LiquidMind Core Engine/
Scoring & Ranking Engine
TL;DR
Every setup is mathematically graded 0–100 before execution. LiquidMind weights each component differently depending on the role of the timeframe being evaluated — Origin, LTF Entry, or other HTF confluence — and then aggregates across all timeframes using a per-origin weighting matrix.
POI timeframe — main context
Entry execution timeframe
HTF confluence timeframes
| Origin | W | D | 4H | 1H | 15m | 5m | 1m |
|---|---|---|---|---|---|---|---|
| W | 50% | 10% | 10% | 5% | 25% | — | — |
| D | 10% | 45% | 10% | 5% | 30% | — | — |
| 4H | 5% | 10% | 45% | 10% | 30% | — | — |
| 1H | 2% | 3% | 10% | 45% | 10% | 30% | 5% |
| 15m | 2% | 3% | 5% | 5% | 45% | 10% | 30% |
How It Works
- 1
Origin TF (main context): Trend Alignment 44% · POI Strength 33% · FVG Liquidity 11% · MCB Momentum 11%. Trend and POI receive a 1.5× priority boost because this is the timeframe that defines the setup.
- 2
LTF Entry TF: Displacement 75% · MCB Momentum 15% · FVG Liquidity 10%. Entry execution is dominated by displacement quality — a clean, impulsive move away from the POI is the primary gating condition.
- 3
Other HTF (confluence): Trend Alignment 40% · MCB Momentum 40% · FVG Liquidity 20%. Higher timeframe alignment is judged equally on trend direction and MCB momentum structure.
- 4
Multi-TF Aggregation: The final score is a weighted sum across all active timeframes. For a 1H origin, the 1H score carries 45%, the 5m entry carries 30%, and 4H/D/W provide confluence at 10%/3%/2%. Each origin TF has its own aggregation map.
- 5
Grade Thresholds: A+ (90–100) Excellent · A (80–89) Very Good · B (65–79) Good · C (50–64) Average · D (30–49) Below Average · E (0–29) Poor/Rejected.
The engine never treats all timeframes equally. A Weekly origin setup assigns 50% of its final score to the Weekly itself and 25% to the 15m entry — reflecting the reality that a monthly-scale POI demands precise entry timing. Grades D and E result in automatic rejection. Grades A and A+ unlock full position sizing. This mathematical structure ensures the bot scales into conviction, not guesswork.