This checklist evaluates verifiability, not profitability. A provider can pass all seven checks and still lose money ahead; honesty and skill are different tests, and this page only measures the first. We apply the same list to ourselves at the end, including the checks we cannot yet pass. Nothing here is investment advice. Decisions are yours.
You do not need a statistics degree to filter crypto signal providers. You need seven questions, each answerable in minutes from public information, each with a clean pass or fail. The list is ruthless by design: in our survey of this market, almost nothing passes more than two, and the industry's own review sites concede that most paid groups are outright scams. Fifteen minutes with this page is worth more than any subscription it disqualifies.
Check 1: can the history be edited?
Where does the track record physically live? A Telegram feed, a Discord server, a website database, all editable or deletable by the owner at will, all FAIL. A PASS requires an append-only record the provider cannot rewrite without detection: cryptographic commitments of each call at publication time, third-party archives outside their control, or exchange-verified execution records. This check comes first because every other number inherits from it; an editable history makes accuracy, expectancy, and drawdown figures decorative. The mechanics of doing this right cost nothing and are documented in the attestation explainer, which is exactly why declining to do it is informative. Phrasing to use: "Is there any way for me to verify a call you made six months ago was published then and hasn't changed?"
Check 2: are the losses on display?
Go looking for the worst month, the longest losing streak, and the maximum drawdown, stated as numbers. If the public materials show only winning screenshots and an equity curve that never breathes, FAIL. Any genuine trading history contains stretches that look like failure; systematic strategies in particular spend most of their calendar time below a previous high-water mark. A presented history without visible pain has been curated, and curation is the polite word. PASS requires the complete ledger, wins and losses, with the ugly summary statistics computed and shown. Phrasing: "What was your worst month, and where can I see it?"
Check 3: is "win" defined, and is expectancy stated?
Ask for the scoring rule in writing. If touching a first take-profit placed inside daily noise counts as a win regardless of the stop-out that follows, the advertised accuracy is manufactured arithmetic, worked through in the 95% lie, and the provider FAILS. PASS requires two things: a written definition of a completed trade's outcome, and an expectancy or profit-factor figure net of fees. A provider who cannot state their average loss has either never computed it or will not say; both are answers. Phrasing: "If a signal hits TP1 and then stops out, how is it scored, and what is your expectancy per trade after costs?"
Check 4: does the record span a bear market?
Check the start date of the verifiable record, not the founding date in the marketing. A history that begins after the last major bottom has never been tested by the only conditions that matter, and crypto's winters are where signal operations quietly die, delete, and relaunch under new names. FAIL is a bull-only record presented as proof of skill. PASS is either a record through at least one full regime change, or the rare honest sentence "we have not traded a winter yet," which deserves real credit. The reasoning, and the reason bull-phase brilliance is the cheapest commodity in this market, is the subject of everyone is a genius in a bull market. Phrasing: "Show me your 2022, or its equivalent."
Check 5: is there a methodology you could criticize?
You are not owed the secret sauce; you are owed the shape of the kitchen. What data feeds the process? What conditions produce a signal? What risk rules govern position size and exit? What environments should this approach be expected to lose in? "Proprietary AI" with nothing further is not a methodology, it is a mood, and it FAILS. PASS requires disclosure specific enough that a knowledgeable reader could argue with the choices, because secrecy and accountability cannot occupy the same claim. Note the corollary: a provider with no stated losing environment is claiming an edge in all conditions, which is a claim nobody in the history of markets has honestly sustained. Phrasing: "In what market conditions does your approach lose, and how do I recognize them?"
Check 6: how does the provider actually make money?
Read the links before the testimonials. Exchange referral codes pay the provider a share of your trading fees, forever, which rewards making you trade more and bigger, not better; leverage platform kickbacks sharpen the same conflict. Paid placements from token projects convert the feed into advertising sold as conviction. FAIL is undisclosed referral or promotional revenue. PASS is disclosed economics dominated by subscriptions, where the provider prospers only if subscribers stay because the product deserves it. A referral-funded provider is not necessarily dishonest, but you now know which direction their errors will lean, and the volume-hungry arithmetic behind that lean is worked through in the Telegram assessment. Phrasing: "What percentage of your revenue is referral or promotional?"
Check 7: what survives without the screenshots?
The synthesis check. Mentally delete every image of a green PnL, every testimonial, every follower count, all of it either unverifiable or purchasable, and inventory what remains. Formulas? A complete ledger? A hash chain? Third-party tracking? Exchange-verified records? If the residue is empty, the screenshots were the product, and you have your answer. If the residue is substantial, you may be looking at one of the rare honest operators, and your remaining question is whether their honestly-documented process actually suits your temperament and horizon, which no checklist can answer for you.
Scoring, and the fifteen-minute walkthrough
Score one point per pass. Seven is essentially unheard of. Five or six merits a serious look. Three or four means demand the missing artifacts before any money moves. Two or fewer, walk, whatever the returns claimed, BECAUSE of the returns claimed. The walkthrough in practice: open their public channel and website (checks 1 and 2, five minutes), send the two scoring questions (check 3, one message), find the record's true start date (check 4, two minutes), read the methodology page if one exists (check 5, three minutes), and inspect their links for referral parameters (check 6, two minutes). Check 7 is the minute of honest inventory at the end.
Applied to this site, both directions
Fairness demands we sit our own exam. Check 1: pass, every daily publication is committed to a public chain. Check 2: pass by construction today, because the board makes no performance claims whose losses could be hidden, and our founding experiment published its losses in full. Check 3: pass in advance, our scoring philosophy is public before any strategy exists, per measurement vs advice. Check 4: CANNOT PASS YET, and we say so plainly: this site's attested record is young, and years of chain are the only cure. Check 5: pass, every formula is published. Check 6: pass, subscriptions only, no referral links, disclosed. Check 7: the residue is the formulas, the chain, and the experiments, which is the entire product. Five and a half of seven, with the deficit stated and dated. Hold everyone to this list, including us, especially when our record gets old enough to be worth bragging about. Decisions are yours.