The Gauntlet Research Methodology

Updated 2026-07-11

Gauntlet Gallery's institutional methodology for historical, object-level, authentication, provenance, condition and market research.

How We Build Historical, Market, Authentication and Object-Level Intelligence

Institutional document: GG-002
Updated: July 11, 2026

This institutional standard is published by Gauntlet Gallery. It is part of the The Gauntlet Curatorial Philosophy, The Gauntlet Research Methodology, The Gauntlet Cataloging Standard, Gauntlet Photography and Presentation Standards, The Gauntlet Authentication Framework, The Gauntlet Historical Preservation Policy, The Gauntlet Collector Experience and The Gauntlet Field Exhibition Philosophy series.

Introduction

Collecting depends on information.

A collector may first respond to an object visually or emotionally, but nearly every serious acquisition eventually raises factual questions:

• Who made it?
• When was it created?
• Was it authorized?
• How many were produced?
• Is the signature authentic?
• What materials were used?
• Has the object been altered?
• Where has it appeared before?
• How often does it sell?
• What have comparable examples actually sold for?
• Which claims are documented?
• Which claims remain uncertain?

The quality of the answers depends on the quality of the research.

Gauntlet Gallery treats research as part of the object itself.

An artwork, signed instrument, designer figure or historical artifact is not fully represented by photographs and a price. Its identity is also carried by the documentary record surrounding it:

• release information,
• edition details,
• exhibition history,
• authentication records,
• provenance,
• market activity,
• condition history,
• and the broader cultural context that explains why the object matters.

Our research process is designed to assemble that record as completely and responsibly as possible.

The purpose is not to create an appearance of certainty where certainty does not exist.

The purpose is to:

1. gather the strongest available evidence,
2. distinguish reliable sources from weak ones,
3. reconcile conflicting information,
4. make assumptions visible,
5. preserve useful historical records,
6. and provide collectors with a more defensible basis for judgment.

Gauntlet’s research methodology applies across four primary areas:

• historical and cultural research,
• object identification and cataloging,
• authentication and provenance research,
• and market and pricing analysis.

Each area uses different sources and different standards of proof.

A museum archive may be excellent for exhibition history but irrelevant to a recent market price.

A major auction result may be useful for valuation but incomplete for original release details.

An artist’s studio may be authoritative about an edition but may not maintain a comprehensive secondary-market record.

A collector forum may contain rare first-hand information, but that information must be independently tested.

Research quality therefore depends not merely on how many sources are found, but on whether each source is appropriate to the claim being evaluated.

This document explains how Gauntlet approaches that task.

1. Research as an Institutional Function

Gauntlet does not treat research as a final step added after an object is acquired.

It is a continuing institutional function.

Research informs:

• what we collect,
• what we decline,
• how we describe objects,
• how we distinguish editions,
• how we evaluate authenticity,
• how we price,
• how we photograph,
• how we educate collectors,
• and how we preserve records for future owners.

The process begins before acquisition and may continue after sale.

New information can emerge.

An archive may be updated.

A missing exhibition catalogue may become available.

A previously hidden sale price may be confirmed.

A signature opinion may be revised.

A production detail may be clarified by the artist or publisher.

Gauntlet’s records are therefore treated as living records.

They should become more accurate and complete over time.

Research Before Acquisition

Before acquiring or accepting an object, we seek to understand the central claims.

For example:

• Is the work a recognized edition?
• Does the edition size match established records?
• Is the item an official release?
• Does the authentication number verify?
• Is the signature consistent with the claimed signer?
• Is the object complete?
• Does the provenance make sense?
• Are there known counterfeits?
• Is the asking price supported by the market?

This process helps prevent weak or misleading objects from entering the collection.

Research During Cataloging

Once an object is accepted, research becomes more detailed.

We document:

• title,
• creator,
• date,
• medium,
• dimensions,
• edition,
• signatures,
• markings,
• publisher,
• manufacturer,
• condition,
• authentication,
• provenance,
• release history,
• historical context,
• and available market evidence.

The object is compared with known examples.

