Evidence & Verification Policy

GachaMeister’s Five Public Evidence States

GachaMeister uses five public evidence states when reviewing or rating platforms and services, each of which is explained below:

Observed

Observed means we have direct first-party evidence of the specific event, process or result being described. This could include a documented purchase, pack opening, withdrawal, shipment, redemption, customer support interaction, or other test as carried out as part of our own research/review process.

An Observed label only applies to what the evidence actually establishes, so, for example, if one card was successfully delivered to us, we can say that delivery was completed in that case. It does not mean we can use one successful shipment to claim that a platform always delivers successfully, or that every customer can expect the same experience.

Official-Only

Official-only means the information is supported by a current source that comes first-hand from the relevant company, platform, vendor, grading company, regulator or other responsible organisation. Crucially, it means that it has not been independently established through our own testing.

This could, for example, include things like published fees, terms, service availability, withdrawal rules, shipping policies, grading tiers, or stated pack odds. While official information is often the best evidence for what a company currently says its rules and terms are, it does not automatically mean that everything claimed is how it works in reality.

Quoted

Quoted means that a named source has provided us with a statement that we are reporting as their statement. Where relevant, we display who made the statement, when it was made, and the context in which it was provided to GachaMeister. A quoted statement does not automatically verify that everything contained within the statement is an independently verified fact, however.

Community-Reported

Community-reported means the information comes from collector reports, customer reviews, community discussions or similar third-party accounts and has not been independently established by GachaMeister as a general fact.

Community evidence can be extremely useful for identifying potential issues with a platform or service that may deserve further investigation by the GachaMeister team.

However, an individual review or Reddit post, for example, can’t, on its own, be a reliable metric of how a platform/service normally operates. Even when we see multiple reports, the context surrounding them needs to be considered, including things like their age, independence, specificity, and whether the underlying claim can be corroborated anywhere else.

Unverified

Unverified means GachaMeister is aware of a claim, question or possible issue, but we do not currently have enough evidence to give it a stronger evidence status.

Unverified doesn’t mean that something is false, nor does it mean something is true. It means exactly what it says: the evidence available to us isn’t enough for us to establish the claim to meet the required standard, so it remains Unverified.

How We Choose the Right Evidence

GachaMeister believes that no single source is automatically best for every question, and for us, the right evidence depends on what we’re actually trying to establish.

As a general rule of thumb:

• For laws, regulations and legal status, we look for legislation, regulators, courts and other appropriate legal authorities.
• For platform, vendor or grading-company rules, we use current official sources for things like fees, terms, stated odds, shipping policies and withdrawal rules.
• For what happened during GachaMeister testing, we use our own first-party testing evidence.
• For independent corroboration and discovering possible issues, we may use reputable reporting and carefully assessed community evidence.

Search engines and AI tools can also help us to find information and locate possible sources, but they are not treated as reliable evidence sources on their own. For us, what’s most important is that any source used can support the particular claim being made.

What We Mean by “Observed”

When we describe something as Observed, it means that our conclusion has been reached due to evidence we experienced first-hand during our testing.

Depending on what we’re testing, that evidence might include receipts, transaction records, screenshots, timestamps, support correspondence, withdrawal records, tracking information, photographs or verifiable blockchain transactions.

We always do our best to preserve appropriate evidence and enough context to show what happened, when it happened, and which platform, account, product, or transaction was involved. We don’t rely on somebody’s memory alone for an important first-party claim and, where suitable, we may publish a redacted version of the supporting evidence for GachaMeister readers.

Evidence and Testing Aren’t the Same Thing

An evidence label tells you what supports a claim, whereas a testing state tells you whether we tested the relevant feature or process.

For example, a withdrawal policy might be Official-Only because we checked the platform’s published rules. In this case, the withdrawal process might separately be Not Tested, simply because we haven’t attempted a withdrawal ourselves.

If you want to know more about Tested/Not Tested/Not Testable/N/A/Unknown labels, including how those states affect GachaMeister scoring and review conclusions, check out our Review and Evidence Methodology guide.

When the Evidence Doesn’t Give a Clear Answer

Sometimes sources conflict with one another, important information is missing, or good evidence has become too old to support a current claim. Where this is the case, we don’t hide those limitations, or fill in the gaps ourselves.

• When Sources Don’t Agree

A platform’s help centre might say one thing, while its terms and conditions page says something completely different. A member of the customer support team might contradict writing on the public-facing website, or an old policy page may remain accessible long after the rules have changed.

These are, of course, all just examples - but situations like these do happen, and when a disagreement matters, GachaMeister looks at which source is responsible for the claim, which is most current - and whether the difference/dispute could be explained by location, account, type, product, or date. We may also ask the platform/company/seller/organisation to clarify the discrepancy.

If we can’t resolve a conflict of information, we make that uncertainty visible. Depending on the situation, that could mean leaving the claim Unverified, leaving the relevant review part Unknown, withholding a conclusion, or explaining the conflicting information directly.

