Best chart patterns for crypto: what 39,000+ measured setups show

Watch your chosen patterns with ChartScout

Once you have chosen a few patterns, ChartScout can monitor them without requiring you to keep every chart open. A Pattern Watcher follows one exchange, pair, timeframe and pattern. For example, a 15m SOL/USDT bull flag on Binance is one combination; it does not cover every SOL setup.

Watchers send alerts through the platform, Telegram, Discord and email, letting you review a detection away from your desk. The chart image shows the formation, while confidence and maturity describe aspects of its structure. Neither field is a probability of profit.

The Pattern Screener helps you discover candidates outside that configured watchlist. It lists published futures detections with a chart image and exchange link, but sends no notifications. Configure a watcher when you want personal alerts for a particular combination.

Coverage spans 1,000+ pairs with more than $1M in 24-hour volume across Binance, Bybit, KuCoin, MEXC and Hyperliquid. The threshold excludes thinner markets, although current spread and depth still need checking before a trade.

The plans define watcher capacity and timeframe access. The 7-day Pro trial requires a card; its screener has a 15-minute delay, while watcher alerts are not subject to that intentional delay.

ChartScout uses public market data without exchange API keys or custody, and it places no trades. It also has no candlestick charting terminal. Keep your charting workflow for checking the detection against current price, nearby levels and your own trading rules.

Best chart patterns for crypto depend on the metric

No pattern wins every comparison. Our three published studies contain 39,408 setups in total, each using a universe of 59 Binance USDT-margined perpetual markets. This adds their separate sample counts; it is not one pooled experiment or a count of independent trades.

For the fixed 5% endpoint, the overall counts are 1,208 of 6,013 eligible bull flag breakouts, 2,916 of 17,572 evaluable bear flag breakdowns, and 1,322 of 11,387 resolved cup and handle breakouts. The cup and handle study starts earlier in the process: 73.12% of its ready candidates broke upward. That breakout rate is not its target-hit rate.

Cup and handle's full-depth target differs from a fixed 5% move. A shallow cup can reach its full target without rising 5%, so its higher target-hit rate is not contradictory.

These are descriptive detection outcomes, not strategy win rates. Entries, stops and fees can change whether a trade makes money. The studies do not establish a universal winner or predict your next setup. Their sampling periods and formation definitions differ, so compare each result with its own method and denominator.

Why stock percentages do not transfer

The market label matters as much as the pattern name. For upward flags, Bulkowski's stock benchmark is 56% reaching a 5% move, against our crypto figures of 20.09% overall and 35.68% at 15m and above.

For downward flags, the comparison is 55% on stocks versus 16.59% overall and 44.49% at 15m and above. For cup and handle, Bulkowski reports 61% reaching target across 913 stock samples, versus 34.19% in our crypto data.

Three differences prevent direct transfer. Crypto trades continuously rather than within stock-market sessions. Our samples are heavy in intraday candles, which changes the scale of a fixed percentage move. The market universe is different: crypto perpetuals are not the same instruments as stocks.

Endpoint horizons also differ. Our observation rules stop measuring after defined reversal, invalidation or time conditions. These comparisons describe different samples; they do not prove which market is harder to trade.

Treat a published percentage as a result attached to its sample and method. Applying it to another market requires new measurement, rather than assuming a familiar pattern name makes the populations equivalent.

What makes a pattern worth considering

A useful pattern gives you a decision you can define before entry. Check four things:

• Invalidation: identify the price action that would disprove the setup before you risk money on it.
• Trigger: specify the event you require, such as a candle close beyond a named boundary.
• Target: measure the available move from your planned entry and compare it with the distance to invalidation.
• Liquidity: check whether the pair can support your entry and exit at the size you intend to trade.

A modest hit rate can still suit a plan when gains compensate for losses. Close invalidation can improve the potential reward relative to risk, but moving a stop closer does not create an edge. Use risk and reward math with realistic entries, exits and costs, not the pattern percentage alone.