Any discrepancies are investigated.

Research After Sale

The research record remains relevant after ownership changes.

Collectors may later need:

• documentation for insurance,
• authentication verification,
• resale support,
• condition history,
• estate planning,
• or provenance transfer.

A well-researched object is easier to preserve and easier to represent accurately in future transactions.

Gauntlet’s objective is that each object should leave with a stronger record than it had when it arrived.

2. The Research Question Comes First

Good research begins with a precise question.

Weak research often begins with a broad search and then collects whatever appears first.

Gauntlet instead defines the claim that must be evaluated.

For example:

“Is this object authentic?”

is too broad.

The useful questions are:

• Was this object manufactured by the claimed company?
• Was this print authorized by the artist?
• Is this handwritten signature genuine?
• Does the certificate correspond to this specific item?
• Is this object from the stated edition?
• Was this artifact actually used during the mission?
• Is this an original release or a later reproduction?

Each question may require a different source.

Similarly:

“What is this worth?”

must be divided into more specific questions:

• What have truly comparable examples sold for?
• How recent are those transactions?
• Were the examples authenticated?
• Were they framed?
• Were fees included?
• What condition differences existed?
• How frequently does the object trade?
• Is the current market rising, falling or inactive?

Precise questions prevent unrelated facts from being treated as evidence.

3. The Gauntlet Source Hierarchy

Not all sources carry equal weight.

Gauntlet uses a source hierarchy to evaluate reliability.

The hierarchy is not absolute.

A lower-tier source may contain important first-hand evidence. A higher-tier source may contain an error. However, the hierarchy provides a disciplined starting point.

Tier One: Primary and Authoritative Sources

Tier One sources have the closest direct relationship to the claim being evaluated.

Examples include:

• the artist,
• artist studio,
• artist estate,
• original publisher,
• original printer,
• official manufacturer,
• issuing gallery,
• museum archive,
• government agency,
• NASA,
• mission contractor,
• record label,
• official tour organization,
• official foundation,
• recognized catalogue raisonné,
• original certificate,
• original invoice,
• production document,
• direct first-hand correspondence,
• signed statement from a knowledgeable participant,
• or contemporaneous legal or corporate filing.

These sources generally receive the greatest weight.

However, authority must still match the claim.

An artist may be authoritative about whether a work was created or authorized but may not know every secondary-market transaction.

A manufacturer may confirm a release but may not authenticate every example in circulation.

A museum may provide reliable historical context but may not have inspected the object being sold.

Tier One evidence is strongest when it directly addresses the question.

Tier Two: Highly Credible Specialist Sources

Tier Two sources are independent but professionally recognized.

Examples include:

• major auction houses,
• respected specialist dealers,
• recognized authentication firms,
• established galleries,
• museum publications,
• academic books,
• peer-reviewed research,
• exhibition catalogues,
• artist monographs,
• specialist reference books,
• established market databases,
• experienced conservators,
• and subject-matter experts with verifiable credentials.

Tier Two sources can be highly persuasive, particularly when several independent sources agree.

They are often essential when primary records are unavailable.

Tier Three: Reputable Secondary Publications

Tier Three sources include established publications that report, review or summarize cultural activity.

Examples include:

• major newspapers,
• respected magazines,
• recognized art publications,
• specialist media,
• contemporaneous exhibition reviews,
• industry reporting,
• and established local press.

These sources are useful for:

• dates,
• exhibition coverage,
• quotations,
• cultural context,
• public reception,
• and evidence that an event occurred.

They may be less reliable for precise edition details, market pricing or authentication unless they cite stronger underlying evidence.

Tier Four: Collector and Community Sources

Tier Four sources include:

• collector forums,
• specialist social groups,
• hobby communities,
• archived marketplace discussions,
• collector spreadsheets,
• enthusiast databases,
• informal interviews,
• and long-running blogs.

These sources can contain exceptional knowledge.

Collectors sometimes preserve details that official institutions failed to record, including:

• release dates,
• packaging variants,
• retailer exclusives,
• production errors,
• historical asking prices,
• obscure colorways,
• and first-hand purchase information.