• When We Don’t Know, We Say So

When we can’t find enough evidence, we don’t automatically issue a rating of zero, nor say that something is automatically false, fraudulent, unsafe, or illegal. A lack of information or evidence also isn’t treated as proof that something didn’t happen.

If a material claim depends on evidence we don’t have, the claim, rating or other GachaMeister output may need to be limited or withheld. In a scored review, for example, we don’t invent a category score when the evidence needed to assess that category is unavailable.

Sometimes, Unknown is the most accurate - and the fairest - answer we can give.

• What Old Evidence Can - and Can’t - Tell Us

Platforms change their fees, terms, shipping policies, withdrawal methods and service availability all the time, so evidence that accurately shows what happened in the past may no longer establish what is true today.

A six-month-old screenshot might be excellent evidence of what a platform displayed six months ago, but terrible evidence of its current policy. Similarly, a completed GachaMeister test remains a historical record of what happened on that date; we don’t rewrite that result simply because the platform later changes its process.

Where current information matters, GachaMeister checks freshness separately. This doesn’t mean that small cosmetic changes to a website mean the underlying factual research needs to be done again, but we do monitor changes, and, where necessary, re-test and re-investigate.

Bigger Claims Require Stronger Evidence

Claims involving fraud, scams, rigging, theft, illegality or other pretty serious forms of misconduct need particular care, because the consequences of us getting them wrong are much greater.

To give you an example: a customer complaint may give us a good reason to investigate a platform or service, but on its own, it does not establish that an operator committed fraud or acted in bad faith.

Where our evidence supports a specific factual description - but not the wider allegation - we publish the factual description. In very serious cases, we may also ask the organisation or person involved to respond before publication, as part of our Right of Reply policy.

If New/Better Evidence Appears

Evidence doesn’t mean that our verdict is “frozen” forever. New information may appear that helps to resolve an uncertain claim, prove that an old source is no longer current, or even show that something we published was incorrect.

Something that was accurate at the time of publication could later change - and when this is the case, an update is usually required. If we publish a statement as fact that was wrong when published, then a correction may be required. If the evidence supporting a claim becomes insufficient, then that claim may also need to be changed, or removed.

Our Corrections & Right of Reply Policy explains how we handle those situations, and what you can do if you feel we got something wrong.

What GachaMeister Publishes - and What We Keep Private

Sometimes, useful evidence may contain personal information, addresses, account identifiers, payment details or other sensitive material that readers don’t need in order to understand our finding.

When we publish evidence, we may, where suitable, redact sensitive information. We preserve the underlying original separately so that removing private details from the public version doesn’t change what the evidence actually means.

We will never intentionally publish passwords, recovery codes, authentication tokens, full identity documents, complete payment-card details or similar credentials simply to prove that a test took place.

Similarly, in cases where we’re dealing with a complaint from a reader and they supply us with documents or evidence, this doesn’t automatically give us permission to reproduce or automatically collect it, let alone publish it.

The same principle also applies, more broadly, to data and information we find elsewhere. Just because we’re technically able to access something does not automatically give GachaMeister the permission to reproduce, republish, or automatically collect it. If, at any point, the relevant rights are unclear, restricted, or withdrawn, we don’t use the affected material until we have a suitable permitted way to proceed.

GachaMeister’s Commercial Relationship Standards

An affiliate relationship, test access, or another commercial connection does not change the evidence threshold that’s required to support a claim.

Just because we have an affiliate relationship with a platform, for example, does not mean it can turn an Official-Only claim into an Observed one, make weak evidence sufficient, remove an inconvenient result, increase a score, or suppress a supported negative finding.

For more information, please see our Editorial Independence Policy and Affiliate Disclosure which explain these protections in more detail.

Evidence Rules for Research Reviews

A Research Review may be published before GachaMeister has used the platform when the page is substantial, current, sourced and prominently labelled Not Yet Hands-On Tested. Current official terms, policies, fees, supported-country information and published odds may be stated as official facts; marketing or performance claims remain attributed to the operator unless independently established. Community experiences remain Community-Reported and are not converted into general facts by repetition alone.

GachaMeister may give editorial analysis based on documented facts and clearly identified uncertainties. Until first-party testing supports the relevant claim, the page must not imply a deposit, purchase, opening, buyback, redemption, shipment, support interaction or withdrawal occurred; must not claim operational reliability; and must not present a recommendation as hands-on tested. Research-based scores and recommendations are allowed when supported by current evidence and the applicable scoring rules; any rating markup must match the visible assessment and pass technical eligibility checks. Serious adverse claims require evidence proportionate to the allegation, precise factual wording and, where appropriate, a right-of-reply step. Later testing updates the same canonical URL, shows the test scope and date, preserves n=1 limits and records material corrections or changes.