Comprehensive guide to the most statistically reliable chart patterns based on Thomas Bulkowski's analysis of over 30,000 patterns across multiple decades
The Cup with Handle and Bump-and-Run Reversal Bottom dominate statistical rankings with sub-10% failure rates and average moves exceeding 50%. Thomas Bulkowski's analysis reveals that only a handful of formations consistently outperform, with the top patterns sharing common characteristics: extended formation periods, declining volume during consolidation, and clear supply/demand exhaustion signatures before breakout.
The Cup with Handle ranks as the most reliable pattern in Bulkowski's research, combining an extraordinarily low 5% break-even failure rate with a 54% average price rise post-breakout. This pattern, popularized by William O'Neil, reflects a complete cycle of institutional accumulation that creates powerful upside momentum upon completion.
The pattern forms over 1-6 months for the cup portion plus 1-4 weeks for the handle. Volume declines during the cup's descent, reaches its lowest point at the bottom, then rises as recovery begins. The handle shows declining volume—a critical confirmation of selling exhaustion. Breakouts require volume 50% above the 20-day average minimum.
The cup represents weak retail hands selling in panic while institutions quietly accumulate at discount prices. The rounded U-shape (versus sharp V-shape) indicates controlled, patient accumulation. The handle formation shakes out final nervous holders before breakout—Bulkowski warns that 47% of patterns experience substantial drops within 2 months of breakout, emphasizing the importance of proper handle formation.
The efficacy lies in the "Handle." After stocks recover from the bottom of the cup to near highs, latecomers rush to sell to break even. The handle represents a final "shakeout" of weak holders. When price refuses to drop significantly during the handle formation, it signals that supply is scarce, paving the way for a clean breakout. When volume explodes on breakout, it confirms forced buying from algorithms triggering momentum signals. The clearly defined stop-loss sits 2-3% below handle support, offering a tight risk/reward ratio of 1:3 or better.
Handles shorter than the median 22 days show superior performance, suggesting that prolonged consolidation in the handle phase may indicate weakening institutional conviction.
Discovered by Thomas Bulkowski in 1999, the Bump-and-Run Reversal Bottom (BARR Bottom) claims the #1 overall performance ranking for upward breakouts in bull markets and maintains #2 ranking even in bear markets—remarkable consistency across market conditions.
The pattern consists of three phases:
The pattern typically forms over several weeks to months.
The bump phase represents panic capitulation where selling accelerates beyond sustainable levels. This creates an oversold condition where supply exhaustion becomes extreme. The run phase begins when buyers recognize extreme undervaluation and aggressive accumulation overwhelms remaining sellers.
The accelerating decline in the bump phase creates conditions similar to a "selling climax"—the point where all weak holders have exited and only committed buyers remain. This creates asymmetric upside potential. The pattern's strength in both bull and bear markets suggests it captures fundamental exhaustion dynamics rather than trend-following momentum.
With 3,061 perfect trades analyzed—the largest sample size in Bulkowski's database—the Three Rising Valleys pattern provides the most statistically robust performance data. The pattern combines a solid 10% failure rate with 48% average rise.
The pattern appears frequently in markets and confirms when price closes above the highest peak in the pattern. Duration varies but typically spans several weeks. Volume trends downward in 64% of occurrences—interestingly, below-average breakout volume correlates with better performance, contradicting conventional wisdom about volume confirmation.
Each successively higher valley demonstrates buyers stepping in at progressively higher price levels—direct evidence of increasing demand. The three valleys represent three opportunities for sellers to regain control; their failure to push prices to new lows on each test confirms supply exhaustion.
In bear markets, the average rise is just over half the bull market performance, making this pattern significantly more effective in uptrending environments. Traders should weight positions accordingly based on broader market conditions.
The Rectangle Bottom combines a moderate 15% failure rate with 48% average rise and an exceptional 79% probability of reaching the measured price target—the highest target achievement rate among top patterns.
Price enters the pattern from the top (hence "bottom" designation) and consolidates between parallel support and resistance levels for a minimum of 30 days on daily charts. The horizontal boundaries should be defined by at least 2 significant swing highs and 2 swing lows. Volume remains relatively low during consolidation then sharply increases as breakout approaches.
The rectangle represents equilibrium where buying and selling pressure are roughly equal. Each bounce from support confirms demand absorption; each rejection from resistance confirms remaining supply. The pattern acts as a battlefield where one side gradually exhausts itself.
Partial rises (where price moves toward resistance but fails before reaching it) predict downward breakout 75% of the time. Conversely, partial declines predict upward breakout 77% of the time—providing valuable early warning of likely breakout direction.
The Rounding Bottom claims the distinction of lowest break-even failure rate among all chart patterns in Bulkowski's research, making it the premier choice for risk-averse traders prioritizing capital preservation.
The pattern forms over several months to several years, creating a gradual U-shaped bottom that reflects slow, steady accumulation. The extended timeframe filters out weak patterns and builds substantial supply absorption before reversal. Duration on the longer end correlates with more significant reversals.
The gradual nature of the rounding bottom indicates methodical institutional accumulation without triggering rapid price increases that would attract attention. Unlike sharp V-bottoms driven by panic buying, rounding bottoms represent controlled position-building over extended periods. This patient accumulation creates a solid foundation where virtually all available sellers have been absorbed before price advances.
The extended formation time means the pattern self-selects for strength—weak patterns fail before completion. The gradual slope prevents the pattern from attracting short-term traders who might create selling pressure, allowing institutions to build significant positions. This creates a high-probability, low-failure outcome once the pattern confirms.