However, these sources require verification.

A repeated collector claim is not automatically true.

Gauntlet uses community sources to generate leads, identify patterns and locate evidence, not as unquestioned authority.

Tier Five: General Commercial and User-Generated Sources

Tier Five sources include:

• marketplace listings,
• resale descriptions,
• social posts,
• unsourced blogs,
• image aggregators,
• general price-guide sites,
• anonymous seller claims,
• and copied product descriptions.

These sources are useful for discovery but weak as proof.

A marketplace listing may show:

• how an object is being represented,
• how frequently it appears,
• which claims are common,
• or which photographs circulate.

It does not necessarily establish:

• authenticity,
• sale price,
• edition accuracy,
• provenance,
• or historical fact.

Tier Five claims must be corroborated before they are used as authoritative information.

4. Evidence Weighting

Source quality is only one part of the evaluation.

Gauntlet also considers:

• proximity,
• specificity,
• independence,
• contemporaneity,
• documentation,
• consistency,
• and possible bias.

Proximity

How close is the source to the event or object?

An original publisher’s release record is closer than a seller’s description written ten years later.

A photograph of the object at an event is closer than a later verbal claim that it was present.

Specificity

Does the source address this exact object or only the general category?

A certificate tied to a unique serial number is stronger than a generic statement that similar objects were produced.

Independence

Do multiple sources rely on one another, or are they independently confirming the same fact?

Ten websites repeating the same original error do not equal ten independent confirmations.

Contemporaneity

Was the source created at the time of the event?

Contemporaneous sources are often valuable because they predate later market narratives.

An archived 2013 exhibition announcement may be stronger for event timing than a retrospective article published years later.

Documentation

Does the source include photographs, records, serial numbers, invoices, archived pages or direct quotations?

Documented evidence carries more weight than unsupported recollection.

Consistency

Does the source align with known physical evidence and other reliable records?

A stated edition size should be tested against:

• markings,
• publisher records,
• known examples,
• and release history.

Bias

What incentive might the source have?

A seller may benefit from exaggerating rarity.

An auction house may emphasize prestige.

A collector may be emotionally invested in a particular attribution.

An artist’s estate may have commercial or legal reasons for taking a position.

Bias does not invalidate a source, but it should be considered.

5. Historical and Cultural Research

An object becomes more meaningful when its historical context is understood.

Gauntlet’s historical research seeks to answer:

• What was happening when the object was created?
• What was the artist or creator doing at the time?
• How did the object relate to earlier and later work?
• Was it connected to an exhibition, campaign, tour, mission or collaboration?
• How was it received?
• What larger cultural transition did it represent?

Artist Histories

For artists, we research:

• biography,
• training,
• influences,
• major career periods,
• exhibitions,
• collaborations,
• recurring imagery,
• materials,
• political and cultural context,
• studio methods,
• institutional collections,
• and critical reception.

The goal is not to reproduce a generic biography.

The goal is to place the specific object within the artist’s development.

For example:

• Is the work early or late?
• Does it introduce a major motif?
• Does it belong to a recognized series?
• Was it created during an important political period?
• Does it differ from the artist’s usual format?
• Was it associated with a significant exhibition?

Exhibition Research

Exhibition history can establish context and sometimes provenance.

Gauntlet searches for:

• exhibition titles,
• dates,
• venues,
• cities,
• participating artists,
• checklists,
• catalogues,
• installation photographs,
• invitations,
• flyers,
• press coverage,
• interviews,
• and known works shown.

Useful sources include:

• gallery archives,
• artist websites,
• museum records,
• archived event pages,
• local press,
• art publications,
• social-media records,
• and collector photographs.

When an exhibition history cannot be fully verified, we distinguish between:

• confirmed participation,
• reported participation,
• and probable association.

Historical Events

For memorabilia and artifacts, research may extend beyond art history.

Examples include:

• tour dates,
• recording sessions,
• film releases,
• political events,
• mission timelines,
• launch records,
• corporate collaborations,
• product releases,
• and historical anniversaries.

