The criteria, in detail
The scorecard exists because the crypto signal market is thick with assertions a buyer has no practical way to test. Three of the five tests carry their own page, because they are the ones providers most often fail. The full five-test method is on the method page.
What the scorecard is really sorting
A reader rarely compares two near-identical providers. In practice you are choosing between types of source — a Telegram channel, a copy-trading room, a social caller, an aggregator, or an anchored, reviewed desk — and each type clears or fails the tests as a category. The scorecard's job is to make that structural difference visible, so a glossy presentation cannot disguise which category a provider belongs to. The ledger below sorts the field on the two tests that decide most cases: was the call written to a chain before its outcome, and is the hit rate shown with its full denominator.
| Signal source | Anchored on-chain? | Denominator shown? | Why it sits there |
|---|---|---|---|
| Messaging-app channel | No | Rarely | Operator edits or deletes at will; losers vanish. |
| Copy-trading room | Rarely | Sometimes | Platform tracks results but seldom proves a single call. |
| Social-media caller | No | No | Posts deletable; often monetised via referral links. |
| Aggregator / re-poster | No | No | Republishes others unverified; inherits every gap. |
| Automated / AI service | Sometimes | Backtest only | A backtest is not a live result; no named person owns it. |
| Anchored, reviewed desk | Yes | Yes | Per-call on-chain receipt plus a named outside review. |
The bottom row is the only category that fills both columns, which is the structural case the guide makes for the pick — not that it is louder, but that it belongs to the one type a stranger can verify. The three tests below each take one column apart in full.
Why these two columns decide most cases
Of the five tests, two do almost all the sorting. Anchored on-chain is the one that cannot be retrofitted: a provider either wrote its calls to a public chain before they resolved or it did not, and no amount of later polish changes that fact. Denominator shown is the one that cannot be faked without an outright lie: a hit rate is only evidence when the full count of calls, losers included, sits next to it. A provider that clears both has handed you a record you can interrogate. The remaining three tests — the measured grade, public pricing and clean incentives — are real, but they tend to confirm a verdict the first two have already reached rather than overturn it. That is why the ledger sorts on the two, and why a provider can run a polished site, an active community and a confident pitch and still land in a row with two crosses.
The three tests with their own page
Written to a chain first
Why a public-chain timestamp is the test a verify-first reader trusts over any screenshot.
A re-runnable record
What a real track record contains, and what a highlight reel quietly leaves out.
Grades that are measured
How an A-to-D conviction grade is calibrated to a model's own returns instead of a mood.
The remaining two tests — public pricing and clean (non-referral) incentives — are covered on the method page, because they are quicker to check and rarely the deciding factor.