Research Note 07

BOS: A Structural Scanner, Not a Predictor

BOS is a standalone structural-scanning lab, built alongside Klines rather than inside it, and the note frames its own standard plainly: a scanner, not an oracle. The core mechanism identifies confirmed swing highs and lows in BTC price action, records the first causal break of that structure, and tracks the resulting impulse until an opposing structural break occurs. The same detection logic runs independently across three timeframes — one hour, four hours, and one day — with no timeframe voting between them and no forward-looking information used at any stage.

Coverage and the excluded faster signal

Tested against 134 BTC waves of 8% or larger from 2022 through 2026, the one-hour layer covered 93% of those waves, meaning the great majority of major moves in this window were captured by a break-of-structure signal on the fastest timeframe tracked. A faster, thirty-minute candidate signal was tested separately as a way to lead the one-hour confirmation earlier: it did lead by a median of fifteen hours, but carried a 44.8% false-positive rate, and was excluded from the core system on that basis rather than kept as a faster but noisier alternative.

The negative result

The negative result is presented as the more important finding of the two: prior candle behavior did not predict the direction of a subsequent break of structure, and divergence between timeframes did not hold up as a usable predictor either, once tested directly rather than assumed. BOS is described as useful specifically as a mechanical description of structure as it forms, not as an autonomous entry signal in its own right.

What remains open

Two threads remain explicitly open rather than resolved. Path continuity between confirmed breaks has not been established as a reliable property. A specific sequential effect — a break of structure followed by one-hour confirmation followed by four-hour alignment — shows a small effect that has not yet been validated to the same standard as the rest of the system. A separate options and gamma overlay is being tested independently of the core structural scanner, with a decision point on that overlay fixed for the end of September 2026, disclosed in advance rather than left open-ended.

As with Klines, no profitability claim is made for BOS. It is presented as a structural description tool whose false-positive behavior and prediction limits have been measured and stated directly, with the same discipline the rest of the series applies to published third-party claims applied here to the project's own system.

Note: this note's source lives on the site itself (/labs/bos/), not in the searching-for-edge repository.