Government records, official archives and contemporaneous reporting are often central.

Archival Web Research

The public internet changes continuously.

Pages are removed.

Retailers close.

Artist websites are redesigned.

Product listings disappear.

Gauntlet uses archived web records when available to recover:

• original product pages,
• release dates,
• pricing,
• dimensions,
• edition information,
• artist statements,
• exhibition announcements,
• and historical images.

Archived pages are especially useful when current secondary sources have copied incomplete information.

However, archived content must still be evaluated for authenticity and date accuracy.

6. Object Identification

Correct identification is the foundation of every later conclusion.

A misidentified object can produce:

• incorrect historical context,
• incorrect pricing,
• incorrect edition claims,
• and incorrect authentication assumptions.

Gauntlet uses a field-based identification process.

Core Identification Fields

Whenever applicable, we document:

• creator or brand,
• title,
• alternate title,
• series,
• year,
• medium,
• materials,
• dimensions,
• edition size,
• edition number,
• proof designation,
• signature,
• inscription,
• publisher,
• printer,
• manufacturer,
• production country,
• model number,
• serial number,
• packaging,
• accessories,
• certificate,
• and visible markings.

Image Comparison

We compare the object with known examples.

This may include:

• front,
• reverse,
• edges,
• labels,
• stamps,
• signatures,
• serial numbers,
• copyright marks,
• packaging,
• and production details.

Image comparison should identify both similarities and differences.

A difference is not automatically evidence of a counterfeit.

Genuine variation can occur because of:

• hand production,
• manufacturing tolerances,
• different production runs,
• regional packaging,
• restoration,
• aging,
• or documented variants.

The significance of the difference must be researched.

Measurements

Dimensions are often essential.

For prints, we may distinguish:

• image dimensions,
• sheet dimensions,
• and framed dimensions.

For figures:

• height,
• width,
• depth,
• weight,
• and scale.

For instruments:

• model,
• body shape,
• finish,
• and manufacturer specifications.

Measurements can help distinguish:

• originals from reproductions,
• editions from posters,
• large and small variants,
• official figures from copied versions,
• and altered objects from complete examples.

Titles and Naming

Object titles are frequently inconsistent.

An artist, gallery, auction house and collector community may use different names.

Gauntlet records:

• the preferred official title,
• known alternate titles,
• common marketplace names,
• and misspellings when useful for search.

However, common usage does not override an established official title.

7. Release and Edition Research

Edition information strongly affects identity and value.

Relevant questions include:

• Was the edition limited, open or timed?
• What was the stated edition size?
• Were artist proofs produced?
• Were printer’s proofs produced?
• Were multiple colorways released?
• Were there regional or retailer exclusives?
• Was the edition signed?
• Were unsigned versions also produced?
• Did the release include certificates?
• Was the object sold individually or as a set?

Official Release Records

We prioritize:

• artist announcements,
• publisher pages,
• gallery pages,
• manufacturer records,
• original retailer pages,
• archived newsletters,
• and contemporaneous social posts from official accounts.

Edition Markings

We inspect:

• handwritten numbers,
• stamped numbers,
• embossed marks,
• labels,
• packaging identifiers,
• and certificates.

An edition marking should be interpreted in context.

For example:

• “AP” may indicate artist proof.
• “PP” may indicate printer’s proof.
• “HC” may indicate hors commerce.
• “EA” may indicate épreuve d’artiste.
• A fraction may indicate the item number and total edition.

However, not every object uses these conventions consistently.

Variants

Variants can materially affect value.

Gauntlet tracks:

• colorways,
• sizes,
• proof types,
• finishes,
• materials,
• packaging variants,
• signed and unsigned editions,
• event editions,
• and regional releases.

A complete release history helps prevent weak comparables.

8. Authentication Research

Authentication research begins by defining the claim.

Signature Authentication

For signatures, we consider:

• recognized third-party opinions,
• database verification,
• letter formation,
• signature placement,
• medium,
• age,
• known signing habits,
• provenance,
• signing photographs,
• and comparison with documented examples.

