Best Practices
This guide outlines recommended best practices to help you get the most accurate signals, efficient strategies, and consistent performance when using the Comps platform.
General Best Practices
Start simple before building complex strategies
Test preset and custom strategies across different assets and timeframes
Review signal logic and confirmation rules regularly
Avoid over-optimization, which may reduce real-world performance
Monitor market conditions—no strategy works in all markets
Signal Generation Best Practices
Choose the Right Strategy
Use preset strategies for quick and reliable signal generation
Use custom strategies once you understand how Comps interprets logic
Ensure your strategy is marked Run Strategy to appear in the signal cockpit
Select Appropriate Timeframes
Lower timeframes (1m–15m) are more sensitive to noise and volatility
Higher timeframes (1h–12m) provide more stable signals
Match timeframe selection to your trading style (scalping, swing, position trading)
Use Risk:Reward Wisely
Conservative settings reduce drawdowns but may generate fewer trades
Aggressive settings increase opportunity but also risk
Use custom Stop Loss and Take Profit levels only if you understand their impact
Always define a stop loss before entering any trade
Apply News Filters Strategically
Enable high-impact news filters during major economic events
Disable news filters only if your strategy is designed for volatility
Combining news filters with confirmation logic improves signal stability
Strategy Creation Best Practices:
Prompt Mode
Be specific and structured in your prompts
Clearly define:
Entry conditions
Exit rules
Stop loss and take profit logic
Avoid vague language that can lead to inconsistent interpretation
Composition Mode
Combine logic intentionally—more components do not always mean better results
Use confirmation logic to reduce false signals
Balance indicators across trend, momentum, and volume
Avoid conflicting conditions within the same strategy
Configuration Mode
Limit indicators to what adds real value
Use AI optimization unless you fully understand manual parameters
Keep Signal Generation and Signal Confirmation logic distinct
Regularly review indicator performance
Indicator Best Practices
Use no more than:
10–15 indicators for Signal Generation
10–15 indicators for Signal Confirmation
Avoid stacking indicators that measure the same thing
Let Comps AI optimize parameters for dynamic markets
Use manual settings only for proven, tested indicator values
AI Instruction Best Practices
Write clear execution instructions for AI behavior
Define priority rules if conditions conflict
Use templates as a baseline and customize as needed
Avoid unnecessary overrides unless required
Win Rate Optimization Best Practices
Increase optimization gradually
Monitor how optimization affects:
Signal frequency
Confidence scores
Trade quality
Do not rely on win rate alone—consider risk-to-reward balance
Luma Chat AI Best Practices
Ask specific, actionable questions
Increase chat power levels for deeper analysis
Use Luma Chat for validation, not blind confirmation
Credits & Power Level Management
Use lower power levels for testing and experimentation
Reserve higher power levels for critical trades
Enable Auto-Buy to avoid interruptions
Track credit usage in the dashboard
Performance & Scaling Best Practices
Save strategy versions before major changes
Export strategies as backups
Review strategy performance regularly
Disable underperforming strategies
Avoid running too many strategies simultaneously on the same asset
Risk & Responsibility Best Practices
Never risk more than you can afford to lose
Do not rely solely on automated signals
Always perform your own due diligence
Use Comps as a decision-support tool, not financial advice
When to Adjust Your Approach
Consider refining your strategy when:
Market conditions change
Volatility increases or decreases significantly
Performance deviates from expectations
Drawdowns exceed your risk tolerance
Final Recommendations
Comps is most effective when used with:
Clear strategy logic
Disciplined risk management
Ongoing review and refinement
A balance between AI automation and human judgment
Last updated