Evidence quality is the degree to which information can reliably support a particular valuation conclusion. It is not the same as evidence quantity: three verified sales of genuinely comparable objects can be more persuasive than twenty loosely related listings, repeated database entries or unsupported asking prices.
Good evidence is relevant to the precise object and valuation purpose, drawn from the appropriate market, accurately interpreted, sufficiently recent, independently verified and transparent enough for another informed person to examine. The valuer must then explain how that evidence leads to the conclusion. Collecting prices without analysing them is research activity, not valuation reasoning.
Central principle
Weak evidence requires more assumptions. A defensible valuation does not hide those assumptions: it identifies them, tests them and expresses only the degree of precision that the evidence can support.
Evidence is not universally strong or weak. Its usefulness depends on what is being valued, for what purpose, in which market and at what date. A current dealer listing may be highly relevant to replacement cost because it shows what a collector would have to pay to source another example. The same listing is weaker evidence of open-market value because no buyer has yet accepted the price.
A forced-sale result may help estimate liquidation value but understate an orderly specialist-market value. A ten-year-old sale of the exact rare variant may be important because it proves market recognition, while a current sale of a common look-alike may be timely but materially less comparable. Evidence quality must therefore be judged against a stated basis of value, effective date, market and set of assumptions.
Collector scenario
A rare role-playing book appears in two records. One is a beautifully catalogued sale at a major international auction house, but the book is a different printing, has a restored cover and was sold within a heavily promoted creator archive. The other is a modest specialist-market listing for the exact printing, in similar condition, sold alone with clear photographs and a confirmed price.
The first record is more prestigious. The second may be more probative. The correct question is not which source looks most impressive, but which transaction most closely represents the economic circumstances of the subject.
Evidence should not be labelled simply “good” or “bad.” Each record has separate strengths and weaknesses. A weakness in one dimension may justify reduced weight rather than complete exclusion, but some defects - such as the wrong edition, an unknown price basis or a false sale status - can be decisive.
Did an exchange actually occur, and was the reported amount paid?
Strong evidence
A confirmed completed sale, paid invoice, published sold result or reliably documented private transaction.
Weak evidence
An asking price, auction estimate, hidden best offer, withdrawn listing or seller recollection.
Collector risk
Treating an offer or aspiration as proof of buyer acceptance can materially overstate value.
Can the object sold be identified precisely enough to compare it with the subject?
Strong evidence
The record identifies edition, issue, printing, variant, region, language, condition, completeness and distinguishing features.
Weak evidence
A generic title, distant photograph or category description such as ‘old role-playing book’.
Collector risk
False comparability arises when materially different editions or configurations are treated as the same market object.
Do the characteristics that form price align with those of the subject?
Strong evidence
Comparable identity, condition, completeness, authenticity, provenance, packaging and commercial context.
Weak evidence
A visually similar object from a different edition, grade, format, market level or demand profile.
Collector risk
Superficial similarity can conceal differences large enough to make the price unusable as a direct benchmark.
Does the evidence come from the market assumed by the valuation purpose?
Strong evidence
A sale in the venue, geography, market level and buyer population in which the subject would ordinarily transact.
Weak evidence
A local clearance sale used to infer specialist international retail value, or elite retail used to infer liquidation value.
Collector risk
The right object in the wrong market can produce the wrong valuation conclusion.
Does the transaction represent conditions close enough to the valuation date?
Strong evidence
Recent evidence from a market whose supply, demand and trading conditions remain comparable.
Weak evidence
An old result carried forward without considering market movement, inflation, currency or category change.
Collector risk
Chronologically old evidence may still matter, but unexplained reliance creates false confidence in a stale price level.
Is the reported number understood on a consistent basis?
Strong evidence
Hammer, buyer's premium, tax, shipping, currency, discounts and lot contents are separately identified.
Weak evidence
A headline figure with no indication whether it is hammer, total paid, asking price or bundled consideration.
Collector risk
Comparing unlike price bases can create an apparent spread that is only a reporting difference.
Who created the record, how was it produced and can it be checked?
Strong evidence
A traceable auction archive, invoice, preserved sold listing, specialist database or reasoned appraisal with identified sources.
Weak evidence
An undated screenshot, social post, search snippet, unsupported aggregation or unattributed AI summary.
Collector risk
Source prestige does not cure missing item-level detail, and niche records still require verification.
Is this a separate market observation rather than a copy of another record?
Strong evidence
A distinct transaction involving a different object, buyer, seller and sale event.
Weak evidence
The same auction result reproduced by several databases, articles and social posts.
Collector risk
Duplicate evidence can make one exceptional transaction look like a market trend.
Did the transaction occur under conditions consistent with the valuation assumptions?
Strong evidence
Adequate exposure, informed participants, ordinary motivations and no unusual relationship or pressure.
Weak evidence
Forced sale, charity bidding, poor cataloguing, restricted shipping, related parties, speculative bidding or possible non-payment.
