Notes

• A chart needs at least 5 historical signals, the same minimum as the scanner posts. Every figure is measured from the close of the signal candle (candle 0, labeled “Signal”).
• Per-candle figures: Avg % Move, Median % Move and Win Rate are calculated separately at each candle after the signal (1–10), as if the position were held for exactly that many candles.
• Scanner posts use best-of-5 figures: the Win Rate and Median % Move in the posts credit each signal with its best close within candles 1–5. That is why they are higher than any single candle shown here.
• Compare with an ordinary candle, not with 50%: markets have tended to rise over time, so even after an ordinary candle, price was more often higher than lower a few candles later. Part of every bullish figure here is that long-term drift, and the same drift works against bearish signals.
• Bearish signals mark a stretched advance, not a reversal call. In this backtest, price was lower 5 candles later after about half or fewer of them. Their best-of-5 win rate in the scanner posts is higher, because it counts a win when any one of the next 5 closes is below the signal close.
• When a filter covers more than one combination (any “All” selection), each line combines them, weighted by how many signals each has. Avg % Move and Win Rate are exact. Median % Move is a weighted average of each combination’s median, not a true median of all the signals; select a single ticker, bias, level and timeframe to see an exact median. With Bias set to All, moves are shown in each signal’s direction: a fall counts as a positive move for a Bearish signal.
• The charts include every signal that matches the filters. They are not limited to the signals marked ◆ (higher-timeframe confluence) in the daily posts.
• These figures are backtested and hypothetical. They are context, not a recommendation or trade instruction.