How Chart Pattern Win Rate Statistics Are Calculated

The Numbers Behind the Patterns

Every trader wants to know: "What is this pattern's win rate?" But few ask the harder question: "How was that win rate calculated, and should I trust it?"

Chart pattern win rate statistics are only as good as their methodology. A win rate calculated over 50 samples is not the same as one calculated over 5,000 samples. A win rate that counts any move in the right direction as a "win" is not the same as one that requires the full measured move target to be reached. Understanding these nuances separates traders who use data wisely from those who are fooled by statistics.

Our Methodology: How Win Rates Are Calculated

1. Pattern Detection and Logging

Every time our detection engine identifies a chart pattern, it logs the following:

• Pattern type (e.g., cup and handle, head and shoulders)
• Ticker symbol and asset class (stock, ETF, crypto)
• Detection date and timeframe (daily, weekly, 4-hour)
• Key price levels: breakout/breakdown point, measured move target, stop-loss level
• Confidence score at the time of detection

2. Outcome Tracking

After a pattern is detected and confirmed (price breaks out or breaks down), we track what happens next:

• Win: Price reaches the measured move target before hitting the stop-loss level
• Loss: Price hits the stop-loss level before reaching the target
• Partial win: Price moves 50%+ toward the target but reverses before reaching it
• Timeout: Price neither reaches the target nor the stop within the tracking window (typically 2x the pattern's formation duration)

3. Win Rate Calculation

The primary win rate is calculated as:

Win Rate = (Wins) / (Wins + Losses) x 100

Timeouts are excluded from the primary calculation but reported separately. Partial wins are reported as a separate metric — they show the pattern's "directional accuracy" even when the full target is not reached.

4. Sample Size Requirements

We do not publish a win rate for any pattern/timeframe combination with fewer than 100 completed outcomes. Small sample sizes produce unreliable statistics. A pattern that "wins" 90% of the time over 10 samples is not meaningful — over 1,000 samples, that number might be 60%.

Our database currently contains over 50,000 tracked pattern outcomes across all pattern types and timeframes.

What Affects Win Rates

Market Conditions

Win rates vary significantly based on overall market conditions:

• Bull market: Bullish patterns (cup and handle, bull flags, ascending triangles) have higher win rates. Bearish patterns (head and shoulders, descending triangles) have lower win rates.
• Bear market: The reverse — bearish patterns outperform, bullish patterns underperform.
• Sideways market: Most patterns underperform their historical averages because breakouts lack follow-through.

Our win rate data is segmented by market condition so you can see how patterns perform in the current environment, not just historically.

Timeframe

Pattern win rates differ by timeframe:

• Weekly chart patterns: Highest win rates (~70-75% for the best patterns). Larger timeframes filter out noise.
• Daily chart patterns: Moderate win rates (~60-70%). The most commonly traded timeframe.
• 4-hour chart patterns: Lower win rates (~55-65%). More noise, more false breakouts.
• 1-hour and below: Win rates drop further. Intraday patterns are more affected by randomness.

Confidence Score

As discussed in our confidence score article, higher confidence scores correlate with higher win rates. An 80+ confidence bull flag wins ~75% of the time; a 40-score bull flag wins ~45%.

Volume Confirmation

Patterns that break out or break down on above-average volume have significantly higher win rates than those on light volume. Our data shows a 10-15% improvement in completion rates when volume exceeds the 50-day average at the breakout.

Published Win Rates for Key Patterns

Based on our database of 50,000+ outcomes (daily chart, all market conditions combined):

• Ascending triangle breakout: ~73%
• Inverse head and shoulders: ~71%
• Double bottom: ~70%
• Bull flag: ~67%
• Cup and handle: ~65%
• Descending triangle breakdown: ~64%
• Symmetrical triangle (trend direction): ~60%
• Head and shoulders (standard): ~68%
• Rising wedge breakdown: ~65%
• Falling wedge breakout: ~68%

These numbers represent the percentage of patterns that reach their measured move target after a confirmed breakout/breakdown. Your individual results will vary based on entry timing, stop placement, and trade management.

Why Win Rate Alone Is Misleading

A 65% win rate does not tell you whether a pattern is profitable. Profitability depends on the relationship between win rate and risk-to-reward ratio.

Consider two scenarios:

• Pattern A: 70% win rate, average win = 1R, average loss = 1R. Expectancy = +0.40R per trade. Profitable.
• Pattern B: 50% win rate, average win = 3R, average loss = 1R. Expectancy = +1.00R per trade. More profitable, despite a lower win rate.

Our platform reports both win rates and average risk-to-reward ratios for each pattern type, giving you a complete picture of expected profitability.

How to Use Win Rates in Your Trading

1. Set Realistic Expectations

No pattern wins 100% of the time. Even the best patterns fail 25-30%. Plan for losses.

2. Weight Your Position Sizes

Use higher confidence scores (which correlate with higher win rates) to adjust position size. Full size on A-grade setups, reduced size on B-grade.

3. Track Your Own Statistics

Your personal win rates will differ from the published averages because of your entry timing, stop management, and trade selection. Track your results and compare them to our baseline.

4. Focus on Expectancy, Not Win Rate

A 55% win rate with 3:1 reward-to-risk is more profitable than a 75% win rate with 1:1.

View Win Rates on TradingPatterns.io

Every pattern detection on our platform includes the historical win rate for that specific pattern type and timeframe. Our track record page shows how past detections have performed.

Explore our pattern win rates — free accounts can view win rates for the three free-tier patterns.