Breakouts dominated while the sweet spot stayed selective
Backtest patterns, sweet spot performance, and missed-trade analysis for August 24 to August 28.
THE EDGE THIS WEEK
The Edge This Week
This week favored selectivity over volume. The scanner produced 504 setups, but only 100 became breakouts, which points to a tape where continuation existed in pockets rather than across the board.
The strongest pattern was concentration of edge in a narrow group of symbols and sectors, while weaker areas produced repeated false starts. That split matters because the backtest still showed a strong 64.2% average win rate across 2,610 trades, which suggests the edge remained intact at the symbol level even as broad scanner conversion weakened.
BY THE NUMBERS
By The Numbers
The headline tension this week was simple: strong backtest breadth across tracked symbols, but low scanner conversion in real-time setup flow. In practice, that usually means the edge was present only when pressure, structure, and follow-through aligned cleanly. Traders who leaned on raw setup count likely saw noise. Traders who filtered aggressively likely saw a better experience.
Scanner volume alone was not predictive. The useful pattern was quality concentration, not activity concentration.
SWEET SPOT REPORT
Sweet Spot Report
The sweet spot cohort underperformed its historical baseline by 13.1 percentage points this week. That is a meaningful drop, especially because this setup family is usually designed to capture cleaner continuation conditions rather than high-conflict environments.
The likely explanation is that low-pressure, short-duration setups did not receive enough immediate expansion after triggering. In weaker tape conditions, these patterns can look technically sound but fail to attract sustained participation. That aligns with the broader scanner result, where breakouts were scarce relative to total setup volume.
For traders reviewing see live setups in the scanner, the takeaway is not to abandon the sweet spot. It is to demand confirmation. When the weekly hit rate falls materially below the historical average, the pattern often needs stronger context, such as sector alignment, cleaner relative strength, or reduced overhead supply.
SYMBOL SPOTLIGHT
Symbol Spotlight
GDX stands out because it paired scale with efficiency. A 95.45% win rate over 22 trades and 0.9097 average R suggests not just isolated success, but repeated clean execution. That type of profile usually appears when a symbol offers consistent directional clarity and fewer mid-pattern reversals.
ISRG is notable for showing up in both the top performers and the most active scanner names. Its 92.31% win rate across 13 backtest trades, combined with five scanner setups this week, suggests a symbol that repeatedly presented structure the model could actually monetize. This is the kind of name worth monitoring closely in future weeks because recurrence plus quality often signals a durable behavior pattern.
ETN is the opposite case and may be the more important research lesson. It was one of the most active scanner symbols with five setups, yet it finished among the worst performers with a 33.33% win rate across nine trades and negative expectancy. Activity without edge is a trap. $ETN reinforces that repeated appearance in the scanner is not enough if follow-through is inconsistent.
The contrast between $GDX, $ISRG, and $ETN shows the difference between frequency and tradability. The useful question is not which symbols appeared most often, but which symbols converted structure into follow-through.
WHAT THE BOTS MISSED
What The Bots Missed
The cleanest interpretation is operational consistency. There were no missed trades, no missed runners, and no blocking filter that systematically kept valid opportunities out of the system. Traders can watch the bots in the Edge Lab, but this week's result suggests the issue was not missed execution. It was market selectivity.
That distinction matters. A week with zero misses but weak scanner conversion tells us the process captured what it was designed to capture, yet the available pool was simply less productive. In research terms, that shifts attention away from tooling and toward environment, symbol selection, and setup discrimination.
SECTOR HEAT MAP
Sector Heat Map
Breakouts concentrated most heavily in Technology with 23, followed by ETF with 17. Consumer-related groups combined for 18 breakouts if the consumer and Consumer categories are treated together, while Energy produced 7, Healthcare 6, and Industrials 5.
The first pattern is concentration in growth and index-linked exposure rather than broad uniform participation. Technology leading the breakout count suggests there were still pockets of momentum and institutional interest, but the low overall breakout rate implies that leadership was not deep enough to lift the full opportunity set.
The second pattern is category inconsistency in the sector feed itself, with ETF and etf, along with consumer and Consumer, listed separately. Even with that normalization issue, the broader conclusion does not change. Breakouts clustered in a few areas, while broad diversification across sectors was limited.
This helps explain why some symbols posted exceptional win rates while others repeatedly failed. In a concentrated market, the best results usually come from staying aligned with the active groups and avoiding weak sectors that are still generating setups but not producing sustained resolution.
RESEARCH NOTE
Research Note
This week highlights a useful distinction between symbol-level edge and scanner-level edge. The backtest average win rate of 64.2% across 189 symbols says the system can still identify profitable behavior over a broad sample. But the scanner breakout rate of 19.8% says the real-time environment was less forgiving, with many setups unable to complete the full breakout sequence.
For traders, the implication is that setup quality is conditional, not static. A pattern can remain valid in the archive while becoming harder to monetize in the present if the market stops rewarding early expansion. The practical response is to tighten context filters when scanner conversion falls below normal: prioritize symbols with strong internal consistency, favor sectors already proving they can produce breakouts, and be skeptical of names that are merely active without positive expectancy.
When broad setup conversion weakens, the edge does not disappear. It narrows. The job shifts from finding more setups to identifying the smaller subset that still has structural follow-through.