Overfitting and Curve Fitting

With enough tweaking, any set of rules can look brilliant on old data. Here is how overfitting happens, why testing many ideas guarantees some lucky winners and how to spot a curve-fit system.

Fitting noise, not signal

Overfitting happens when rules are tuned so closely to past data that they capture random noise instead of a real pattern. Each extra condition, such as a specific moving-average length, a day-of-week filter or a special exception, makes past results look better and future results less reliable.

Unlock all of Investing School

The rest of this lesson, its chart examples and quiz are part of Investing School. Start free with Stock Market Fundamentals and the first lesson in every category.