OPEN-SOURCE SCRIPT
Order Block AI [PickMyTrade]
Most Order Block indicators draw zones and leave you guessing which ones to trade. Order Block AI adds a machine learning engine that learns from every historical order block retest on your chart — comparing current market conditions to past setups and returning a data-driven confidence score before you enter.
The core engine is K-Nearest Neighbors (KNN) with Lorentzian distance. Rather than applying fixed rules, it stores every past order block retest as a feature vector — RSI level, volume strength, and ADX trend reading at the moment of retest. It then records what happened: did price reach the take profit or reverse to the stop? On each new signal, the model finds the K most similar historical episodes and votes on the likely outcome. The confidence % you see on the label is the model's output — not a static threshold, but a score that adapts as new market data resolves.
────────────────────────────────────
3 Layers of Confirmation
Layer 1 · Market Structure
Pivot-based detection of confirmed swing highs and lows:
Layer 2 · Order Block + Liquidity Sweep
A signal requires price to actively retest the OB zone. No retest = no signal.
Layer 3 · KNN Classification
When a retest occurs, the KNN engine retrieves the K most similar historical setups using Lorentzian distance — a log-space metric that handles outliers better than standard Euclidean distance. The confidence score reflects how many of those K analogues resolved in the signal direction. The model updates continuously as past signals resolve, meaning it gets more informed over time without any manual adjustment.
Signal fires only when all 3 align:
CHoCH or Liquidity Sweep → Order Block Retest → KNN Confidence ≥ Threshold
────────────────────────────────────
Non-Repainting
Pivots require full left + right bar confirmation before any signal is issued. The label appears at the confirmed bar — what you see is what was tradeable at the time.
────────────────────────────────────
Settings
Market Structure
Order Blocks
Fair Value Gaps
KNN Engine
Risk Management
Display
────────────────────────────────────
Alerts
────────────────────────────────────
Notes
────────────────────────────────────
Risk Disclosure
Confidence scores reflect the frequency of past patterns in the KNN training history — they do not guarantee future outcomes. Use in combination with your own position sizing and risk management. Not financial advice.
The core engine is K-Nearest Neighbors (KNN) with Lorentzian distance. Rather than applying fixed rules, it stores every past order block retest as a feature vector — RSI level, volume strength, and ADX trend reading at the moment of retest. It then records what happened: did price reach the take profit or reverse to the stop? On each new signal, the model finds the K most similar historical episodes and votes on the likely outcome. The confidence % you see on the label is the model's output — not a static threshold, but a score that adapts as new market data resolves.
────────────────────────────────────
3 Layers of Confirmation
Layer 1 · Market Structure
Pivot-based detection of confirmed swing highs and lows:
- BOS (Break of Structure) — trend continuation
- CHoCH (Change of Character) — first counter-trend break, potential reversal
Layer 2 · Order Block + Liquidity Sweep
- Bullish OB — last bearish candle before a bullish structure break
- Bearish OB — last bullish candle before a bearish structure break
- Liquidity Sweep — wick beyond last swing that closes back inside (stop hunt)
A signal requires price to actively retest the OB zone. No retest = no signal.
Layer 3 · KNN Classification
When a retest occurs, the KNN engine retrieves the K most similar historical setups using Lorentzian distance — a log-space metric that handles outliers better than standard Euclidean distance. The confidence score reflects how many of those K analogues resolved in the signal direction. The model updates continuously as past signals resolve, meaning it gets more informed over time without any manual adjustment.
Signal fires only when all 3 align:
CHoCH or Liquidity Sweep → Order Block Retest → KNN Confidence ≥ Threshold
────────────────────────────────────
Non-Repainting
Pivots require full left + right bar confirmation before any signal is issued. The label appears at the confirmed bar — what you see is what was tradeable at the time.
────────────────────────────────────
Settings
Market Structure
- Pivot Left / Right Bars — confirmation bars for swing detection
Order Blocks
- Show/hide zones, bull and bear colors
Fair Value Gaps
- Show/hide 3-candle imbalance zones (visual context only — does not affect signals)
KNN Engine
- K Neighbors — how many historical analogues vote on the signal (default 5)
- History Length — maximum past signals stored for KNN learning (default 200)
- Min Confidence — minimum fraction of K neighbors that must agree to fire a signal (default 0.60)
- Outcome Bars — bars after a historical signal to evaluate its outcome (default 20)
Risk Management
- ATR Stop Multiplier and R:R Ratio — used by the KNN engine to score whether past signals reached TP or SL
Display
- Zen Mode — hides all visuals except entry signal labels
- Show/hide BOS and CHoCH structure labels
────────────────────────────────────
Alerts
- OB AI Long Signal
- OB AI Short Signal
────────────────────────────────────
Notes
- The KNN engine improves as it accumulates more resolved historical setups — give it time on the first load
- Works on any symbol and timeframe
- Lower the confidence threshold to see more signals; raise it to filter more aggressively
- Fair Value Gap zones are displayed as context — they do not affect signal logic
────────────────────────────────────
Risk Disclosure
Confidence scores reflect the frequency of past patterns in the KNN training history — they do not guarantee future outcomes. Use in combination with your own position sizing and risk management. Not financial advice.
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.
Automate TradingView alerts to brokers no coding needed. Trade futures, forex, or crypto, copy trades across unlimited accounts, & control risk with ease.
For Tradovate: bit.ly/4a1KBrY
For all other supported brokers: bit.ly/4qEbU2G
For Tradovate: bit.ly/4a1KBrY
For all other supported brokers: bit.ly/4qEbU2G
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.
Automate TradingView alerts to brokers no coding needed. Trade futures, forex, or crypto, copy trades across unlimited accounts, & control risk with ease.
For Tradovate: bit.ly/4a1KBrY
For all other supported brokers: bit.ly/4qEbU2G
For Tradovate: bit.ly/4a1KBrY
For all other supported brokers: bit.ly/4qEbU2G
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.