> ## Documentation Index > Fetch the complete documentation index at: https://docs.trinigence.com/llms.txt > Use this file to discover all available pages before exploring further. # Parameter Ranges > How to choose safe parameter ranges and avoid curve-fitting. ## Why ranges matter Optimization is only useful if the **range you search makes sense**. If the range is too wide, you’re effectively testing different strategies. If it’s too narrow, you can’t see whether the logic is stable. A good range should: * preserve the *intent* of the indicator * reflect the timeframe and holding period you care about * reveal whether results are stable or fragile *** ## How to choose a sane range Start from a **reasonable baseline** and expand only if the behavior remains consistent. Practical guidance: * Use common defaults as a starting point (e.g., RSI 14, ATR 14, EMA 20/50). * Keep ranges proportional to your timeframe and expected trade duration. * Avoid extremes that change the indicator’s meaning (very short or very long periods). * If the best result sits on the **edge of your range**, it’s a sign the range is wrong or the strategy is fragile. *** ## Robust vs fragile parameters A robust strategy **tolerates small changes** without collapsing. A fragile one only works at a single, narrow setting. Signals of robustness: * multiple adjacent values perform similarly * trades and drawdowns remain consistent * behavior is explainable, not random *** ## Quick checklist Before you optimize, confirm: * the baseline strategy works in principle * the range matches the timeframe and intent * you can explain why each parameter matters *** ## What to read next