FRACTAL AI AGENT Docs ← Terminal

Bar Pattern Self-Similarity PRO

Find where this exact bar sequence happened before.

Bar Pattern Match studies the recent sequence of candles and searches the loaded history for similar sequences. It helps you judge whether the current price action is a recurring setup or just a shape that looks familiar.

What this tool is for

Use it when the order and shape of recent candles matter—for example, a compression, rejection sequence or repeated impulse-and-pullback. The tool compares normalized shapes, groups related matches and shows whether the resemblance appears on macro, mid or micro scales.

How to read the result

OutputWhat it means
Dominant patternThe tool’s description of the most important repeated candle sequence.
Signal & confidenceThe directional implication and how clearly it is supported.
Self-similarity scoreHow strongly the current sequence resembles patterns found elsewhere in the loaded data.
Scale levelsWhether repetition is visible across macro, mid or micro windows, including each scale’s strength.
Pattern clustersGroups of similar sequences, their chart locations and individual similarity percentages.
Fractal dimensionA roughness measure: higher values generally indicate more complex, jagged price action.

Practical workflow

  1. Run it after a distinct sequence has formed, not in the middle of one unfinished candle.
  2. Review the similarity score, then inspect where each cluster occurred on the chart.
  3. Compare context as well as shape: trend, volatility, nearby support/resistance and volume can change the outcome.
  4. Use a repeated pattern to form a hypothesis, then define confirmation and invalidation from current structure.
Difference from Pattern Matcher: Fractal Pattern Matcher produces a broader trade-style analysis with historical analogues, levels and a forward outlook. Bar Pattern Match focuses on repeated candle sequences and their similarity across scales.
Similar is not identical: normalized candle shapes can match while the surrounding market context differs. Always inspect the original locations before treating a cluster as useful evidence.