Continuation patterns give you boundaries to watch

Flags are a practical starting point because their boundaries make the continuation idea easy to define. A bull flag follows a sharp rise with a compact, roughly parallel pullback. Its breakout trigger is above flag resistance, with structural invalidation below the flag low.

A bear flag reverses that picture: a sharp decline, an upward or sideways pause, a trigger below support and invalidation above the flag high.

Pennants replace the parallel pause with converging trendlines after a sharp move. Bull and bear versions favor continuation in the pole's direction; the opposite boundary defines invalidation.

Ascending triangles combine flat resistance with rising lows and favor an upward break. Descending triangles combine flat support with falling highs and favor a downward break. Define invalidation beyond the opposite side before entry, since either formation can break against its usual bias.

We have not published outcome studies for pennants or triangles. Use our pattern reference for identification, not its stock percentages as crypto forecasts.

Reversal patterns need confirmation, not anticipation

Double bottoms and tops are straightforward to define: two similar lows or highs separated by a bounce or pullback. A bottom favors an upward neckline break, invalidated below the lows; a top favors a downward break, invalidated above the highs.

Head and shoulders has three peaks, with the middle highest; its inverse has three troughs, with the middle lowest. The neckline break triggers the reversal idea. A conservative structural invalidation lies beyond the head; a tighter right-shoulder exit is a different trade rule.

Falling and rising wedges squeeze candles between converging, downward or upward boundaries. They favor bullish and bearish breaks respectively, but can also appear as continuation structures. Define invalidation below the wedge low for a bullish setup or above its high for a bearish one.

Cup and handle is a bullish base, often a continuation: a rounded recovery followed by a smaller pullback. The trigger is a close above the rim, invalidation below the handle low, and the target projects cup depth upward from the rim.

On fast crypto charts, an apparent reversal can keep developing after it looks complete. Wait for your defined confirmation rather than treating the outline alone as a finished pattern. Review fake breakouts. Apart from cup and handle, we have not published studies for these patterns.

Timeframe changes what the percentage means

Bull flags reached a 5% move in 10.69% of cases at 1m, 29.68% at 15m and 39.04% at 1h. These are individual timeframe results, not the pooled 15m-and-above figure.

A fixed 5% move is larger relative to a small 1m formation than to a larger hourly formation. The study measures follow-through across candles, not a move inside one candle. Part of the difference therefore comes from the threshold, alongside sample composition and stopping rules.

Consult the timeframe table in the bull flag study. Do not read higher percentages as proof that changing timeframe improves a strategy. Sparse higher-timeframe samples also limit comparisons.

Choose two or three patterns you can define consistently

For scalping on 1m to 5m charts, begin by logging bull and bear flags. Their boundaries make a compact practice set; the studies do not validate a scalping strategy or make 5% a suitable scalping target.

For day trading on 15m to 1h charts, compare those flags with cup and handle. You can evaluate continuation against a base using the same trigger, invalidation and target checklist.

For swing trading on 1h to 4h charts, study cup and handle with double bottoms and tops. The latter are learning choices because their levels are clear, not because we have measured superior crypto outcomes.

For two weeks, log every instance you would have taken, including failures. Record the pair, timeframe, formation, trigger, invalidation and target before the outcome is known. Add the subsequent outcome without changing those rules. Keep unresolved setups unresolved. Separate chart-target attainment from hypothetical net results, including costs. This exercise tests whether your definitions are consistent, not whether two weeks prove profitability.

Automation finds candidates; you still judge the trade

Software can watch more pairs than a person, but it cannot decide whether a setup fits your plan. In ChartScout, one watcher is one exchange, pair, timeframe and pattern combination. Watching the same pair for three patterns uses three watchers.

A forming pattern typically triggers at about 75 to 80% maturity, before the breakout completes. An alert is a reason to look, not a confirmed breakout. Check the detection chart against your trigger and invalidation before considering an order.

Keep the alert, your trade plan and the study result separate. Confidence and maturity describe the detected structure; the studies describe what happened to defined historical samples. Neither determines the outcome of your next trade.

Frequently asked questions