Appeals & Corrections

TL;DR

If RINGER has flagged an account you own or represent, email appeals@ringer.report. We review every good-faith appeal within 5 business days, re-examine the underlying evidence, and respond with one of three outcomes — correction, annotation, or removal — with corrections published as prominently as the original claim, annotations attached visibly to the report, and removals logged publicly. We keep a public corrections log. We will never demand you prove your real-world identity to be heard. We also will not remove truthful analysis solely because its subject objects.

Who can appeal

Anyone who owns or represents a flagged account — the account holder, their counsel, or an authorized representative. You do not need to tell us who you are in real life (see below). You do not need to be the subject at all: if you are a researcher, journalist, or reader who believes a specific published claim is factually wrong, the same address and the same standards apply.

How to appeal

Email appeals@ringer.report with:

1. Which report and which claim. Link the report; quote or precisely identify the specific finding you dispute. “The whole report is wrong” is not reviewable; “the funding transfer you attribute to address X on date Y did not occur / is misread” is.
2. Why it’s wrong. The most effective appeals point at evidence — a transaction hash, a settlement record, a timing fact we misread, a wallet we mis-resolved. Everything we publish is built from public data, so the strongest rebuttals are also public data.
3. If ownership matters to the appeal (for example, “these five wallets are not commonly operated”), a signed message from the wallet in question is sufficient proof of control. That’s it — a signature, not a name.

Identity is never the price of admission

We never demand identity disclosure to process an appeal. RINGER names wallets, not people; it would be incoherent — and wrong — to require a human being to unmask themselves to dispute a claim about an address. A cryptographic signature proves control of an address when control is the point in dispute. An anonymous appeal that points at a real error gets the error fixed exactly as fast as a signed one.

What we re-examine

An appeal triggers a re-review of the disputed claim against the original evidence archive — the hash-stamped activity tape, funding transfers, and on-chain settlement reads behind the report — plus any new evidence the appeal supplies. Specifically, we re-check:

• The raw data: were the transactions, fills, and settlements read correctly from chain and tape?
• The reconstruction: did our known coverage gaps (incomplete public activity history, unobserved sells and REDEEM/MERGE exits, unresolved markets) distort this specific claim?
• The statistics: does the finding still clear its gates and nulls when the disputed input is corrected or excluded?
• The language: even if the numbers hold, did we state anything more strongly than the evidence supports?

The re-review is done by a human, against the same standards the original CRITICAL review used — and with the same posture: the nulls exist to kill our findings, and an appeal is one more chance for them to do their job.

The three outcomes

Within 5 business days of receipt, we respond with one of:

• Correction — the claim was wrong. We fix it in the report, state plainly what was wrong and why, and publish the correction as prominently as the original claim — not buried in a footnote. If the correction removes the basis for the flag, the flag comes down with it.
• Annotation — the claim stands, but the dispute adds material context (an alternative explanation consistent with the data, a coverage caveat we should have named, a subject’s on-record response). We attach the annotation to the report, visibly.
• Removal — the finding no longer clears our own bar once the appeal’s evidence is in. We take it down and say so publicly. A removed report is logged, not memory-holed.

If a review genuinely needs more than 5 business days (for example, an on-chain re-verification job across a large wallet set), we say so within the 5 days, with a date.

What an appeal cannot do: we do not remove truthful analysis solely because its subject objects. If the data is public, the reading is accurate, and the language stays within what the evidence supports, the report stands — with your objection annotated if you want it on the record.

The corrections log

Every correction, annotation, and removal is recorded in a public, permanent corrections log — what changed, when, and why. We treat this log as a feature, not a liability: a detection service with no visible corrections is either perfect or hiding something, and we are not perfect. Our reports already publish our blind spots and the places our numbers came out smaller than the public headline; the corrections log is the same honesty applied over time.

Good faith, both directions

We owe you: a real human review, on the clock, against the evidence, with the outcome published. You owe the process: a specific claim, disputed on substance. We reserve the right to consolidate duplicate appeals and to decline re-review of a claim already reviewed on identical grounds with no new evidence — but a new fact always reopens the file.

Full detection methodology and limits: Methodology & Limits.

RINGER analyzes public on-chain and prediction-market data. Informational only — not trading advice. Findings are probabilistic statements about accounts, never people — see the methodology.