Some traders believe that curve fitting is the same as optimisation, thinking that tuning a strategy’s parameters to perform better on past data is always beneficial. However, curve fitting and optimisation are not the same. While optimisation is a necessary part of refining a trading strategy, curve fitting goes too far — tailoring a system so tightly to historical data that it loses the ability to perform in live, unpredictable markets.

The belief that curve fitting is the same as optimisation dangerously confuses healthy refinement with destructive overfitting.

Why Traders Confuse Curve Fitting With Optimisation

Several reasons explain why traders blur the distinction:

• Surface similarity: Both involve adjusting parameters to improve backtest results.
• Excitement about better metrics: Seeing win rates and profits increase feels rewarding, even if the improvements are misleading.
• Lack of forward testing: Traders often fail to test systems on new data, missing the warning signs of overfitting.
• Pressure to find ‘perfect’ systems: Emotional desire for certainty drives traders to tweak strategies until they look flawless — even if they are fragile.

Thus, what starts as simple optimisation can easily turn into dangerous curve fitting without clear discipline.

What Optimisation Actually Is

Optimisation, done properly, means:

• Refining sensible parameters: Adjusting reasonable values like stop-loss distance, moving average periods, or risk percentage based on logical, market-driven reasoning.
• Testing general robustness: Seeking parameter ranges that work reasonably well across different assets, timeframes, and market conditions.
• Keeping it simple: Reducing unnecessary complexity to avoid making the system dependent on specific past behaviours.
• Prioritising stability over performance: Choosing settings that offer consistent, realistic returns — not just the best historical figures.

True optimisation improves a system’s adaptability, not just its backtest statistics.

What Curve Fitting Is

Curve fitting happens when:

• Overfitting past data: The strategy is tuned so specifically to past price movements that it loses all generality.
• Chasing perfect results: Maximising backtest metrics like profit factor, win rate, or maximum drawdown at the expense of live viability.
• Overcomplicating rules: Adding unnecessary filters, conditions, and parameters that only worked historically but have no logical basis.
• Failure in unseen data: The system performs brilliantly on backtested data but collapses when forward tested or traded live.

Curve fitting creates beautiful backtests — and disastrous live trading results.

How to Avoid Curve Fitting While Optimising

To optimise properly without falling into curve fitting:

• Use out-of-sample testing: After backtesting and optimising on one dataset, test performance on different, unseen periods.
• Look for robustness, not perfection: Accept imperfect but consistent performance across varying conditions.
• Limit the number of variables: Fewer parameters mean less risk of fitting random noise instead of genuine patterns.
• Apply logical reasoning: Only adjust parameters when there is a strong market-based reason to do so, not just to improve metrics.
• Monitor performance consistency: A truly robust strategy should perform decently across different assets, sessions, and volatility environments.

This disciplined approach ensures your strategy is built on real edges — not random chance.

Examples of Curve Fitting Versus Proper Optimisation

• Curve fitting: A system only enters trades when a 17-period RSI crosses above 48.3 during the second half of the London session after a 1.4% drop — overly specific, unlikely to hold up live.
• Optimisation: Testing several reasonable RSI periods (like 14, 21, or 30) and choosing one that performs consistently across multiple years and assets — logical and sustainable.

Each example shows the clear line between useful refinement and harmful overfitting.

Conclusion

It is completely false to believe that curve fitting is the same as optimisation. Proper optimisation strengthens a trading strategy by making it more consistent and resilient, while curve fitting weakens it by chasing perfect but unreliable backtest results. Traders who focus on robustness, simplicity, and logic — rather than perfection — give themselves the best chance of success in real-world trading.

To learn how to build, optimise, and test professional trading strategies without falling into the trap of curve fitting, enrol in our expertly developed Trading Courses today.