What I look for:
- Three or more touches of the same level within 20–30 minutes.
- Each push smaller than the last, measured in pips.
- Tick activity staying high while price goes nowhere.
- Spread normal. If the spread is wide, it's not absorption, it's just a thin book.
I don't have a sample big enough to quote a hit rate, so treat this as a hypothesis. The structural appeal is the risk: the stop sits just beyond the absorbed level, so a loser is small compared with a reversal that runs back through the range. To test it, log every stall that meets all four criteria, not only the ones that reversed, and score the 08:00 to 08:30 move afterwards.
I'd like to collect more examples. If you have a screenshot of a pre-London stall that either reversed or broke cleanly, please post it with the pair, date and time. Failures are just as useful as successes.