1. Conceal what came later
A paper cover or chart replay stops at a chosen bar. The learner records timeframe, moving-average settings, and the filter state without seeing the eventual move. This does not reproduce live-market pressure, but it removes the easiest form of hindsight.
2. Make four marks
Each sheet receives a regime label, two pieces of visible evidence, an invalidation condition, and an alternative reading. The labels are deliberately short. If the explanation depends on a long story, the observation is probably not yet defined.
3. Reveal a measured segment
The chart advances by a fixed number of bars rather than jumping to a dramatic endpoint. The learner notes whether the rule changed state and whether that change followed the written conditions.
4. Audit language before outcome
Review begins with consistency: Were the same words used the same way? Was an exception invented after seeing later price? Only then do we discuss what the market did. The purpose is to strengthen observation, not celebrate a lucky label.
5. Keep the rejected sheets
Crossed-out definitions are part of the record. They show where a lookback was too sensitive for the intended timeframe, where a trend filter added unacceptable delay, or where “unclear” was missing from the classification set.
This method appears throughout the flagship workshop and practice circle. To discuss whether it fits your current level, send a training enquiry.