How to judge a reversion record
Understanding the method is half the skill. The other half is reading the evidence a service puts behind it — and that is where most mean-reversion claims fall apart. Three tests do most of the work; each has its own page below.
What you are really sorting
A reader rarely compares two near-identical services. In practice you are choosing between kinds of record — a chat-room call list, a mirror-trading room, an influencer's feed, a backtest screenshot, or an audited model book — and each kind can or cannot answer the key questions as a category. The job of these tests is to make that structural difference visible, so a polished presentation cannot disguise which category a record belongs to. The table below sorts the field on the two questions that decide most cases.
| Kind of record | Timestamped? | Full count shown? | Why it lands there |
|---|---|---|---|
| Chat-room call list | No | Rarely | The poster edits and deletes at will; losing calls vanish. |
| Mirror-trading room | Rarely | Sometimes | A platform logs results but seldom per-trade proof. |
| Influencer caller | No | No | Posts are deletable; income often comes from referral links. |
| Backtest screenshot | No | Yes | A fitted curve, not a live record; no one issued the trades. |
| Automated / AI feed | Sometimes | Backtest only | A simulation is not a forward result; no owner of the live record. |
| Audited model book | Yes | Yes | Per-trade on-chain receipt plus a named independent reviewer. |
The bottom row is the only kind of record that answers both questions, which is the structural case for the worked example — not that it is louder, but that it belongs to the one category a stranger can audit. Each of the three tests below takes one question apart in full.
Why these two questions decide most cases
Of the tests that matter, two do nearly all the sorting. Issued before the outcome cannot be retrofitted: a record either fixed its trades in public before they resolved or it did not, and no later polish changes that. The full trade count cannot be faked without an outright lie: a win rate is evidence only when the whole population of trades, losers included, sits beside it. A record that answers both has handed you something you can interrogate. The remaining tests — the measured grade, public pricing, aligned revenue — are real, but they tend to confirm a verdict the first two have already reached. That is why a service can have a confident pitch, an active community and a beautiful equity curve and still land in a row with two crosses.
The three tests, each on its own page
Issued before the outcome
Why a public timestamp is the test a reversion record cannot fake its way around.
A re-runnable record
What a real reversion track record contains, and what a fitted backtest leaves out.
The remaining two tests — published pricing and aligned (non-referral) revenue — are covered on the profitability guide, because they are quick to check and rarely the deciding factor.