Moving Average

A moving average replaces each point with the average of several points around it, so short-term swings are smoothed and the underlying trend emerges.

A moving average is the main tool for separating trend from noise: by averaging over a sliding window (seven days, say), random peaks and troughs cancel out and the underlying line calms. Choosing the window length is a trade-off: a short window reacts to fast trends but keeps more noise; a long window is smoother but shows a real change with a lag. A subtle point is that a moving average is inherently lagging; because it is built from past data, it turns late at inflection points, so it is a poor tool for early warning. A seven-day window is popular on the web because it covers exactly one weekly cycle and removes the day-of-week effect.