OPEN-SOURCE SCRIPT
AetherEdge - Order Block Quality
🖊️ Overview
AE-OBQ does not just draw order blocks (OBs) — it learns and scores each one's quality = P(it holds and reacts), and applies the same learned model to scan a basket of symbols. Order blocks are not equal: some produce clean reactions, most do not. AE-OBQ learns the difference from the real outcome of past blocks (held / broken) and ranks which symbols currently show the highest-quality setups.
🔶 Key Features
Automatic order-block detection from a BOS plus strong displacement, drawn as zones
A quality score (ML) — online logistic learns P(hold and react) per OB; zone shading reflects quality
A multi-symbol screener — the identical learned model applied to up to 6 symbols, ranked by quality (symbol / bias / quality % / distance)
BUY/SELL signals on the first retest of a high-quality OB
ATR-normalized features so the learned quality model transfers across instruments
A gold HUD with active OB count, nearest OB's bias/quality/distance, and realized hold rate
OBs resolve on realized price — training and signals are no-repaint
🧠 Technical Architecture OB detection:
when a break of the recent swing high/low (BOS) comes with strong displacement (body/ATR), the last opposing candle before it is taken as the order block and its high/low drawn as a zone. Quality model (ML): five features — OB size, displacement strength, trend alignment (higher trend vs OB direction), momentum, and formation volume ratio — are all ATR-normalized and bounded (no standardization layer), and an online logistic regression learns quality = P(hold and react). Each OB trains at resolution with its realized label (held = rejected after retest; broken = zone violated) — no lookahead. Because features are ATR-normalized, the same weights apply directly to other symbols. Screener: request.security fetches the same features for each symbol, and the chart-learned weights are applied as-is to produce a quality score. A selection sort ranks them descending; the chart symbol runs through the identical pipeline and appears first. Honest scope: a linear classifier over order-block features. Not deep learning, and not a guarantee of win rate.
⚙️ Recommended Settings & Tuning Guide
Key parameters: structure length (swingLen), min displacement (body/ATR), quality threshold (qThr), resolve horizon (resN), screener symbols/timeframe.
Larger swingLen → only major BOS, focusing on important OBs; smaller → more detections
Raise min displacement → only blocks at the origin of forceful moves
Raise qThr → restrict highlights/signals to high-quality OBs
Crypto starting points (tune on your chart):
BTC / ETH (15m–4H): defaults are the baseline (swingLen 10, dispK 1.0, qThr 0.60)
SOL / XRP and high-vol alts: dispK 1.5 to exclude noisy OBs
Scalping (1–5m): swingLen ~5, shorter resN
Swing (4H–daily): swingLen 15+, longer resN to give the hold judgment room
Keep screener symbols within one asset class for stable quality comparison
💡 How to Use in Practice
Trade reactions off OBs: buy the dip when price returns to a high-quality bullish OB, sell the rally into a bearish one
Reading quality: a darker zone means the model rates it high quality; faint OBs are skip candidates
Hold rate: the share of OBs that actually held on this symbol/settings — a gauge of the model's reliability in that context
Use the screener: see at a glance which watchlist symbols currently rate highest, and focus on nearby high-quality OBs
Combinations: pair with AE-AMF's big-picture momentum or AE-FAM's anomaly detection, and use OBQ's quality to prioritize entry zones
⚠️ Important Notes
Learning period: no signals until warmup bars; the model needs enough resolved OBs to learn
Learning reset: changing inputs, symbol, or timeframe re-learns the weights
Screener granularity: the screener is a multi-symbol radar and computes each symbol's features with compact approximations (the chart symbol is the most precise) — confirm by charting the target symbol
Probability, not a guarantee: even high-quality OBs can break — use stops and position sizing
🚨 Disclaimer
This indicator is for educational and informational purposes only and is not financial advice or a recommendation to buy or sell. No method guarantees future profits; past performance does not indicate future results, and trading carries the risk of loss. All trading decisions are your own — use proper backtesting and disciplined risk management.
AE-OBQ does not just draw order blocks (OBs) — it learns and scores each one's quality = P(it holds and reacts), and applies the same learned model to scan a basket of symbols. Order blocks are not equal: some produce clean reactions, most do not. AE-OBQ learns the difference from the real outcome of past blocks (held / broken) and ranks which symbols currently show the highest-quality setups.
