Order Block After a Sweep: The Highest-Win-Rate SMC Confluence

A plain-language guide with live examples and backtest data across BTC, ETH, SOL, and ZEC.

Jay Sharma · Founder, Botsfolio

Published August 18, 2026

What it is

An order block after a sweep is a compound Smart Money Concepts pattern that requires two events to occur in strict sequence:

1. A liquidity sweep of a prior swing high or low (the trap event). Price wicks past the swing, triggers stops resting beyond it, and closes back inside the range.
2. A valid order block forms in the reversal impulse that follows the sweep. The last opposite-color candle before the reversal impulse creates a fresh OB in the direction of the reversal.

The pattern captures the moment where stop-hunt liquidity has been engineered and the smart-money entry immediately follows. Both the trap thesis (sweep-and-reclaim signals a failed break) and the structural entry thesis (OB provides the specific reaction level) point at the same setup, from complementary angles.

The mechanic behind why this confluence outperforms its components is straightforward. A liquidity sweep alone tells you a trap occurred but not exactly where to enter. An order block alone tells you a structural entry level exists but not why price is likely to react from it right now. The two together specify both the timing (right after the sweep) and the price (the refined level inside the OB) with unusually strong evidence for both.

Two things people commonly get wrong about this pattern, which this guide will not:

• The sequence matters. Sweep must come first, then OB in the reversal. An OB followed by a sweep (in the wrong order) is a different pattern that historically underperforms.
• The 1-hour dominance in our data does not mean higher timeframes don't work. It means the strict sequence conditions produce enough clean setups to publish on 1H, while higher timeframes either produce fewer instances (too rare to backtest reliably) or resolve differently for reasons we are still investigating. The pattern's structural logic works on any timeframe; the empirical publish gate is what limits current coverage.

How to spot one on a chart

The pattern requires the strict two-event sequence in the correct order.

1. The prior swing

Start by identifying a clear swing high or low on the timeframe you are analyzing. For a bullish sequence: a well-defined swing low with visible stop-loss liquidity likely resting below. For a bearish sequence: a well-defined swing high with visible buy-side stops likely resting above.

2. The sweep event

Wait for a candle that wicks past the swing and closes back inside the range. This is the sweep-and-reclaim event covered in detail in our liquidity sweep article. The reclaim close is essential: a wick past that closes below (for a bullish sweep) is a sustained break, not a sweep.

3. The order block in the reversal

After the sweep, look for a reversal impulse that breaks structure in the direction opposite the sweep. The last opposite-color candle before the impulse (last bearish candle before a bullish break, or last bullish candle before a bearish break) forms a fresh order block. That candle's body and wick define the OB zone, and the wick cluster inside the zone defines the refined level.

4. The retest of the OB

The trigger for entry is when price returns to test the newly-formed OB after the initial impulse extends. Same as any OB setup, the retest reaction at the refined level is what confirms the pattern.

How Botsfolio detects and measures it

Every candle close, our analysis engine tracks both sweep events and order block formations. When we detect a valid sweep event immediately followed by a valid order block formation in the reversal impulse, we tag the OB with a "post-sweep" designation. This is a specific compound pattern separate from a standalone order block detection.

Three pieces of data are recorded:

• The sweep event (level swept, wick depth, reclaim strength)
• The order block (formation candle, refined level, invalidation)
• The sequence timing (how many candles between sweep and OB formation)

The reaction is tracked continuously through the OB's normal lifecycle: retested and held, in progress, or invalidated. Every outcome becomes a row in our backtest.

Two things worth naming:

• We only surface OB-after-sweep setups where the sweep-to-OB sequence occurs within 5 candles. Longer gaps break the causal link between the two events.
• Our cost model assumes 0.12 percent round-trip fees, already deducted from every expectancy R and annual gain figure.

Full methodology at our methodology page.

A live example on BTC right now

Here is a live order-block-after-sweep on BTC, drawn as it looked when it formed, alongside how the pattern has performed across timeframes. If nothing is currently active, the widget shows the most recent formed example.

