Pick a coin. Type a name in the Analyze tab. Free accounts cover eight major pairs; Premium covers all Binance spot pairs.
Read the decision, not just the number. Every result names the conditions behind it.
Check the risk panel. Volatility, liquidity and data quality decide whether a signal is worth anything.
Follow the paper record. The Validation page shows how previous signals actually resolved — losses included.
LONG, SHORT and WAIT
LONG
Conditions currently favour upward movement. A hypothetical entry, stop and two targets are recorded.
SHORT
Conditions currently favour downward movement, recorded the same way.
WAIT
The evidence does not clear the thresholds. This is a real answer — see Risk Center.
Confidence is not probability
Confidence is a checklist pass-rate. Eight of ten checks passing
shows 80% — that means eight conditions currently line up. It does
not mean an 80% chance of profit.
The two would only be the same if the checklist had been calibrated against
realised outcomes across a large sample. That calibration has not been done, so we
do not describe confidence as a probability anywhere.
Key terms
Entry
The reference price at the moment a signal was recorded.
Stop-loss (SL)
The level at which the hypothetical position is treated as wrong.
TP1 / TP2
Targets at 1.15R and 2.1R.
R
One unit of risk — the entry-to-stop distance.
Paper signal
A recorded hypothetical. No order is placed, ever.
Expired
Neither target nor stop was reached inside the observation window. Counted and published, not hidden.
Cost-adjusted
The same result after subtracting estimated fees and slippage. Always shown separately from the raw figure.
How win rate is calculated
Win rate = wins ÷ closed signals
Only closed signals count — those that reached a target or a stop.
Open and expired positions are excluded from the rate because neither reached a
barrier. Every published percentage carries its sample size, and the counts for
open, expired and WAIT are shown alongside so the denominator is never hidden.
The engines, and how mature each one is
Feature status is stated plainly so research is never presented as a finished
product.
Core 10-check scorecard LIVE
The directional checklist behind every signal. Details
Options, network and market-making research approximations. Details
SARVAM-369 BETA
Master decision layer that can veto to WAIT. Details
ML ensemble ROADMAP
Not shipped. No model is connected; its weight is redistributed to the active engines.
Walk-forward validation ROADMAP
Not built. Thresholds are reasoned, not fitted to out-of-sample results.
LIVE = in production and exercised by tests · BETA = shipped, still changing ·
EXPERIMENTAL = research approximation, treat output as indicative ·
ROADMAP = not implemented.
Subscription
Free
Eight major USDT pairs, signals, paper record and validation data.
Premium
All Binance spot pairs, market scanner, deeper history and research tools.
Premium unlocks coverage and tooling. It does not improve the
probability that any signal succeeds, and we will not claim that it does.
Payment is not automated yet — an administrator grants Premium manually.
Frequently asked
Does it trade for me?
No. There is no exchange connection and no order placement anywhere in the system.
Do you need my API keys?
Never. We do not ask for exchange keys, passwords, wallet keys or seed phrases.
Is this financial advice?
No. It is educational research and is not personalised to your circumstances.
Can I verify past signals?
Yes — the public ledger is hash-chained and independently verifiable.
Why so many WAIT results?
Because most of the time there is no edge worth acting on. See the Risk Center.
How do I delete my account?
Account → Delete account. It removes your data; see the privacy policy.