Understanding Confidence Scores
What the Confidence Score Tells You
Every PatternPilotAI analysis includes a confidence score, expressed as a percentage from 0 to 100. This score reflects how clearly and reliably the AI has identified a chart pattern. Understanding what it means helps you use it effectively in your trading decisions.
How the Confidence Score Is Calculated
The confidence score is derived from several factors:
Pattern Clarity
How closely the detected pattern matches its textbook definition. A cup and handle pattern with a smooth, rounded cup and a well-defined handle scores higher than one with an irregular or choppy formation.
Trend Alignment
Whether the broader market trend supports the pattern's implied direction. A bullish pattern found in an uptrend scores higher than the same pattern found in a downtrend, because trend-aligned patterns have historically higher success rates.
Risk-to-Reward Quality
The attractiveness of the potential trade setup. Patterns that offer a clear entry point, a logical stop-loss level, and a meaningful profit target score higher than setups where these levels are ambiguous or the risk-to-reward ratio is unfavorable.
Volume Confirmation
When volume data is visible in the chart, the AI checks whether volume behavior supports the pattern. For example, breakout patterns are more reliable when accompanied by increasing volume.
Supporting Technical Factors
Additional signals that strengthen or weaken the setup, such as alignment with support and resistance levels, candlestick confirmation patterns, or proximity to key moving average levels.
Interpreting Different Confidence Levels
High Confidence (75-100%)
The pattern is clear, well-formed, and supported by multiple confirming factors. These setups represent the AI's strongest identifications. You can treat these with higher conviction in your analysis process.
Moderate Confidence (50-74%)
The pattern is recognizable but may have some ambiguity, conflicting signals, or a less-than-ideal risk-to-reward setup. These are reasonable trade candidates but warrant closer personal analysis.
Low Confidence (25-49%)
The pattern is partially formed or has significant conflicting signals. Use these results as a starting point for further research rather than as standalone trade signals.
Very Low Confidence (0-24%)
The AI found little to no clear pattern, or the detected pattern is too ambiguous to be actionable. Consider that the chart may not have a tradeable setup at this time.
What the Confidence Score Is NOT
The confidence score does not predict:
- Whether the trade will be profitable
- The probability of the price reaching the target
- Market direction with certainty
It measures how clearly the AI identified the pattern and how well the setup aligns with factors that have historically correlated with pattern reliability. Even high-confidence patterns can fail, and low-confidence results can sometimes lead to successful trades.
How to Use Confidence Scores in Your Process
- Set a personal threshold - Many traders decide they will only act on analyses above a certain confidence level, such as 60% or 70%
- Combine with your own analysis - Use the confidence score as one input alongside your own chart reading, fundamental analysis, and market context
- Review the confidence breakdown - Check which factors are boosting or lowering the score to understand the strengths and weaknesses of the setup
- Track outcomes over time - Use the trade journal to log your trades and see how confidence scores correlate with your actual results
Was this article helpful?