Signature comparison alone is rarely sufficient for a definitive conclusion.

Autographs can vary across:

• time,
• signing conditions,
• speed,
• health,
• writing surface,
• and context.

For commercially significant signatures, established third-party authentication is generally preferred.

Print and Artwork Authentication

For artist works, we consider:

• edition records,
• paper,
• image size,
• sheet size,
• printing method,
• signature,
• numbering,
• blind stamps,
• publisher marks,
• release history,
• provenance,
• and comparison with known examples.

Manufacturer Authentication

For designer objects, we examine:

• mold,
• materials,
• finish,
• weight,
• articulation,
• markings,
• packaging,
• accessories,
• and release records.

Artifact Authentication

For artifacts, we examine:

• chain of custody,
• contemporaneous documentation,
• mission or event records,
• serial numbers,
• material analysis,
• specialist opinion,
• and photographic matching.

Authentication Databases

Where available, certification numbers should be checked directly with the issuing organization.

The verified record should be compared with the item:

• category,
• signer,
• description,
• image,
• and certificate details.

A valid number attached to the wrong object is not valid authentication.

9. Provenance Research

Provenance is strongest when it forms a continuous documented chain.

However, the standard required depends on the object.

A modern open-edition figure may not require a complete ownership history.

A unique artwork or historically significant artifact may.

Types of Provenance Evidence

Relevant evidence includes:

• artist or studio invoice,
• publisher invoice,
• gallery invoice,
• auction invoice,
• collection label,
• shipping record,
• correspondence,
• photographs,
• estate documentation,
• authentication record,
• prior catalogue entry,
• and transfer agreement.

Ownership Statements

A previous owner’s written statement may be useful, especially when supported by photographs or contemporaneous records.

However, an ownership statement is not equivalent to independent proof.

Gaps in Provenance

A gap does not automatically make an object inauthentic.

Many legitimate objects lack complete records.

The correct approach is to disclose the gap and avoid claims that require a stronger chain than the evidence supports.

10. Condition Research

Condition evaluation combines physical inspection with category-specific knowledge.

Inspection Methods

Depending on the object, Gauntlet may use:

• normal light,
• angled light,
• magnification,
• front and reverse inspection,
• measurement,
• high-resolution photography,
• packaging inspection,
• and comparison with production norms.

Condition History

When possible, we preserve:

• prior condition reports,
• restoration records,
• framing history,
• and historical photographs.

These can help identify later changes.

Material Knowledge

Different materials fail differently.

Paper may crease, fox, fade or ripple.

Painted vinyl may scuff or transfer.

Metal may oxidize.

Signatures may fade or smear.

Packaging may yellow, crush or delaminate.

Understanding the material helps distinguish:

• production characteristics,
• natural aging,
• handling damage,
• and restoration.

11. Market Research

Market research seeks to identify what buyers have actually paid for sufficiently comparable objects.

It does not seek to justify a predetermined asking price.

Primary Market Data

Primary market information may include:

• original retail price,
• artist release price,
• publisher price,
• gallery price,
• and initial distribution method.

Primary-market price is historically useful but may not reflect current value.

Secondary Market Data

Secondary market data includes:

• completed auction sales,
• verified marketplace sales,
• dealer transactions,
• direct sales,
• and private-sale evidence where available.

Asking Prices

Current asking prices may help measure:

• available supply,
• seller expectations,
• and market positioning.

They are not treated as completed-sale evidence.

Unsold Lots

Unsold lots are informative.

They may show that the market rejected a particular reserve or estimate.

When available, buy-in data can provide important context.

Hidden Best Offers

Some platforms record a sale but conceal the accepted amount.

These transactions should not be treated as if the original asking price was paid.

Where the true price is unknown, the record should be labeled accordingly or excluded from precise calculations.

12. Comparable-Sale Selection

A sale is only useful when its relationship to the subject object is understood.

Gauntlet evaluates comparability across:

• creator,
• title,
• edition,
• size,
• colorway,
• date,
• signature,
• proof type,
• condition,
• framing,
• authentication,
• provenance,
• venue,
• sale date,
• and fees.

