Methodology
Methodology
We publish exactly how CopyRadar's quality score and red flags are computed. We avoid “black-box recommendations” — the rationale is open so anyone can verify it.
1. Data source
The analysis is based on each lead trader's publicly available closed position records. That means we aggregate individual trades ourselves (entry price, exit price, P&L, leverage, holding time, symbol) rather than relying on the exchange's summary figures. Only fully closed positions are used; open positions are excluded. We currently use public copy-trading data from Binance and OKX.
2. Measured metrics
| Metric | Definition | Why it matters |
|---|---|---|
| Sample size | Number of closed trades | With a small sample you cannot separate luck from skill. |
| Profit concentration | Share of total profit from the single largest winner | Returns that depend on one big hit rarely repeat. |
| Median holding time | Median position duration | Scalping cannot be followed because of copy latency. |
| Max leverage | Highest leverage across trades | High leverage increases single-event liquidation risk. |
| Profit factor | Gross profit ÷ gross loss | Shows how healthy profit is relative to losses. |
| Diversification | Number of distinct symbols traded | Concentration in a few symbols is exposed to event risk. |
| Copier P&L | Combined realized P&L of followers (exchange-provided) | Whether the people who copied actually made money. |
Note: available metrics differ per exchange. For example, OKX does not publish copier P&L or max drawdown, so those bonuses and penalties are not applied to OKX traders.
3. Red-flag criteria
4. Quality score
Every trader starts from a base of 60 points. Healthy factors add points (up to roughly +40) and risky factors deduct, so even the top of the table is meaningfully differentiated. The final score is clamped to 0–100.
- Sample confidence (more trades)up to +14
- Profit factor (max at PF 2.0)up to +12
- Diversification (max at 25 symbols)up to +10
- Copier profit+5
- Healthy win rate (40–75%)+3
- Small sample (<20 trades)−8 to −25
- High / elevated leverage−18 / −6
- Profit concentration / some−20 / −8
- Scalping (<1h)−15
- Copier loss−20
- Deep / moderate drawdown−12 / −5
- Symbol concentration (≤3)−8
5. Grades
An A requires 95+ and is deliberately strict — only about the top 8% earn it.
6. Limitations
This methodology is a risk-screening tool, not a prediction. It is based on past data and does not guarantee future performance; a high score does not mean there is no risk of loss. The data is a snapshot from the time of collection, and a trader's style can change at any time. Thresholds may be adjusted as the service improves, and this document is updated when they are.
Everything here is for reference and is not investment advice. Please read the investment disclaimer.