Analyzer

Analyzer detects 23 price and candlestick pattern types on Bitcoin. Binance and Bybit retain separate data and results. Up to 40 price and 40 candlestick findings are displayed, prioritizing scenarios still under observation. Each timeframe uses up to the last 2,000 available closed candles; zooming does not change the analysis. The actual evaluated range is shown next to the chart.

Price highs and lows are evaluated at three scales. A pivot is confirmed by a closed reversal of at least one ATR, or by 3, 6 or 12 subsequent candles without exceeding the extreme. Triangle, wedge, rectangle and channel boundaries can emerge from two highs and two lows. A dashed shape awaits a breakout; its chart marker shows when the finding became available, rather than backdating it to the last peak. Tolerances adapt to volatility using Wilder ATR 14, initialized with the mean of the first 14 true ranges and continued with Wilder smoothing. Source gaps split analysis into consecutive segments; missing candles are not fabricated.

Conditions met describe pattern recognition, not the probability of a future move. A measured target comes from the geometry of a confirmed price pattern. The actual confirmation close and remaining distance are shown. If price touched the target after the final anchor but before or during the confirmation candle, the finding is late and is not counted as a newly predicted target. Observation lasts twice the pattern length from confirmation; this is not a universal pattern lifetime. If one candle hits both target and invalidation, the order is unknown because OHLC data alone cannot establish it. Candlestick patterns are not assigned price targets. Main findings are selected by span, geometry, recency, confirmation and volume. Overlaps are grouped; ranking is not a success probability. Trend context and price zones use only the latest consecutive segment from the same market. A zone requires at least two price reactions and exists independently of a particular pattern outcome.

The combined outlook evaluates nine standard timeframes from the same exchange. It separates short-term movement, the broader trend and long-term structure. Each horizon has one reference timeframe; other frames can challenge its conclusion, but are not added as independent confirmations. Second-level data do not contribute to the combined directional assessment.

Optional AI analysis explains the same verified evidence. The application supplies price levels from the source snapshot. Generated text is not a live prediction and is marked as older when its evidence changes. AI does not search news or replace pattern calculations.