Three Ways In, Three Different Trades
Immediate, retest, or probe-and-add. Each buys something and pays for it somewhere else. Set your range and see the trade-off in the numbers rather than in the argument.
Read the capture and miss figures as structural, not measured. The share of the impulse each model captures and the share of setups each one misses follow from where in the sequence you commit — they are reasoning about the mechanism, not results from our sample. The numbers that are measured, on 489 attempts across 490 SPY sessions: 75% of attempts closed back inside the range, and the median best price within ten minutes was 0.72 range widths when the breakout held against 0.24 when it failed. The risk figure uses the boundary-zone width, because the invalidation belongs at the inner level rather than under the whole range.