⟳ Updated every 2h
Simulation — what a small account would have made by following the AI
Simulation based only on signals where the AI proposed an SL and TP and whose result is known. Starting budget 5000 $, fixed risk 100 $/trade (2%), no compounding, no fees or slippage. Indicative only — not a real track record. Expired signals are excluded.
3 trades over 2 days
Simulation — what we avoided (or missed) by ignoring weak signals
Same simulation (5000 $, 100 $/trade) but on the discarded signals (score too low). We calculate what would have happened if they had been taken: when SL would have been hit → money saved by the filter; when TP would have been hit → gain missed because of the filter.
11 signals over 5 days
Comparison — does the AI outperform the indicator alone?
We replay the exact same simulation (5,000 $, 100 $/trade) twice on the same signals: once using the raw indicator SL/TP levels, once using the AI-suggested levels. The greener column shows which approach was more profitable.
3 signals compared
⚠️ Trades taken with warning (score 55–59)
These signals have an imperfect context (e.g. opposing trend, unvalidated filter) — tracked separately to determine whether taking them despite the warning is worth it, or whether they should be systematically ignored.
Indicator Stop Loss vs AI-suggested Stop Loss
The indicator always places its SL exactly at 1R (fixed distance). The AI suggests an SL positioned on market structure — often a bit further away, but better placed. This section compares both approaches on taken signals (≥55) to determine which better protects the trade without cutting winners.
3 signals analyzed (taken signals with AI SL available)
Results of taken signals
These figures concern only signals deemed strong enough to follow (score above the threshold). The win rate indicates how often price reached the target (TP) before the stop-loss (SL). The higher this figure, the more reliable the system.
TP reached 3
Zoom on performance by direction (buy vs sell), quality of targets set, and recent trend to see if the system is improving.
2 (67%)
1 (33%)
0 (0%)
1 (33%)
0 (0%)
2 (67%)
% winners by signal strength (score)
A signal with a high score should produce more winning trades than a weak signal — this table verifies whether that's actually the case.
100% (1/1)
100% (2/2)
% winners by trigger type
Some trigger types work better than others — this table shows which ones, so you know which setups to focus on.
100%
✅ 1
❌ 0
/1
100%
✅ 2
❌ 0
/2
Ignored signals — was the filter right to skip them?
These signals had a score too low to trigger an alert. We verify here in hindsight what would have happened if they had been taken: losses avoided or gains missed?
Well ignored: 3
Missed gains: 8
11 resolved ignored signals
Global score — is the system making the right decisions?
Combines taken AND ignored signals to answer one question: in what proportion did the system get it right — whether by alerting on a winning signal, or ignoring one that would have lost?
Correct: 6
Incorrect: 8
14
1✅1❌0
2✅2❌0
11✅3❌8
Analysis quality — is the system well calibrated?
These statistics cover all signals (taken and ignored) and the automatic post-trade reports generated after each resolution. They verify whether the score, SL and TP are correctly calibrated, and whether the AI correctly reads market direction.
Signal distribution by score (all signals)
3 TP too ambitious (price couldn't reach it) · 5 TP too conservative (price went much further)
🕐 Impact des sessions & Kill Zones sur les signaux
Chaque ligne correspond à une fenêtre horaire UTC. Kill Zones = moments de liquidité maximale où les institutions entrent sur le marché. Le win rate, le nombre de signaux pris et ignorés, et le taux de bonnes décisions permettent de voir quelles sessions sont les plus favorables.
| Session / KZ | Total | Conv. | Warn. | Ignored | Ignored decisions | Win% | Good dec.% |
|---|---|---|---|---|---|---|---|
|
Tokyo
|
2 (14%) | — | — | 2 | ✅0(0%) ❌2(100%) | — | 0% |
|
London
|
2 (14%) | — | 1 | 1 | ✅1(100%) ❌0(0%) | 100% (1/1) | 100% |
|
NY KZ Kill Zone |
1 (7%) | — | — | 1 | ✅0(0%) ❌1(100%) | — | 0% |
|
NY/London
|
4 (29%) | 1 | 1 | 2 | ✅2(100%) ❌0(0%) | 100% (2/2) | 100% |
|
NY
|
4 (29%) | — | — | 4 | ✅0(0%) ❌4(100%) | — | 0% |
|
Hors session
|
1 (7%) | — | — | 1 | ✅0(0%) ❌1(100%) | — | 0% |
KZ = Kill Zone (max liquidity) · Win% = taken resolved signals · Good dec.% = taken winner + ignored loser avoided