The idea is reasonable: the last opposite candle before a strong move marks an area where large orders sat, and price may react when it returns. On higher timeframes I can see why traders like it.
On M1, my problems were:
- Too many of them. By mid-session every chart had eight or ten "valid" blocks. If everything is a level, nothing is.
- Definition drift. Is it the body or the wick? The last candle or the last two? I noticed I picked the version that fit the trade I already wanted.
- Spread swallows the reaction. Many M1 reactions were 2–4 pips. After costs, that is barely anything.
So my current view is that on M1 it is mostly clutter, unless you have a strict mechanical definition and you've tested it.
If you use them on low timeframes, how do you define them, and did you test that definition or adopt it from a video?