Grades that are measured
A conviction grade should be calculated, not vibed in a chat message.
Most rooms attach confidence words to calls — “strong”, “high conviction”, a row of fire emoji — that mean whatever the sender wants on the day. A measured conviction grade works differently: it pins a call to its place in its own model's return spread, with arithmetic underneath.
With the pick, each call is tagged from A (top) down to D (bottom), and the cutoff is fixed model by model. The E grade was pulled from the live product, which keeps the four-step scale meaningful. The bar that earns an A is set against each model's own returns, which is why the same letter means different absolute moves on different clocks:
| Model | Horizon | What earns an A (per trade) |
|---|---|---|
| Intraday | opened and closed inside one session | ~0.70% per trade |
| Multi-Session | carried across part of a day to a couple of sessions | ~4.50% per trade |
| Swing | held roughly one to four weeks | ~6.00% per trade |
| Position | long-horizon, higher-conviction holds | long-form grading (no single bar) |
An A marks the top band of a model's own measured return spread; D is the weakest band still sent. The threshold is set per horizon, so an A on a same-session call (around 0.70% a trade) and an A on a multi-week swing (around 6.00%) both read as “top band for this clock” rather than one absolute target forced across very different holding times. The E band is gone — it left the live product in 2026, leaving four steps that each still carry weight.
Why per-model calibration matters
One blanket cutoff spanning all four clocks would quietly shortchange the fast model and inflate the slow one: a 0.70% same-session move and a 6.00% multi-week move differ too much in raw size for a single absolute target to say anything. Grading each call against its own model's spread means a B on a fast call and a B on a swing call each say the same thing — “above-typical for this clock” — which is exactly the signal a trader who cannot take every call needs in order to know when to size up.
And since the grade is folded into the on-chain hash (see sealed before the result), it sets before the result and stays put once the trade is done. That is what stops a grade from being a marketing dial turned up after a winner prints — the kind of quiet edit a chat window invites.
What failing this test looks like
A grade fails this test as soon as it is a word in place of a number — and an unhashed grade fails twice over, since it can be re-typed after the event as well.
- Free Discord rooms. The owner decides what gets posted and when, and a message can be edited in place or deleted with no trace, so the room fails sealed before the result outright — and the count too, because the losing calls can simply be removed before anyone tallies them.
- Paid Discord rooms. Charging a fee changes the price, not the proof. The post history still lives inside a chat the operator controls, so a paid room fails sealed before the result exactly as a free one does; the pricing being public is the only test it reliably clears.
- Copy-trade communities. More checkable than a chat, since a platform tracks participant results — but the calls are seldom stamped per signal and seldom graded, so the community fails sealed before the result and a measured grade even where a rough count exists.
- Re-poster and mirror bots. They forward other people's calls into a new channel without auditing them, so every verification gap in the original is inherited unfixed. They fail a re-openable record by construction.
This is why the guide compares a field of channel types rather than reviewing one product: a grade tied to measured returns and frozen before the outcome is the test most of the field cannot clear, which is exactly what makes clearing it worth paying for.
The grade is among the fields the timestamp guards, and it carries weight only atop a re-openable record — a grade tuned to a curated highlight reel is tuned to a fiction. To watch the grade and the rest of the fields verified together on one call, work through the verification walkthrough.