Files
gogo2/WILLIAMS_PIVOT_SYSTEM.md
Dobromir Popov e993bc2831 cache, pivots wip
2025-10-20 15:21:44 +03:00

117 lines
4.1 KiB
Markdown

# Williams Market Structure Pivot System
## Overview
The dashboard now displays the complete Williams Market Structure analysis with **5 recursive levels of pivot points** and **trend analysis** (higher highs/higher lows for uptrends).
## What's Displayed
### 1. Five Levels of Recursive Pivots
Each level analyzes pivot points at different timeframes:
- **Level 1** (Short-term): Direct swing highs/lows from 1-minute OHLCV data
- Small red/blue triangles (size 6)
- Most granular view of market structure
- **Level 2** (Medium-short): Pivots from Level 1 pivots
- Medium orange/light blue triangles (size 8)
- Filters out noise from Level 1
- **Level 3** (Medium): Pivots from Level 2 pivots
- Light orange/lighter blue triangles (size 10)
- Intermediate trend structure
- **Level 4** (Medium-long): Pivots from Level 3 pivots
- Yellow/very light blue triangles (size 12)
- Longer-term swing points
- **Level 5** (Long-term): Pivots from Level 4 pivots
- Bright yellow/pale blue triangles (size 14)
- Broadest market structure view
### 2. Trend Analysis (For Each Level)
The system calculates trend direction using classic Williams methodology:
- **Uptrend** ↑: Higher highs AND higher lows
- **Downtrend** ↓: Lower highs AND lower lows
- **Sideways** →: Mixed or insufficient pattern
### 3. Trend Strength (0-100%)
Each level also provides a trend strength score based on:
- Consistency of pivot point progression
- Average strength of individual pivots
- Number of confirming pivots
## Dashboard Display
### Pivot Markers
- **High Pivots**: Triangle pointing DOWN (potential resistance)
- **Low Pivots**: Triangle pointing UP (potential support)
- **Color/Size**: Larger and brighter = higher level = longer-term significance
### Trend Annotation (Top-left corner)
Shows all 5 levels of trend analysis:
```
Williams Trends:
L1: ↑ 75% (Short-term uptrend, strong)
L2: ↑ 68% (Medium-short uptrend, strong)
L3: → 52% (Medium sideways)
L4: ↓ 61% (Medium-long downtrend, moderate)
L5: ↓ 71% (Long-term downtrend, strong)
```
The border color indicates the **Level 5 (longest-term) trend**:
- Green border = Long-term uptrend
- Red border = Long-term downtrend
- Orange border = Long-term sideways
## How It Works
### Recursive Calculation
1. **Level 1**: Find swing highs/lows in raw 1m OHLCV data
2. **Level 2**: Treat Level 1 pivots as "candles" and find swing points among them
3. **Level 3**: Treat Level 2 pivots as "candles" and find swing points
4. **Level 4**: Treat Level 3 pivots as "candles" and find swing points
5. **Level 5**: Treat Level 4 pivots as "candles" and find swing points
### Trend Detection
For each level, the system:
1. Gets the last 10 pivot points (or all if < 10)
2. Separates into highs and lows
3. Compares most recent highs/lows to previous ones
4. Determines if forming higher highs/higher lows (uptrend) or lower highs/lower lows (downtrend)
## Usage
### Toggle Pivots On/Off
Use the **"Show Pivot Points"** toggle in the Session Controls panel.
### Interpret Multiple Levels
- **All levels aligned**: Strong directional move
- Example: L1 L2 L3 L4 L5 = Strong sustained uptrend
- **Lower levels diverging**: Potential reversal
- Example: L1 L2 L3 L4 L5 = Short-term correction in long-term uptrend
- **Mixed signals**: Market indecision or transition
- Example: L1 L2 L3 L4 L5 = Choppy market within long-term uptrend
### Trading Applications
- **Level 1-2**: Scalping entries/exits
- **Level 3**: Intraday swing trades
- **Level 4-5**: Position trading, trend confirmation
## Data Source
- Uses last **2000 candles** of 1-minute data
- Recalculated on each chart refresh
- Cached in pivot bounds system (refreshes daily)
## Technical Details
- **Algorithm**: Larry Williams Market Structure Analysis
- **Minimum pivot distance**: 3 candles (configurable)
- **Pivot strength**: Based on distance from surrounding highs/lows, volume, and formation duration
- **Files**:
- `core/williams_market_structure.py`: Core algorithm
- `core/data_provider.py`: Data integration
- `web/clean_dashboard.py`: Visualization