Historical performance across coins and timeframes

The table below is aggregate performance of order-block-after-sweep setups across the coins we backfill, updated live from our backtest database. Book-early management means partial off at first target with the remainder trailed. Both directions combined.

Reading the table honestly, three observations:

The measured win rates for this pattern are unusually high, in the 85 to 97 percent range on the 1-hour timeframe across BTC, ETH, SOL, and ZEC. This is materially higher than either component pattern alone: standalone order blocks on 1H sit around 83 percent; standalone sweeps around 76 percent. The confluence effect is real and substantial.

Expectancy R is also higher than either component: +0.72R to +1.46R per instance depending on coin, compared to +0.55R to +0.69R for the individual patterns. This is not just about win rate; the average winner is meaningfully bigger when the setup meets the strict compound criteria.

The pattern currently only publishes on 1H timeframe. On higher timeframes (4H, 8H, 1D), the strict two-event sequence within a short window occurs less frequently, so sample sizes fall below our minimum-N threshold for publication. The pattern's structural logic works on any timeframe; only the sample sufficiency limits current coverage. As more historical data accumulates and more instances form on higher timeframes, we will surface them when sample crosses the threshold.

What tends to invalidate the confluence

Every backtested setup carries a reversal rate: the percentage of instances that reached +1R at some point, then finished at or below breakeven.

Three failure patterns account for most invalidations in our data:

Weak reclaim close on the sweep. The sweep must produce a genuine reclaim close, not a marginal close-back-inside. Sweeps that barely reclaim (close just above the swept low for a bullish sequence) often produce weaker OB setups that fail on retest.

Weak break-of-structure in the reversal impulse. The OB formation requires that the impulse following the sweep actually breaks recent structure. Impulses that fall short of breaking structure produce OBs that lack strong structural anchoring and underperform on retest.

Extended time between sweep and OB retest. Setups where price rallies far from the OB before retesting have lower retest hold rates than setups where price retests promptly. Our data shows the strongest retest reactions occur within 5 to 15 candles of OB formation.

How this confluence interacts with other patterns

The order-block-after-sweep pattern is itself a confluence. Additional confluence with other patterns provides further edge.

• OB-after-sweep plus fair value gap. When the reversal impulse that forms the OB also creates an FVG, the FVG and OB retest at the same time. Three-way confluence in a single setup.
• OB-after-sweep plus momentum divergence. When RSI or wave momentum divergence coincides with the sweep event, the reversal has both structural (sweep + OB) and momentum (divergence) confirmation.
• OB-after-sweep at a liquidity level. When the swept swing was itself a well-tested liquidity level, the reversal read is strongest because the sweep event captured genuinely clustered orders.

How reads of this concept commonly go wrong

Five patterns show up repeatedly when we look at how the confluence gets misread.

1. Skipping the strict sequence. The pattern requires sweep first, then OB in reversal. An OB followed by a sweep is a different setup with different base-rate outcomes. Order matters.
2. Accepting weak sweeps. Marginal reclaim closes on the sweep candle produce weaker OB setups. The sweep needs to be a clear, unambiguous reclaim event.
3. Waiting too long for the retest. Setups where price extends far from the OB before retesting have lower base rate. If price runs 3R or more from the OB without retesting, treating a later pullback as the same setup is over-extending the pattern's causal logic.
4. Assuming higher win rate means guaranteed win. 90 percent win rates in our backtest still mean 1 in 10 setups fail. Position sizing off "this always works" produces oversized risk relative to the residual 10 percent failure case.
5. Trading the pattern on sub-1H timeframes. We do not publish sub-1H OB-after-sweep setups because the compound sequence produces too much noise at lower timeframes. Traders attempting to spot the pattern on 5-minute or 15-minute charts will find many technically-qualifying but functionally-weak setups.

Frequently asked questions

Related concepts

Educational analysis, not financial advice. Past performance does not predict future results.