Exact Comparables

Exact comparables match the same work or object variant closely.

These generally receive the highest weight.

Near Comparables

Near comparables may differ in:

• colorway,
• proof status,
• framing,
• or condition.

Adjustments may be necessary.

Contextual Comparables

Contextual comparables may include:

• related works from the same series,
• similar edition sizes,
• or comparable objects by the same creator.

These are useful when exact sales are scarce, but they should not be presented as equivalent.

13. Cleaning Market Data

Raw market data frequently contains errors.

Gauntlet applies cleaning procedures before drawing conclusions.

Duplicate Removal

The same sale may appear across:

• auction databases,
• syndication sites,
• marketplace archives,
• and dealer records.

Duplicate records should not be counted multiple times.

Currency Normalization

Historical prices may be converted into a common currency.

The method and date of conversion should be documented when material.

Fee Treatment

Auction results may report:

• hammer price,
• price including buyer’s premium,
• or total cost.

These should not be mixed without adjustment.

Incorrect Identification

Listings may misidentify:

• title,
• edition,
• size,
• colorway,
• or artist.

Images and descriptions should be reviewed before including the sale.

Outlier Review

Extreme values may be legitimate.

They may also result from:

• bundled lots,
• incorrect records,
• unusual provenance,
• charity sales,
• celebrity ownership,
• heavy damage,
• or market manipulation.

Outliers should be investigated rather than automatically removed.

14. Statistical Interpretation

Statistics summarize data.

They do not eliminate the need for judgment.

Gauntlet may use:

• minimum,
• maximum,
• average,
• median,
• mode,
• lower quartile,
• upper quartile,
• standard deviation,
• transaction count,
• annual sale count,
• time between sales,
• and historical growth rate.

Median

The median is often useful when a small number of extreme sales distort the average.

Average

The average may be useful when the distribution is relatively balanced.

Range

The range reveals dispersion but may overemphasize unusual extremes.

Quartiles

Quartiles help show the central trading band.

Frequency

Sale frequency helps evaluate liquidity.

An object with a high estimated value but one transaction every several years may be difficult to sell.

Time Period

Recent sales often deserve more weight, but older records can reveal long-term patterns.

The appropriate period depends on the object and market.

15. Confidence Levels

Gauntlet assigns an internal confidence level to important conclusions.

High Confidence

High-confidence conclusions are supported by:

• direct authoritative records,
• multiple independent confirmations,
• consistent physical evidence,
• and minimal contradiction.

Moderate Confidence

Moderate-confidence conclusions are supported by credible evidence but contain:

• incomplete records,
• limited comparables,
• minor inconsistencies,
• or reliance on strong secondary sources.

Low Confidence

Low-confidence conclusions may rely on:

• limited evidence,
• collector reports,
• uncertain identification,
• incomplete provenance,
• hidden sale prices,
• or unresolved contradictions.

Low confidence does not mean false.

It means the conclusion should be presented cautiously.

16. Resolving Conflicting Information

Conflicts are common.

Different sources may report different:

• dates,
• edition sizes,
• dimensions,
• titles,
• release prices,
• and historical facts.

Gauntlet resolves conflicts through a structured process.

Step One: Identify the Exact Conflict

The conflicting claims are stated clearly.

Step Two: Trace Each Claim to Its Earliest Available Source

Later repetition may originate from one initial error.

Step Three: Evaluate Source Authority

Primary records generally outweigh later secondary descriptions.

Step Four: Compare Physical Evidence

The object itself may help resolve dimensions, markings or edition details.

Step Five: Preserve the Conflict When Necessary

When the evidence remains inconclusive, both positions may be recorded.

Example:

Official release material lists the edition as 200, while several later marketplace records describe an edition of 250. Gauntlet has not located sufficient primary evidence to reconcile the discrepancy.

That is more responsible than selecting the more favorable number.

17. Research Documentation

Research should be reproducible.

Gauntlet preserves source information so important conclusions can be reviewed later.

