Setup Quality Detection

Purpose

"How good is this setup? Is it worth trading? Risk:Reward"

This component scores every potential trade from 0-100 and assigns a grade (A+ to F). Higher grades get larger position sizes. The scoring uses a 5-factor weighted system with penalties and bonuses.

Trading Market Principle

"Not all setups are created equal. Grade them, position size them accordingly."

A+ setup = everything aligned, maximum conviction → 2 lots
Poor setup = missing key factors → 1 lot or skip
Penalties protect capital → creeper, ma_struggle, wrong phase = reduce score
Bonuses reward quality → railroad, clean entry, near key level = boost score
A+ requires ALL criteria → one missing = capped at 79 (no A+)

5-Factor Weighted Scoring

Every setup is scored across 5 factors, each with a specific weight:

Factor	Weight	What It Measures	Score Source
Timeframe Alignment	30%	Are all timeframes pointing same direction?	MTF alignment_score × 100
Trend Strength	20%	Is the trend strong with no adverse conditions?	100 - penalties + bonuses
Entry Quality	15%	Are we entering at a good spot?	Near MA, clean entry bonuses
Key Level Proximity	20%	Are we near support/resistance/institutional levels?	Near key level check
Risk/Reward	15%	Is the R:R ratio acceptable?	Lookup table

Formula:

Weighted_Score = (TF × 0.30) + (Trend × 0.20) + (Entry × 0.15) + (KeyLevel × 0.20) + (RR × 0.15)

Factor Calculations

Factor 1: Timeframe Alignment (30%)

Input:  TimeframeAnalysisResult.alignment_score (0-1)
Output: Score 0-100

score = alignment_score × 100

Factor 2: Trend Strength (20%)

Starts at 100, then applies penalties and bonuses:

trend_score = 100

// PENALTIES (subtracted)
IF is_creeper_move:        trend_score -= 50
IF price_ma_struggle:      trend_score -= 30
IF NOT has_two_day_trend:  trend_score -= 30
IF trend_phase != MIDDLE:  trend_score -= 25

// BONUSES (added)
IF is_railroad_trend:      trend_score += 15

RETURN clamp(trend_score, 0, 100)

Factor 3: Entry Quality (15%)

entry_score = 100

IF near_key_level:    entry_score += 10   // Key level bonus
IF NOT near_ma:       entry_score -= 40   // MA distance penalty
IF clean_entry:       entry_score += 10   // Clean entry bonus

RETURN clamp(entry_score, 0, 100)

Factor 4: Key Level Proximity (20%)

Major Institutional Levels Only

Only Institutional levels are used for near_key_level detection:

Level	Description
PDH/PDL	Prior Day High/Low - "ceilings & floors", major makret memory points
200 MA Zone	25pt buffer - MAs are zones/cushions, not skinny lines
Weekly H/L	Prior Week High/Low - institutional reference levels

Excluded (too noisy): Swing levels, 21 MA, Yesterday's Close, HH/HL/LH/LL structure

level_score = 100

# TICKET-25: near_key_level uses major levels only (PDH/PDL, 200 MA zone, Weekly H/L)
IF NOT near_key_level AND NOT near_ma:
    level_score -= 50   // Major penalty

RETURN clamp(level_score, 0, 100)

Factor 5: Risk/Reward (15%)

Uses lookup table:

R:R Ratio	Score	Quality
< 1.0	0	Unacceptable
1.0 - 1.5	40	Poor
1.5 - 2.0	70	Acceptable
2.0 - 3.0	90	Good
≥ 3.0	100	Excellent

Penalties & Bonuses

Type	Condition	Value	Factor Affected
Penalty	Creeper move detected	-50	Trend Strength
Penalty	Price struggling near MA	-30	Trend Strength
Penalty	No two-day trend	-30	Trend Strength
Penalty	Phase != MIDDLE	-25	Trend Strength
Penalty	Not near MA	-40	Entry Quality
Penalty	Not near key level or MA	-50	Key Level Proximity
Bonus	Railroad trend detected	+15	Trend Strength
Bonus	Near key level	+10	Entry Quality
Bonus	Clean entry	+10	Entry Quality

Institutional Fight Multiplier

If institutional fight is in progress, apply 30% penalty:

IF institutional_fight_in_progress:
    final_score = weighted_score × 0.70

This reduces all scores by 30% when big players are battling.

A+ Criteria Enforcement

To get A+ grade, setup must meet ALL three criteria:

Criteria	Requirement
All Timeframes Aligned	`timeframe_analysis.aligned = True`
Entry Near MA	`entry_data.near_ma = True`
Two-Day Trend Present	`market_state.has_two_day_trend = True`

If ANY criteria missing: Score is capped at 79 (maximum A grade, no A+)

Grade Thresholds

Grade	Min Score	Position Size	Risk %	Auto-Trade
A+	90	2 lots	1.5%	Yes
A	80	1 lot	1.2%	Yes
B	70	1 lot	1.0%	No
C	60	1 lot	0.8%	No
D	50	1 lot	0.5%	No
F	<50	1 lot	0.3%	No

Example Calculation

Scenario: MTF aligned at 85%, railroad trend, near MA, R:R = 2.5, no two-day trend

STEP 1: Calculate Factor Scores
─────────────────────────────────────────────────
Factor 1 (TF Alignment):  85% × 100 = 85.0
Factor 2 (Trend):         100 - 30 (no 2-day) + 15 (railroad) = 85.0
Factor 3 (Entry):         100 + 10 (near MA) = 110 → capped at 100.0
Factor 4 (Key Level):     100.0 (near MA counts)
Factor 5 (R:R):           90.0 (R:R 2.5 is in [2.0-3.0) range)

STEP 2: Apply Weights
─────────────────────────────────────────────────
Weighted = (85×0.30) + (85×0.20) + (100×0.15) + (100×0.20) + (90×0.15)
         = 25.5 + 17.0 + 15.0 + 20.0 + 13.5
         = 91.0

STEP 3: A+ Enforcement
─────────────────────────────────────────────────
Missing two-day trend → Cap at 79.0

FINAL: Score = 79.0, Grade = A

Configuration

Parameter	Default	Description
`timeframe_alignment_weight`	0.30	30% weight
`trend_strength_weight`	0.20	20% weight
`entry_technique_weight`	0.15	15% weight
`key_level_proximity_weight`	0.20	20% weight
`risk_reward_weight`	0.15	15% weight
`creeper_move_penalty`	-50	Creeper penalty
`ma_struggle_penalty`	-30	MA struggle penalty
`two_day_trend_penalty`	-30	Missing 2-day penalty
`phase_mismatch_penalty`	-25	Wrong phase penalty
`institutional_fight_multiplier`	0.70	30% reduction

Output

SetupQualityResult:

Field	Type	Description
`grade`	Enum	A_PLUS, A, B, C, D, F
`score`	float	Final score 0-100
`factor_scores`	Dict	Individual factor scores
`position_size`	int	1 or 2 lots
`risk_percent`	float	Risk % for this grade
`can_auto_trade`	bool	Only A+ and A
`penalties_applied`	List	Which penalties were applied
`a_plus_criteria_met`	bool	All 3 criteria met?

Dependencies

Upstream:

multi_timeframe_analysis.py - Provides alignment score (Factor 1)
market_state_analysis.py - Provides creeper, railroad, two-day, phase

Downstream:

signal_generation_trade_management.py - Uses grade for filtering