Collector risk
A genuine price may still deserve reduced weight when the transaction circumstances are abnormal.
Could another informed person relocate the evidence and follow the reasoning?
Strong evidence
Source, date, lot number, images, sale status, price basis, adjustments and inclusion rationale are preserved.
Weak evidence
A remembered figure or cropped screenshot with no searchable reference or explanation.
Collector risk
Without an audit trail, the valuation cannot be reviewed, updated or defended later.
No hierarchy is absolute, because valuation purpose can change the order of importance. The following grades provide a practical starting point for a collector evidence file. They describe how the record should be used, not merely where it came from.
Grade A
A verified completed transaction for a highly comparable object in the appropriate market, with a clear price basis and useful documentation.
Grade B
A verified transaction with material but analysable differences, such as condition, packaging, age, geography or variant.
Grade C
Evidence that helps explain market position, replacement availability, trend or relative demand but does not directly prove the subject's exchange value.
Grade D
Unverified, ambiguous or unsupported information that may suggest where to investigate but should not support a value conclusion by itself.
Excluded evidence
Use a separate excluded status where a record is false, duplicated, materially irrelevant, rescinded, misidentified or too ambiguous to interpret. Retaining the record with an exclusion note is often better than deleting it, because it preserves the research trail and prevents the same misleading observation from being rediscovered later.
Auction records can provide dates, catalogue descriptions, photographs, hammer prices, premiums, estimates, lot numbers and evidence of competitive exposure. They are often powerful evidence of actual market transactions, particularly when the venue reaches the appropriate specialist audience.
Yet the result may be affected by reserve policy, house reputation, poor cataloguing, mixed lot contents, celebrity provenance, guarantees, irrevocable bids, bidding wars, weak attendance, after-sale negotiation or non-payment. A “record price” is an observation, not an automatic revaluation of every related object.
Dealer records help show replacement availability, specialist retail positioning, condition-sensitive pricing, authentication confidence, warranties and the premium attached to prepared, guaranteed inventory. They may therefore be especially relevant to retail or insurance replacement value.
A listing normally records what the dealer asks, not what a buyer paid. The price may include carrying cost, profit margin, restoration, tax, negotiation allowance, reputation and a long holding period. An item that remains listed for years may show the seller's aspiration more clearly than effective demand.
Guides and databases are useful when their methodology is transparent, transactions are traceable, variants remain separate, condition categories are defined and the data is current. They are weaker when they blend dealer retail with auctions, include asking prices as sales, repeatedly publish old values or conceal the records behind one unexplained figure.
A guide is normally best used as context or corroboration rather than as a substitute for direct evidence where direct evidence exists.
Specialist judgement can be crucial where an object is obscure, private-market or technically difficult to identify. The opinion becomes stronger when the expert discloses competence, inspection scope, conflicts of interest, supporting comparables, uncertainty and the reasoning that connects specialist observations to market value. “I know this market” is not a substitute for an analysable conclusion.
Preserve the page, images, description, date, price and sale status when the evidence is collected.
Compare photographs and identifiers to detect copied listings, fraud and duplicate records.
A platform result may conceal cancellation, non-payment, return or relisting.
A visible struck-through price can be materially above the confidential accepted amount.
One object offered on several platforms is still one market observation.
Repeated appearances may reflect failure to sell rather than repeated transactions.
Algorithms often surface highly priced or highly engaged listings instead of representative evidence.
Rarity, provenance, variant and condition claims may never have been independently checked.
Two objects labelled “very good” may be economically incomparable because seller grades are inconsistent. Compare observable defects: fading, staining, foxing, tears, creases, writing, corrosion, cracks, repairs, repainting, mould, odour, brittleness, packaging damage, missing parts and mechanical function.
Clear images, a specific defect description and a verifiable condition or grading report strengthen the record. Stock images, “good for its age,” “untested” and unexplained seller grades weaken it. Where condition cannot be established, use the transaction as a broad boundary rather than a direct benchmark.
A collectible may trade as an item only, complete boxed set, sealed example, partial set, mixed-edition assembly, restored example, dealer-created bundle or collection lot. These configurations can carry completeness premiums, scarce-component premiums, packaging premiums, provenance effects or bulk discounts. A set price should not be divided mechanically by the number of components.
Identity and authenticity must also be secure enough for both the subject and the comparable. Recognised expert opinion, grading verification, serial-number records, manufacturing characteristics, provenance, publication history and forensic testing may determine whether the objects occupy the same market. A record sale based on an attribution later rejected should not remain an unqualified anchor.
Scarcity matters only when collectors recognise and pay for it. A census of five known examples may support rarity, but value still depends on active buyers, cultural importance, visual appeal, completion pressure, liquidity, affordability and confidence in authenticity.
Population reports may also reflect low submission rates, resubmissions, regional gaps and unrecorded private holdings. Rarity becomes stronger valuation evidence when it is connected to verified sales, repeated demand or a demonstrated market premium.
