Trust document · Standards doctrine · v1.0
Editorial independence
ChainBankroll editorial is independent of operator marketing teams, affiliate networks, and pay-for-placement arrangements. Independence is about rubric integrity, not revenue source. We are a publication that monetizes through affiliate links; we are not a marketing arm for any operator.
What independence IS
What independence is NOT
Cross-reference: affiliate disclosure + how we test §How We Make Money.
Scoring oversight
Operator scores are derived by Quinn Hartwell using the 8-dimension rubric. Scores are re-derived every 90 days minimum. The math is unweighted-mean. The rubric is publicly verifiable. Per-operator score = simple unweighted mean of 8 dimension scores: UX, Withdrawals, KYC Friendliness, Game Selection, Sportsbook, VIP, Transparency, Support.
Why unweighted-mean. Weighting schemes are where bias hides. If the methodology hands you a chart where "trust" is weighted 0.18 and "withdrawals" 0.07, someone made that choice with priors. Equal weight is verifiable. The transparency of the math is the credibility signal.
Re-test triggers. Every operator gets re-scored on a 90-day calendar minimum. Sooner when one of these material shifts occurs:
Revision logging. Score revisions are logged with a dated note on the review page. Material revisions also append to launch baseline 2026 change log if the operator is part of the baseline. We do not ghost-edit scores.
Cross-reference: how we test §Scoring Rubric + launch baseline 2026.
Fact-checking
Every quantitative claim sources to one of three categories: ChainBankroll testing dataset, operator-official, or regulator-official. Three tiers, documented citation per claim type.
Withdrawal timing, KYC trigger behavior, support response, deposit confirmation latency. Cited inline with section reference. Underlying CSV downloadable under CC-BY-4.0 license.
Example: "Stake routine withdrawal median 3 min" → launch baseline §3 + withdrawal benchmarks
Operator T&C, footer license number, /wallet pages, /vip pages. Cited with date stamp. Re-verified per re-test cycle because operator T&C changes between cycles.
Example: "Stake VIP rakeback starts Bronze 5%" → help.stake.com canonical extract 2026-06-05
UKGC enforcement register, ATO position statements, CRA gambling-winnings guidance, NZ DIA, Curaçao GCB register. All external regulator links verified via pre-publish external-link audit.
Example: "Stake exited UK on 11 March 2025 post-UKGC investigation" → UKGC enforcement register entry, verified 2026-06-03
Circular sourcing risk. If one site repeats another which repeats another, the chain has no anchor in primary fact.
Cited as sentiment signal where useful; never as fact. A Reddit thread shows what users believe, not what is.
Without a specific operator example tied to a verifiable source. "Industry standard" is the affiliate-page filler we refuse to ship.
The voice-guide §6 honesty marker: when source is partial, we publish the limitation. The phrase "We can't verify X without Y" appears across the site as the editorial-honesty marker for partial coverage. It is not optional.
Voice ratio
The voice ratio is part of the editorial standard, not a stylistic accident. Different page types carry different sentence-count balances between "we" (process and protocol) and "I" (editorial judgment). The ratios are documented because the consistency is verifiable.
Reviews
Higher "I" because reviews carry a Bourdain "What I didn't expect" section. "We tested. I scored."
Methodology / doctrine
/methodology/how-we-test/, this page, /affiliate-disclosure/
Higher "we" because the content describes process. Doctrine docs use editorial-team voice; named-author "I" reserved for judgment.
About / authors
/about/, /authors/quinn-hartwell/, /authors/mark-rogers/
Higher "I" because the page is personal accountability and bias-declaration. Author voice belongs to the named entity.
What "we" means. ChainBankroll editorial entity + its documented standards + the testing protocol it applies. Every "we tested" sentence means the protocol ran against the documented spec at how we test.
What "I" means. Named-author editorial judgment - when a number from the test gets translated into a score, when a dimension scores below the operator's marketing claim, when a Bourdain "What I didn't expect" admission lands. "I" is byline-attributed: "I scored - QH" on Quinn's reviews, "I scoped - MR" on Mark's cryptographic claims.
Why ratios differ by page type. Reviews carry the Bourdain section by design. Methodology is doctrine. Author pages are personal accountability. The ratio reflects what the page is for, not stylistic preference.
Conflicts of interest
The standard conflict-of-interest declaration: affiliate relationships are disclosed inline on every page that contains affiliate links. Full policy at affiliate disclosure. The list below documents what we do NOT carry as a conflict.
What we DO carry
What we do NOT have
If any of the "do not have" items changes, this page updates within 7 days with a dated note in §Version log + the operator-side mention is added to the relevant author profile.
