StratBase.aiStratBase.ai
DashboardCreate BacktestMy BacktestsCatalogBlogNewsToolsHelp

Products

  • Researcher Dashboard
  • Create Backtest
  • My Backtests
  • Catalog
  • Blog
  • News

Alerts

  • Calendar
  • OI Screener
  • Funding Rate
  • REKT
  • Pump/Dump

Company

  • About Us
  • Pricing
  • Affiliate
  • AI Widget
  • Contact

Legal

  • Privacy
  • Terms
  • Refund Policy

Support

  • Help Center
  • Reviews
StratBase.aiStratBase.ai

Think it. Test it.

StratBase.ai does not provide financial advice or trading recommendations. AI only formalizes user ideas into testable strategy configurations for research purposes. Past backtesting performance does not guarantee future results. All trading decisions and associated risks are the sole responsibility of the user. This platform is not a broker and does not facilitate real trading.

© 2026 StratBase.ai · AI-powered strategy research and backtesting platform

support@stratbase.ai
Help Center/Backtest Results/Clone Protection System

Clone Protection System

📋Backtest Results
📌

Clone Protection System

StratBase.ai employs a 3-phase clone protection system to detect copied strategies, protect original creators, and maintain catalog integrity.

📌

Phase 1: Config Hash (Instant)

At backtest creation, a SHA-256 hash is computed from the strategy's defining fields (indicators, conditions, parameters). Identical configurations produce the same hash.

  • Detection: Exact same configuration → instant flag
  • Purpose: Prevent exact clones from appearing in the catalog
  • Catalog dedup: Only one card per (hash + instrument) pair
📌

Phase 2: Equity Curve Fingerprint (Automatic)

After completion (if trade count >= 10), the equity curve is analyzed:

  1. Normalize equity to percentage of deposit
  2. Quantize to hourly intervals
  3. Hash the quantized values (SHA-256)
  4. Compare against all existing fingerprints

Fast Path

If the equity hash matches exactly → instant clone verdict.

Slow Path

For candidates on the same symbol with overlapping periods (>= 7 days):

  • Calculate Pearson correlation on clean windows (periods without open positions)
  • Clone threshold: r >= 0.93
  • Minimum 48 comparison points (2 days of hourly data)

Result

Confirmed clones are automatically archived (hidden from catalog and profiles).

📌

Phase 3: Signal Comparison Report (On Request)

Users can report suspected copies via the Report Copy button:

  1. System compares trade entries and exits with ±1 candle tolerance
  2. Calculates signal match rate (percentage of matching trades)
  3. Assigns verdict:

| Match Rate | Verdict | Action | |-----------|---------|--------| | >= 95% | Auto Clone | Archive backtest, apply strike | | 80–95% | Suspicious | Marked for manual review | | < 80% | Not Clone | No action |

📌

Strike System

Repeated clone violations result in progressive catalog bans:

| Strike | Ban Duration | |--------|-------------| | 1st | 30 days | | 2nd | 90 days | | 3rd+ | Permanent |

During a ban, the user's strategies are hidden from the public catalog.

📌

How It Protects You

As a Creator

  • Your original strategies are protected from unauthorized copying
  • Clones are automatically detected and removed from the catalog
  • You can report suspected copies for deeper analysis

As a User

  • The catalog shows only original strategies (deduped)
  • Cloned content is automatically filtered out
  • Quality and originality are maintained
❓

FAQ

Q: What if I independently create a similar strategy? A: The system compares equity curves, not just indicators. Two strategies with similar logic but different parameters will produce different equity curves. Only nearly identical results trigger clone detection.

Q: Can I clone my own strategy? A: Yes. Self-cloning (same author) is allowed and won't trigger protection. It's a valid workflow for iterating on strategies.

Q: How do I report a suspected copy? A: Open the suspected clone, go to the results page, and click Report Copy. The system will run a signal comparison and provide a verdict.

🔗

Related Articles

  • Exporting Strategies & Results
  • Privacy Levels
  • Publishing a Backtest