How VERIFIED SCORE is calculated
Plenty of sites publish a score. Almost none publish the formula. Here are the weights and the bounds, all of them, so you can reproduce the same number yourself.
Three principles
- The only inputs are measured values from MT5 real-tick backtests. Nobody adds points by hand.
- The formula is published in full (this page).
- Thresholds are fixed absolute values, not percentiles of the current line-up. With percentiles, adding one EA would move every existing EA's score and contradict figures we had already published.
The four axes and their weights
| Axis | Weight | Inputs | 0 points → 100 points |
|---|---|---|---|
| Drawdown | 35% | Maximum drawdown on equity. 50% scores zero, 5% or below scores full marks. | DD 50% → 5% |
| Robustness | 30% | The mean of profit factor and recovery factor. | PF 1.0 → 3.0 / RF 0 → 8 |
| Consistency | 20% | Maximum consecutive losses, and how far the win rate sits from 50%. A 90%-win-rate averaging system and a 25%-win-rate trend system should not be measured with the same ruler. | MCL 25 → 3 / |WR−50| 45 → 0 |
| Sample size | 15% | The mean of total trades (log scale) and years covered. | 30 → 2,000 trades (log) / 3 → 15 years |
Grades
The total is cut into four bands: A from 80, B from 65, C from 50, D below that.
What we left out on purpose
The Sharpe ratio is not an axis. In our measurements some EAs return outliers such as 43.88, and a single anomaly would dominate the total. The raw figure is still shown on each EA's detail page.
Missing data
An axis with no material is not scored zero. It is removed together with its weight, and the remaining weights are renormalised, so that absent data is never treated as a penalty.
The fifth axis (OOS) is still to come
By design, out-of-sample performance belongs as a fifth axis. It scores nothing today because the site has no forward-results data layer yet. It will be added once that material exists, and from that day full marks will be defined across five axes. Until then, full marks means 100 across four.
Score breakdown
| EA | Drawdown35% | Robustness30% | Consistency20% | Sample size15% | Total |
|---|---|---|---|---|---|
| DONCHIAN ENGINE USDJPYduka | 87DD 10.83% | 60PF 1.50 / RF 7.53 | 65MCL 11 / WR 34.5% | 63631 trades / 9.4y | 71B |
| ETHEREUM EA ETHUSDduka | 73DD 17.14% | 87PF 2.83 / RF 6.66 | 67MCL 12 / WR 38.3% | 38154 trades / 7.5y | 71B |
| AIFW XAUUSD | 100DD 4.60% | 45PF 1.77 / RF 4.07 | 60MCL 11 / WR 30.7% | 27228 trades / 3.7y | 65B |
| TRINITY BTCUSDm | 59DD 23.33% | 62PF 1.91 / RF 6.31 | 70MCL 13 / WR 43.1% | 50420 trades / 7.5y | 61C |
| ATLAS BTCUSDm | 53DD 26.29% | 51PF 1.48 / RF 6.17 | 59MCL 13 / WR 33.2% | 631,193 trades / 7.5y | 55C |
| BITCOIN COMET BTCUSDm | 42DD 31.18% | 49PF 1.72 / RF 4.95 | 73MCL 10 / WR 40.1% | 60775 trades / 8y | 53C |
| GOLD VIPER XAUUSDm (Exness) | 44DD 30.30% | 42PF 1.64 / RF 4.14 | 43MCL 40 / WR 56.4% | 7010,340 trades / 7.7y | 47D |
| MEGAMAX DON EURJPY EURJPYduka | 37DD 33.30% | 47PF 1.51 / RF 5.46 | 41MCL 17 / WR 25.2% | 62571 trades / 9.4y | 44D |
| GOLD KING v6 XAUUSDm | 0DD 91.27% | 73PF 4.66 / RF 3.74 | 47MCL 10 / WR 83.1% | 682,827 trades / 7.4y | 42D |
| ETHEREUM NOVA ETHUSDduka | 0DD 72.25% | 64PF 2.79 / RF 3.07 | 79MCL 6 / WR 37.5% | 3372 trades / 8.5y | 40D |
| GOLD NEURON XAUUSDduka | 0DD 55.54% | 25PF 1.35 / RF 2.64 | 80MCL 10 / WR 53.3% | 1002,766 trades / 22.7y | 39D |
| MEGAMAX DONCHIAN USDJPYduka | 0DD 58.59% | 55PF 1.55 / RF 6.64 | 65MCL 11 / WR 35.0% | 63629 trades / 9.4y | 39D |
| MEGAMAX DON GBPJPY GBPJPYduka | 0DD 56.75% | 19PF 1.32 / RF 1.70 | 83MCL 7 / WR 42.6% | 52256 trades / 9.4y | 30D |