A trendline is easy to draw because two points define a line. That geometry does not make the line meaningful. The useful question is not “Can these points connect?” but “Has price treated this diagonal area as relevant to the current structure?”
Begin with two defensible anchors
Choose swings that were visible decisions in price, not tiny turns selected because they produce an attractive slope. On an ascending line, the second low should form after an observable advance from the first. If both anchors sit inside one small congestion, the line probably describes noise rather than the trend you intend to study.
Do not pierce candles merely to improve the fit without recording a rule. Some traders anchor to wicks, others to bodies or zones. Consistency matters more than pretending every market gives a perfect edge.
Treat the third interaction as evidence
A third respect, near the line and at a structurally sensible point, gives the diagonal more weight. It need not touch to the tick. Price is an auction, and a line on a chart has no physical force. Look at the reaction, candle closes, and relationship to nearby horizontal swings.
Repeatedly adjusting the line after every overshoot destroys the test. If the anchor rule changes whenever price disagrees, the line is protecting a story rather than examining one.
Check whether slope still matches the sequence
An extremely steep trendline may break while higher highs and higher lows remain intact. That break can indicate slowing momentum without proving a bearish structure. Conversely, a shallow line may survive even after the local trading opportunity has lost relevance.
Write two separate observations: what happened to the line, and what happened to swing structure. Combining them too early turns a useful warning into an unsupported reversal claim.
State invalidation before the next candle
Before revealing more chart history, write what would make the line irrelevant. It may be a decisive close through the area followed by a failed retest, or a structural lower low beyond a named swing. The exact standard depends on timeframe and method, but it must exist before hindsight arrives.
A good trendline does not forecast with certainty. It organises attention around a slope that price has respected and provides a clear place to notice change.