Staring at a one-minute chart for hours trains the eye to find shapes: double tops, flags, wedges, head and shoulders. The human brain is a pattern-finding machine, and it works overtime when the data is random.
On short timeframes, price movement is dominated by noise. Every bar is a mix of real information and random fluctuation, with the random part often larger. Shapes that look meaningful are formed by chance far more often than by intent.
A test shows it clearly. Generate a random walk, a chart of purely random price changes, and show it alongside real charts. Most people can't tell which is which, and many will happily find "setups" in the random one.
That doesn't mean all patterns are illusions. Some reflect real order flow, such as sweeps and absorptions. The challenge is separating those from the product of imagination.
Statistical testing helps: define the pattern precisely, find all occurrences in history, and measure what happened next. Many cherished patterns show no edge.
Patterns with a clear mechanism, such as a stop sweep, tend to survive. Those without one tend not to.
Which favorite pattern lost its edge once tested on a large sample?
Advertisement
Seeing patterns in noise
Recommended broker for automated trading & scalping