🔶 Key Features
Automatic order-block detection from a BOS plus strong displacement, drawn as zones
A quality score (ML) — online logistic learns P(hold and react) per OB; zone shading reflects quality
A multi-symbol screener — the identical learned model applied to up to 6 symbols, ranked by quality (symbol / bias / quality % / distance)
BUY/SELL signals on the first retest of a high-quality OB
ATR-normalized features so the learned quality model transfers across instruments
A gold HUD with active OB count, nearest OB's bias/quality/distance, and realized hold rate
OBs resolve on realized price — training and signals are no-repaint
🧠 Technical Architecture OB detection:
when a break of the recent swing high/low (BOS) comes with strong displacement (body/ATR), the last opposing candle before it is taken as the order block and its high/low drawn as a zone. Quality model (ML): five features — OB size, displacement strength, trend alignment (higher trend vs OB direction), momentum, and formation volume ratio — are all ATR-normalized and bounded (no standardization layer), and an online logistic regression learns quality = P(hold and react). Each OB trains at resolution with its realized label (held = rejected after retest; broken = zone violated) — no lookahead. Because features are ATR-normalized, the same weights apply directly to other symbols. Screener: request.security fetches the same features for each symbol, and the chart-learned weights are applied as-is to produce a quality score. A selection sort ranks them descending; the chart symbol runs through the identical pipeline and appears first. Honest scope: a linear classifier over order-block features. Not deep learning, and not a guarantee of win rate.
⚙️ Recommended Settings & Tuning Guide
Key parameters: structure length (swingLen), min displacement (body/ATR), quality threshold (qThr), resolve horizon (resN), screener symbols/timeframe.
Larger swingLen → only major BOS, focusing on important OBs; smaller → more detections
Raise min displacement → only blocks at the origin of forceful moves
Raise qThr → restrict highlights/signals to high-quality OBs
Crypto starting points (tune on your chart):
BTC / ETH (15m–4H): defaults are the baseline (swingLen 10, dispK 1.0, qThr 0.60)
SOL / XRP and high-vol alts: dispK 1.5 to exclude noisy OBs
Scalping (1–5m): swingLen ~5, shorter resN
Swing (4H–daily): swingLen 15+, longer resN to give the hold judgment room
Keep screener symbols within one asset class for stable quality comparison
💡 How to Use in Practice
Trade reactions off OBs: buy the dip when price returns to a high-quality bullish OB, sell the rally into a bearish one
Reading quality: a darker zone means the model rates it high quality; faint OBs are skip candidates
Hold rate: the share of OBs that actually held on this symbol/settings — a gauge of the model's reliability in that context
Use the screener: see at a glance which watchlist symbols currently rate highest, and focus on nearby high-quality OBs
Combinations: pair with AE-AMF's big-picture momentum or AE-FAM's anomaly detection, and use OBQ's quality to prioritize entry zones
⚠️ Important Notes
Learning period: no signals until warmup bars; the model needs enough resolved OBs to learn
Learning reset: changing inputs, symbol, or timeframe re-learns the weights
Screener granularity: the screener is a multi-symbol radar and computes each symbol's features with compact approximations (the chart symbol is the most precise) — confirm by charting the target symbol
Probability, not a guarantee: even high-quality OBs can break — use stops and position sizing
🚨 Disclaimer
This indicator is for educational and informational purposes only and is not financial advice or a recommendation to buy or sell. No method guarantees future profits; past performance does not indicate future results, and trading carries the risk of loss. All trading decisions are your own — use proper backtesting and disciplined risk management.
Script open-source
Dans l'esprit TradingView, le créateur de ce script l'a rendu open source afin que les traders puissent examiner et vérifier ses fonctionnalités. Bravo à l'auteur! Bien que vous puissiez l'utiliser gratuitement, n'oubliez pas que la republication du code est soumise à nos Règles.
Clause de non-responsabilité
Les informations et publications ne sont pas destinées à être, et ne constituent pas, des conseils ou recommandations financiers, d'investissement, de trading ou autres fournis ou approuvés par TradingView. Pour en savoir plus, consultez les Conditions d'utilisation.
Script open-source
Dans l'esprit TradingView, le créateur de ce script l'a rendu open source afin que les traders puissent examiner et vérifier ses fonctionnalités. Bravo à l'auteur! Bien que vous puissiez l'utiliser gratuitement, n'oubliez pas que la republication du code est soumise à nos Règles.
Clause de non-responsabilité
Les informations et publications ne sont pas destinées à être, et ne constituent pas, des conseils ou recommandations financiers, d'investissement, de trading ou autres fournis ou approuvés par TradingView. Pour en savoir plus, consultez les Conditions d'utilisation.