeated entities consistent

Manufacturers, publishers, designers, factories, distributors, brands, product lines, places and source publications benefit from authority records.

Preferred names, alternatives, active dates, addresses and corporate relationships prevent one entity fragmenting across inconsistent spellings and historic names.

Structured retrieval

Use fields and relationships for facts collectors need to search

Structured data can retrieve all variants without barcodes, packages carrying one distributor address or provisional records based on a single specimen.

Free text remains essential for nuance, but decisive evidence should not be buried entirely inside prose.

Digital assistance

Automation can find candidate differences, not validate variants

Image comparison and automated analysis may detect text changes, layout shifts, logo differences, colour regions, mould details, dimensions or component substitutions. These tools are useful for triage and comparison at scale.

Their output remains candidate evidence. Camera angle, lens distortion, compression, lighting, cropping, scaling, dirt, wear and restoration can all produce false differences. Human review and object-level verification remain necessary before a catalogue classification changes.

Governance

A catalogue must be able to change without losing its history

New evidence may show that two records describe the same production form, or that one broad record conceals two stable groups. Merging and splitting are signs of healthy research when they are documented rather than silently performed.

Merge

Combine records when separation no longer survives scrutiny

Merge duplicated names, condition-created differences or records shown to describe the same production form.

Preserve old identifiers, add redirects, retain prior descriptions and explain the evidence behind the decision.

Split

Separate a broad record when consistent groups emerge

Split when regional, textual, component or production differences recur systematically and establish reproducible boundaries.

Migrate linked specimens carefully, preserve the former record history and explain why the earlier model is no longer sufficient.

Myth versus reality

Myth: a good catalogue sounds certain.

Confidence appears professional, so provisional language should be hidden until the answer is complete.

Reality: a good catalogue makes limits visible.

Trust comes from showing observations, sources, alternatives, unresolved questions and how the conclusion changed—not from removing every sign of uncertainty.

Worked record

What an evidence-led catalogue entry looks like

Catalogue ID

CV-00427

Preferred name

Blue-logo UK packaging variant

Parent product

Example Adventure Set

Status and confidence

Accepted · strongly supported

Variant definition

Distinguished by a blue upper-left logo, revised UK distributor address, product code EX-1017-UK and a printed barcode on the rear panel.

Diagnostic evidence

blue rather than black logo;
London distributor address;
product code ending -UK;
printed EAN barcode;
revised four-panel instruction insert.

Known variation within the record

Minor differences in blue ink shade and carton depth are recorded but are not currently diagnostic.

Comparison

The preceding UK package has a black logo, earlier distributor address and no barcode. A provisional transitional form has the blue logo and revised address but uses an applied barcode label rather than a printed barcode.

Evidence and dating

Nine independently sourced specimens, two sealed examples, three dated purchase records from 1988–1989 and a 1988 distributor catalogue support an estimated 1988–1990 range. The 1987 copyright statement is recorded separately and is not treated as the sole manufacture date.

Outstanding question

It remains uncertain whether stickered-barcode examples preceded the printed package or were distributed concurrently.

Revision history

2025-02-11: provisional record created;
2025-06-03: third independent specimen added;
2026-01-18: sealed specimen confirmed insert combination;
2026-04-07: status changed to accepted.

Failure modes

Common ways variant catalogues lose trust

Conclusion bias

Starting with the name you want to prove

Labels such as 'first edition', 'prototype' or 'rare error' narrow the search too early and encourage confirming evidence to be noticed while alternatives are ignored.

False repetition

Counting copied claims as independent evidence

Multiple listings, forum posts and dealer descriptions may all trace back to one object or one unsupported sentence.

Category confusion

Treating condition, repair or ownership history as production identity

Fading, replaced parts, retailer labels and later intervention may be important records without defining a variant.

Chronology shortcut

Assuming copyright date, simple design or no barcode proves 'earliest'

Printed dates can remain unchanged, older designs can persist by region and absence requires independent chronological support.

Market contamination

Allowing seller terminology or value to determine validity

Identity must be established before scarcity or price. Commercial names may be retained as search terms but must not substitute for evidence.

Lost accountability

Deleting rejected names or silently rewriting records

Without redirects and revision history, old claims continue circulating while users cannot understand why the classification changed.

Specialist threshold

When collector research should escalate

Seek category, authentication, conservation, printing, manufacturing or scientific expertise when the classification depends on evidence that ordinary inspection cannot resolve safely or reliably. Escalation is particularly appropriate when:

opening, dismantling, cleaning or sampling could damage the object;
adhesives, inks, paper, metal, paint or plastic age must be distinguished;
a prototype, proof, counterfeit or replacement component is suspected;
the claim materially affects authenticity, market value or public scholarship;
specialist archives, factory records or technical process knowledge are required;
credible researchers disagree and the catalogue needs an accountable review decision.

Key takeaways

Record observations before interpretations and keep both separate from the catalogue conclusion.
Require evidence of production, issue, distribution or authorised presentation significance—not merely a visible difference.
Separate individual specimen history from variant identity.
Use independent specimens, contemporary records and feature clusters where possible.
Make decisive, supporting, descriptive and unreliable features explicit.
Keep confidence separate from workflow status and explain the reasoning behind both.
Preserve sources, evidence provenance, contrary examples and unresolved questions.
Allow provisional records, disputes, merges, splits and rejected classifications to remain visible and traceable.
Establish identity before discussing rarity, desirability or value.
A trustworthy catalogue is reproducible and revisable, not merely confident.

Continue learning

Errors, Anomalies and Transitional Examples

Review how collectors separate unusual objects, production errors and transitional configurations before assigning variant status.

→

Back to Editions and Variants

Return to the Editions and Variants sub-domain and its full topic sequence.

→

Significance, Scarcity and Collector Interest

Continue to judging whether an established distinction is historically meaningful, scarce or important to collectors.

→