Single Bar Reversals: Location Filtering

A hammer. A shooting star. A doji. These single-bar reversals are the most taught and most over-traded candlestick patterns in technical analysis. Their reliability is modest in raw form — often not far above coin-flip — but with proper location filtering, the same patterns become genuinely actionable. Location is not a refinement of single-bar analysis; it is the dominant variable.

Why Location Dominates

A hammer forms whenever price dips intrabar and recovers. In a flat range, this happens constantly and means nothing. At the bottom of a 10-bar decline, resting on daily support, with volume expansion — the same hammer is a high-probability reversal. The candlestick is identical in both cases. Only the location differs.

This is why catalog-based pattern trading fails. A trader who enters every hammer trades the same signal in high-quality and low-quality contexts, and the low-quality contexts dominate the sample, dragging expectancy to break-even or below.

The Location Filter

A robust location filter has multiple layers:

1. Prior trend extension — the pattern must form after a meaningful directional move. Define "meaningful" quantitatively: at least 1.5× ATR over the prior 5–10 bars, or a clear higher-higher-high sequence in a decline (for a long reversal).

2. Structural level — the bar should form at a prior swing high or low, a daily or weekly pivot, a moving average, a Fibonacci level, or a known supply/demand zone. Mid-air reversals are discarded.

3. Timeframe alignment — a hammer on H1 is far more reliable when the H4 or daily structure is at support. Lower-timeframe patterns in the direction of higher-timeframe trends carry weight; those against them carry little.

4. Volume confirmation — the reversal wick should form on volume above the recent average. A long wick on declining volume suggests the spike was thin-tape, not genuine rejection.

Worked Example

Consider a shooting star (bearish reversal) candidate on the H1 chart:

• Scenario A: Forms after a 3-bar rally, in the middle of a daily range, on average volume. Filter rejects. The pattern is noise.
• Scenario B: Forms after an 8-bar rally that extended into daily R1 resistance, with a long upper wick and volume 1.8× the prior average. Filter accepts. This is a high-probability short.

The candlestick shape is the same. The filter is the difference.

Common Location Errors

• Trading every hammer at every round number — round numbers are not all support. A round number that aligns with prior structure matters; a random one does not.
• Treating moving averages as automatic support — a 20-EMA touch is not a reversal signal unless accompanied by a confirming candle and prior trend extension.
• Counting trend as "extended" too early — a 2-bar move is not extended. Real extension requires visible commitment.
• Ignoring the higher timeframe — a daily resistance level overrides any H1 reversal signal beneath it.

The Patience Cost

Location filtering eliminates the majority of single-bar signals. A trader scanning for hammers might find a dozen a day across instruments; the filter reduces this to one or two quality setups, or none. This is the point. Frequency is the enemy of expectancy in pattern trading.

The Honest Read

Single-bar reversals are not edge in themselves. They are the final confirmation — the trigger — in a setup built on location and context. The disciplined trader defines the location first, then waits for the candlestick to confirm. Reverse the order, and the pattern becomes a trap. Single-bar patterns filtered by location become a defined-risk entry tool; unfiltered, they are a coin flip with extra steps.