Do market structure indicators draw too much?

 
I have been testing a few market structure indicators lately and noticed something .
At first it feels useful when the chart shows every swing BOS and CHoCH but after scrolling around it starts to look noisy .
Maybe it is just me  but I feel the harder part is not finding structure . It is deciding what should not be drawn.
For those who use or build these tools  do you prefer seeing everything  or only the cleaner confirmed structure  ??
 

For discretionary trading, indicators should prioritize clarity over completeness.

I prefer extracting meaningful structure rather than displaying every micro move.

For my EAs, however, I keep the data closer to raw, because what looks like noise can actually be valid input.

 

Coming at this from the builder side, since I develop market structure tools: "what should not be drawn" is the right question, and in my experience the answer is less about detection filters and more about object lifecycle.

Three things that made the biggest difference for me:

1. Separate confirmed from pending. A BOS/CHoCH should only be labeled once the breaking candle has closed. Anything before the close is a pending state - if you show it at all, make it visually distinct (dotted, faded) and remove it when the break fails. Most "noisy" charts are really mixing anticipated and confirmed events in the same visual style.

2. Automatic cleanup. Mitigated order blocks and invalidated structure should leave the chart on their own after N bars. Users almost never clean charts manually - automating stale-object removal reduced "too much drawing" feedback more than any detection change I made.

3. Default to confirmed-only, make everything else opt-in. The trader who wants every internal swing will find the setting; the majority never should.

The ATR-based minimum swing size mentioned above is worth it too - I would just add that the multiplier has to scale per timeframe. One fixed value that looks clean on H1 will either starve M5 or clutter D1.