The Double Bottom is among the most recognized reversal patterns, but Bulkowski's research reveals that not all variations perform equally. The Eve & Eve configuration—featuring two rounded, U-shaped troughs rather than sharp V-spike "Adam" bottoms—demonstrates statistically superior performance, ranking in the top tier for average post-breakout gains.
The pattern forms a distinctive "W" shape after an established downtrend, consisting of two rounded troughs (Eve & Eve) at approximately the same support level.
| Phase | Volume Characteristic |
|---|---|
| First Bottom (Left) | Higher volume—panic selling climax |
| Intervening Rise | Moderate, increasing |
| Second Bottom (Right) | Lower volume than first—selling exhaustion |
| Breakout | Volume spike required for confirmation |
The wider, more rounded structure of Eve bottoms allows for extended "flushing out" of weak holders. Unlike sharp V-bottoms that trigger quick bounces, the gradual U-shape gives nervous holders multiple opportunities to exit, ensuring only committed buyers remain when the breakout occurs.
| Parameter | Guideline |
|---|---|
| Stop-Loss Placement | Just below the lowest trough (2-3% buffer) |
| Initial Price Target | Neckline + (Neckline - Lowest Trough) |
| Risk/Reward Ratio | Typically 1:2.5 to 1:4 depending on pattern depth |
The High and Tight Flag stands alone as the #1 ranked chart pattern for average percentage gains in Bulkowski's Encyclopedia. Popularized by William O'Neil as "the strongest of patterns," this aggressive momentum formation captures extreme supply/demand imbalances that can produce gains of 200% or more.
Bulkowski's expanded 2,588-pattern study reveals critical nuances: nearly half of all High and Tight Flags fail to produce even a 10% gain. The original 0% failure rate came from a smaller, cherry-picked sample. Real-world trading requires acknowledging this substantial failure risk.
| Component | Requirement |
|---|---|
| Flagpole Rise | ≥90% (ideally 100%+ / a true double) |
| Maximum Duration | 42 trading days (~2 calendar months) |
| Flag Pullback | 10-25% of the move (ideally <20%) |
| Flag Duration | 3-5 weeks |
| Price Floor | Stock must trade above $1 |
| Flag Type | Success Rate | Average Gain |
|---|---|---|
| Tight Flag | 85% | 39% |
| Loose Flag | 45% | 9% |
Patterns following steep declines or V-shaped recoveries systematically underperform. A stock recovering from a 60-70% decline that doubles is simply retracing losses, not exhibiting genuine strength. True High and Tight Flags emerge from positions of strength, not recovery from devastation.
The Doji is perhaps the most recognized candlestick pattern in technical analysis—and also the most misunderstood. Popular trading literature portrays doji as powerful reversal signals, but Bulkowski's statistical research across 4.7 million candle lines reveals an inconvenient truth: most doji patterns perform no better than random chance.
"None of the doji candlesticks are what traders expect. To me, they mean nothing at all. Performance is about random or near random (around 50% to 59%). I do not believe that you can look at a doji and say, with certainty, that price will breakout upward or downward tomorrow." — Thomas Bulkowski, Encyclopedia of Candlestick Charts
| What Traders Believe | What Statistics Show |
|---|---|
| Doji signals reversal | 50-52% reversal rate (coin flip) |
| Doji predicts direction | Breakout direction is random |
| Doji at tops/bottoms is reliable | Slightly better than random at best |
| All doji are tradeable signals | Most provide no statistical edge |
| Doji Type | Reversal Rate | Performance Rank | Verdict |
|---|---|---|---|
| Standard Doji | ~50% | Mid-list | No edge |
| Dragonfly Doji | 50% | 98/103 | Ignore it |
| Gravestone Doji | 51% | 77/103 | Random |
| Long-Legged Doji | 51% | 37/103 | Slightly better |
| Northern Doji | 51% continuation | 83/103 | Random |
| Southern Doji | 52% | 78/103 | Near random |
Despite overall poor performance, narrow conditions may improve reliability:
Do not trade doji patterns as standalone signals. If you insist on incorporating doji into your analysis: use only as secondary confirmation within larger patterns, apply all optimization filters simultaneously, maintain realistic expectations (55-60% win rate maximum), and size positions for a pattern with minimal edge.
Instead of trading doji alone, consider:
Always use additional technical indicators for confirmation. Volume analysis is critical—look for higher volume on the breakout to validate the pattern's strength. Other indicators like RSI or MACD can provide additional confluence.
Even high-probability patterns can fail. Always define your stop-loss and take-profit levels. For reversal patterns like the double top/bottom, a common method is to set a profit target equal to the height of the pattern projected from the neckline.
A pattern that performed well in historical backtesting may not necessarily work in future market conditions. Continuously monitor and validate your strategy's performance, accounting for transaction costs and slippage.
Andrew Lo's landmark 2000 study analyzing 350 stocks over 34 years found that 7 of 10 patterns on NYSE/AMEX stocks showed statistically significant differences in return distributions. The Brock, Lakonishok & LeBaron study examining 89 years of DJIA data found strong support for technical strategies.
The top patterns share common characteristics: declining volume during formation (confirming selling exhaustion), extended formation periods (filtering weak signals), and clear supply/demand dynamics. Trading tall patterns outperforms short ones, buying near yearly lows reduces failure rates, and avoiding throwbacks improves results 97% of the time.