Skip to main content
A study measures how often something happened in the past and what followed. Ask in words, name the market and the window, and the Agent runs it on stored history and shows the sample size.

Studies you can run

Read a study properly

  • Look at the sample size first. A pattern seen eleven times is not a pattern.
  • Look at the spread, not just the average: the typical move against you matters as much as the typical move for you.
  • Ask for a different window to see whether it holds outside the stretch you looked at.
  • Turn a study into a rule with backtest this. See Backtests.