Corrections
Errors get corrected within 7 days of confirmation. Corrections are dated and appended visibly. We don't ghost-edit published claims. Reader-reported errors run through the four-step process below.
Subject "Correction:" + URL + the specific claim you're flagging. editorial inbox. Be specific about what is wrong - a vague "the score is too high" doesn't trigger review; "the cited withdrawal median contradicts the source CSV" does.
Cross-check against operator-official source, testing log, regulator source as applicable. Verification typically takes 1-3 business days. We confirm or rebut with evidence, not opinion.
If error confirmed: page updated with corrected claim + dated correction note appended at the bottom ("Correction [date]: [original claim] was incorrect. Corrected to: [new claim]. Source: [verifiable link]"). Material corrections carry an inline dateline at the top of the affected section.
From the day we confirm the issue. Faster when the underlying data is already public; slower when third-party verification is required. Reporter receives a reply with the diff link or the rebuttal.
What gets corrected without note: typos, broken internal links, image alt-text fixes. These are housekeeping, not editorial reversals.
What does NOT get corrected: scoring decisions readers disagree with. Those remain editorial judgment per the published rubric. Disagreements about whether a given operator deserves a higher score can be filed but won't trigger a revision unless the underlying rubric-data is incorrect. The math is independent; the decision applies the math.
Re-test cadence
Ranking decay without freshness signals is the affiliate-listicle trap. Documented cadence is the alternative. Different content types have different refresh frequencies.
Live testing datasets: withdrawal benchmarks, support response times, KYC severity index. Rolling 30-day window. Refreshed on the 1st of each month.
Operator scores (in /reviews/): re-derived every 90 days minimum. Sooner on material operator-behavior shifts.
Research publications: Curaçao license tracker re-audited every 90 days. Launch baseline continues via testing trinity. Methodology + editorial policy reviewed quarterly with version log entries.
Off-cycle updates: license change, regulator action, public dispute escalation, framework refresh, or major operator-behavior shift. Material-event re-tests publish within 7 days of confirmed trigger.
Feedback channels
Three channels: editorial inbox (corrections + tips), social handles (informal feedback), and Telegram (link-drops + short questions). We do not run a public comments section under reviews - moderation overhead plus Astroturf-vulnerability. Disagreements go through editorial channels.
Editorial inbox
Corrections, tips, partnership inquiries · /contact/
X / Twitter
Informal feedback · @ChainBankroll + @QuinnHartwell
Telegram
Short questions · link-drops · t.me/chainbankroll
We surface valid critique in revision notes when it triggers a fix. We do not adopt critique that disagrees with rubric application but does not contest the underlying data.
FAQ
Errors get corrected within 7 days of confirmation. Corrections are dated and appended visibly. We don't ghost-edit published claims.
Reader reports the error via /contact/ with subject "Correction:" + URL + specific claim. We verify, update, and append a dated correction note. Material corrections carry an inline dateline at the top of the affected section.
Full process documented at §Corrections above.
Live withdrawal data refreshes monthly via the testing-cluster trinity at /testing/.
Operator scores in reviews are re-derived every 90 days minimum, sooner on material operator-behavior shifts (new license, KYC-threshold change, withdrawal-policy change, public dispute, license loss).
Methodology and editorial policy are reviewed quarterly. Cadence summary at §Re-Test Schedule above.
No. The 8-dimension scoring rubric is rubric-defined and operator-payment-independent. Affiliate revenue does not alter dimension scores.
We have published low scores on operators we hold affiliate relationships with. Mechanism documented at affiliate disclosure §How It Doesn't Influence Scoring.
Three-tier hierarchy:
(1) ChainBankroll testing dataset primary for behavioral claims;
(2) operator-official source primary for terms, licenses, jurisdictions;
(3) regulator-official source primary for legality and jurisdiction questions.
We do not accept other affiliate review sites as primary source - circular sourcing risk. Forum threads cited as sentiment signal, never as fact.
Revision history
Each material revision documented with date and change summary. Subsequent versions appended as policy revises quarterly.
Published 8 Jun 2026 · current
Initial editorial policy. Covers independence (rubric + low-scores-on-partners), scoring oversight (unweighted-mean + 90-day re-test calendar), fact-checking three-tier hierarchy, voice ratio (70/30 reviews, 85/15 doctrine, 50/50 about/authors), conflict-of-interest declarations, four-step corrections process, cadence schedule, three reader-feedback channels. Quinn Hartwell author byline.
Standards & Compliance
How we work - standards
Editorial policy document v1.0. Plain-English documentation, not legalese. Independence is rubric integrity, not zero revenue. Material conflicts disclosed in affiliate disclosure. Reader-reported errors corrected within 7 days with a dated correction note.