Source Records

Where practical, we record:

• source title,
• publisher,
• author,
• date,
• URL,
• access date,
• archive link,
• page number,
• quotation,
• image,
• and relevance.

Screenshots and Archived Copies

Important web pages may be preserved through:

• screenshots,
• PDF exports,
• archive links,
• and saved metadata.

This helps protect against link rot.

Object-Level Research Files

Significant objects may have a dedicated research file containing:

• identification,
• images,
• measurements,
• authentication,
• provenance,
• condition,
• historical context,
• comparables,
• sources,
• and unresolved questions.

18. Corrections and Versioning

Research evolves.

Gauntlet supports correction rather than defending outdated claims.

Correction Triggers

A record may be revised when:

• new primary evidence appears,
• an archive is updated,
• authentication changes,
• a market record is corrected,
• an identification error is discovered,
• or condition changes.

Version History

Important revisions should preserve:

• prior statement,
• revised statement,
• date,
• reason,
• and supporting evidence.

This protects institutional memory.

Material Buyer Impact

If a correction materially affects an object already sold, Gauntlet should determine whether the buyer should be notified.

The response depends on:

• severity,
• evidence,
• transaction terms,
• and applicable policies.

19. AI-Assisted Research

Artificial intelligence can improve research efficiency.

It can help:

• organize large datasets,
• identify duplicate records,
• summarize source material,
• generate search variations,
• classify objects,
• flag inconsistencies,
• and surface possible connections.

However, AI-generated output is not treated as a primary source.

Appropriate Uses

AI may support:

• discovery,
• categorization,
• comparison,
• drafting,
• and pattern detection.

Inappropriate Uses

AI should not independently establish:

• authenticity,
• provenance,
• historical fact,
• or market value

without underlying evidence.

Verification

Material AI-assisted findings must be verified against reliable sources.

The responsibility remains with Gauntlet.

20. Research Across Gauntlet Categories

The methodology is adapted to each category.

Street and Contemporary Art

Priority sources include:

• artist archives,
• publishers,
• printers,
• galleries,
• exhibition catalogues,
• museum collections,
• release announcements,
• and completed sales.

Designer Collectibles

Priority sources include:

• official manufacturers,
• collaboration partners,
• original retailers,
• release calendars,
• packaging records,
• collector archives,
• and marketplace sales.

Signed Music Memorabilia

Priority sources include:

• recognized autograph authentication,
• signing provenance,
• official merchandise records,
• tour context,
• instrument identification,
• and comparable authenticated sales.

Space Memorabilia

Priority sources include:

• NASA,
• mission records,
• astronaut estates,
• specialist authentication,
• respected auction archives,
• contractor records,
• and documented provenance.

21. What We Do Not Claim

Gauntlet research is extensive, but it has limits.

We do not claim:

• perfect access to all private sales,
• complete knowledge of every edition,
• infallible authentication,
• exact future value,
• or universal agreement among experts.

We do not present estimates as guarantees.

We do not treat a large dataset as automatically accurate.

We do not use weak evidence simply because stronger evidence is unavailable.

We do not convert probability into certainty for commercial convenience.

22. The Research Standard in Practice

A typical Gauntlet research workflow may include the following:

Phase One: Intake

• photograph the object,
• record measurements,
• identify markings,
• review documentation,
• and define central claims.

Phase Two: Primary Research

• locate official records,
• verify certification,
• search artist or manufacturer archives,
• and inspect release information.

Phase Three: Secondary Research

• review exhibition records,
• publications,
• specialist references,
• and auction archives.

Phase Four: Object Comparison

• compare known examples,
• identify differences,
• inspect condition,
• and reconcile specifications.

Phase Five: Market Analysis

• collect comparables,
• remove duplicates,
• normalize fees and currencies,
• classify sale quality,
• and calculate relevant statistics.

Phase Six: Interpretation

• assign confidence,
• document unresolved issues,
• determine pricing range,
• and prepare the description.

Phase Seven: Publication

• present the object,
• expose material evidence,
• disclose limitations,
• cite relevant sources,
• and preserve the research file.

