Descending and ascending channels fail in student books for a mundane reason: the lower rail is measured from one pair of swings, the upper from another. The eye fills in “parallel,” but a ruler shows a wedge.
Pick the primary line first
Build the trendline that has the clearer reactions. Only after that line survives a touch checklist do you project a parallel copy. The distance should come from a single opposing swing that reacts against the primary — not from an average of several peaks.
Horizontal channels are not exempt
Range boxes look simple, yet students still mix a weekly high with an intraday low. Keep both rails on the same timeframe you intend to trade decisions against.
What we ban in clinic
Freehand “roughly parallel” sketches without a measured copy. If you lack a drafting triangle in the room, use the charting package’s parallel tool once — then print and verify with a physical ruler. The tactile check catches lazy anchors surprisingly often.