Why Score Signals?
A scoring system helps avoid the trap where "every wiggle is a signal."By quantifying quality across multiple dimensions, we can separate high-probability setups from noise and ensure every alert has clear justification.
Score Components (0-100)
Breakout: compression quality + clean range. Mean Reversion: z-score magnitude + deceleration.
Breakout aligns with 1h trend? +10. Mean reversion in range regime? +10. Penalties if misaligned.
Spread acceptable? Expected move vs costs? Liquidity window? Penalties for thin hours.
Near 1h swing level? Key level reclaim/break? Multi-timeframe agreement?
Alert Thresholds
Setup Quality (40 pts)
- Compression ratio— Tighter = better (breakout)
- Z-score magnitude— Deeper overextension (MR)
- Clean boundaries— Clear support/resistance
- Confirmation candle— Strong close in direction
Regime Alignment (20 pts)
- Trend alignment— Breakout with 1h trend = +10
- Range regime— Mean reversion in range = +10
- Counter-trend— Penalty for misalignment
- Volatility state— Matches strategy expectations
Tradeability (25 pts)
- Spread check— Below threshold = +10
- Move vs costs— Expected move > 3-5x cost = +10
- Liquidity window— Active session = +5
- Rollover penalty— Near daily rollover = -pts
Confluence (15 pts)
- 1h swing level— Near key structure = +5
- Level reclaim— Break/reclaim of S/R = +5
- Multi-TF agreement— 1h and 5m aligned = +5
Explainability Requirement
Every alert must include complete transparency:
Score Breakdown
Full breakdown of all 4 components and their individual contributions
Rules That Fired
Exact conditions that triggered the signal (trigger, confirmation)
Cost Estimate
Spread + slippage buffer = total estimated cost to enter