23. The Institutional Value of Historical Data

Historical data creates more than pricing support.

It creates memory.

Without structured preservation:

• old product pages disappear,
• edition details become confused,
• variant names change,
• press coverage becomes difficult to find,
• images detach from titles,
• and seller folklore replaces the original record.

Gauntlet’s historical datasets, artist indexes and research pages are intended to reduce that loss.

A complete record may eventually include:

• release chronology,
• visual archive,
• dimensions,
• materials,
• edition sizes,
• known variants,
• original pricing,
• sales history,
• exhibition references,
• and current market observations.

The value of such a record extends beyond current inventory.

It can support:

• collectors,
• researchers,
• journalists,
• estates,
• insurers,
• auction houses,
• and future authentication work.

This is why Gauntlet views data as cultural infrastructure.

24. Research Ethics

Research can create power.

A party with more information may be able to buy cheaply, sell aggressively or exploit uncertainty.

Gauntlet’s methodology is intended to reduce information asymmetry, not deepen it.

No Selective Evidence

We should not include only evidence that supports a higher price.

Contradictory sales, weak demand and condition risks matter.

No Manufactured Authority

A long report does not become authoritative merely because it is detailed.

Sources and reasoning must support the conclusion.

Respect for Privacy

Private collector information should be protected unless disclosure is authorized or necessary.

Respect for Intellectual Property

Research should summarize and cite sources responsibly.

It should not reproduce protected material unnecessarily.

Respect for Uncertainty

Uncertainty is not a defect to be hidden.

It is part of responsible scholarship.

25. The Gauntlet Research Principles

The methodology can be summarized in the following principles.

1. Define the Question Before Searching

Research without a precise question produces noise.

2. Match the Source to the Claim

Authority in one area does not create authority in every area.

3. Prefer Primary Evidence

Direct records generally outweigh repeated secondary claims.

4. Use Multiple Independent Sources

Agreement is stronger when sources do not depend on one another.

5. Inspect the Object

Digital records cannot replace physical evidence.

6. Preserve Contradictions

Unresolved conflicts should remain visible.

7. Separate Asking Prices From Completed Sales

Seller expectations are not market proof.

8. Clean Data Before Interpreting It

Duplicates, errors and hidden prices distort conclusions.

9. Use Statistics as Tools, Not Answers

Judgment remains necessary.

10. Assign Confidence

Not every conclusion deserves the same certainty.

11. Document Sources

Research should be reviewable and reproducible.

12. Correct the Record

New evidence should improve the conclusion.

13. Use AI as an Assistant

AI can accelerate research but does not replace verification.

14. Protect the Collector

The research burden should not be transferred to the buyer.

15. Preserve History

The record may matter long after the current sale.

Conclusion

Gauntlet Gallery’s research methodology is built around a simple belief:

A collectible object deserves a documentary record equal to the seriousness of the claim being made about it.

If an object is described as authentic, the evidence should be examined.

If it is described as rare, the scarcity should be explained.

If it is presented as historically important, the history should be documented.

If it is priced at a premium, the market basis should be visible.

If uncertainty remains, the uncertainty should be stated.

Research does not remove every risk from collecting.

It does improve the quality of the decision.

It helps distinguish:

• fact from repetition,
• evidence from assertion,
• completed sales from asking prices,
• meaningful scarcity from manufactured urgency,
• strong provenance from convenient storytelling,
• and cultural importance from temporary attention.

Gauntlet’s objective is not to produce the longest description.

It is to produce the strongest available record.

That record combines:

• the object,
• the source,
• the history,
• the evidence,
• the market,
• and the limits of what can responsibly be claimed.

When those elements are assembled carefully, the collector receives more than an item.

They receive an object with context, a price with reasoning and a history that is more likely to survive the next transfer.

That is the Gauntlet Research Methodology.

Editorial image disclosure: Illustrations on this page were generated with Nano Banana for explanatory use. They are not authentication evidence, provenance records, or reproductions of any specific collectible.