High-quality evidence does not always produce the same figure. Differences may represent separate market levels, conditions, editions, dates, currencies, buyer populations or simply random variation in a thin market. The answer is not automatically to average everything.
Evidence density describes how many useful observations exist in the relevant market. Common collectibles may have hundreds of recent transactions; nearly unique material may have none. Sparse evidence does not prevent valuation, but it normally justifies a wider range, more proxy analysis, stronger assumptions and lower precision.
Evidence diversity can strengthen a conclusion when genuinely independent auction sales, private transactions, dealer records, offers, rarity data and category trends converge. Diversity does not rescue irrelevant or duplicated material.
Coverage matters as much as selection. A credible file records evidence above and below the conclusion, relevant unsold lots, contradictory transactions and the reasons for excluding particular observations. Silent cherry-picking - especially of record prices, high grades or only one market level - undermines confidence.
An unsold lot is not a transaction. It may indicate that demand was below the reserve, but without the reserve or bidding level the inference is limited. Poor presentation, wrong venue or authenticity concern may also explain failure.
A written, recent, unconditional offer from an identifiable capable buyer after inspection is stronger than a casual remark. Even a strong offer records one buyer's willingness, not necessarily the wider market.
Absence of results does not prove uniqueness, worthlessness or unlimited value. Document the databases, date range, terminology, variant spellings, venues and geographies searched so the negative result has defined meaning.
Where no direct comparable exists, build the evidence in layers rather than pretending one remote comparison is decisive. Start with the closest objects from the same maker, series, period or production method. Then examine attribute evidence such as premiums for signature, provenance, rarity, size, original packaging and historical importance.
Add evidence about the strength and depth of the relevant market, adjacent sales, specialist-dealer activity and - where appropriate - replacement or recreation cost. Expert judgement then connects the layers. The conclusion rests on a visible chain of inference, and the value range should widen as that chain becomes longer or more assumption-dependent.
Strong evidence
Several recent, verified and highly comparable transactions may support a focused point estimate or narrow range.
Moderate evidence
Direct sales with condition or configuration differences usually support a wider range and explicit adjustment reasoning.
Sparse evidence
Old, proxy or conflicting evidence may justify only a broad range subject to authentication, completeness or specialist review.
Myth
Reality
A record result may reflect exceptional condition, provenance, publicity or a bidding contest. It may define an upper boundary rather than the centre of the market.
Myth
Reality
The listings may be duplicated, unsold, incorrectly identified or copied from one underlying record.
Myth
Reality
Prestige helps with traceability and exposure, but specialist cataloguing errors still occur. Item-level verification remains necessary.
Myth
Reality
Its usefulness depends on methodology, variant separation, condition definitions, update frequency and access to the underlying transactions.
Myth
Reality
Service reputation, grading standards, holder generation, market acceptance and the visibility of underlying defects can still affect comparability.
Myth
Reality
A simple average gives weak and strong evidence equal influence and may combine incompatible markets, configurations and price bases.
Myth
Reality
Uniqueness affects supply, not demand. A unique object can remain commercially weak if few buyers value the distinction.
Myth
Reality
The market may be private, poorly indexed, geographically fragmented or described using different terminology.
A red flag does not always require exclusion. It requires investigation, qualification or reduced weight.
A valuation should remain reviewable after listings disappear and memories fade. Preserve enough information for another informed person to reconstruct the research and understand why each comparable mattered.
Specialist threshold
Consider a qualified specialist appraiser where the value is financially significant; the conclusion will be used for tax, probate, donation, insurance or litigation; authenticity, provenance or title is disputed; the market is private or opaque; direct comparables are scarce; or condition and restoration require technical assessment.
Historical valuation dates, related-party disputes, culturally sensitive material and exceptionally wide evidence ranges also justify independent expertise. The specialist should have appropriate category competence, disclose assumptions and conflicts, select the relevant market and provide a report suitable for the intended use.
“Three completed sales were identified. Comparable A is the same printing and similar in condition, but it sold four years before the valuation date. Comparable B is recent and well documented, although it lacks the subject's original packaging. Comparable C includes the packaging but is a later, more common printing. Two current dealer listings were reviewed but given less weight because they remain unsold. The conclusion is placed above Comparable B to reflect the packaging, but below Comparable A after allowing for its superior condition and stronger provenance. The market is thin, so a range is more supportable than a narrowly precise figure.”
Understand why completed transactions and advertised prices answer different valuation questions.
Return to the full valuation-evidence chapter and its sequence of collector research topics.
Continue to the problems created when direct sales are rare, irregular or privately reported.
Compare auctions, dealer records, private sales, databases, price guides and specialist sources.
Learn how condition, completeness, date, provenance and market differences affect comparable evidence.
Reconcile apparently inconsistent transactions without defaulting to a simple average.
Define the basis, date, market and assumptions that determine which evidence is relevant.
