ODB Microstructure Forensics

Systematic methodology for investigating Open Deviation Bar anomalies by tracing from ClickHouse cache back to raw Parquet trade data. Distinguishes algorithm correctness issues from market microstructure phenomena (liquidation cascades, order book sweeps, matching engine batch effects).

Self-Evolving Skill: This skill improves through use. If instructions are wrong, parameters drifted, or a workaround was needed — fix this file immediately, don't defer. Only update for real, reproducible issues.

When to Use

• ODB bars appear visually larger (taller or wider) than neighbors at the same threshold
• Bars have zero or near-zero duration with extreme price range
• abs_dev_dbps exceeds the threshold (e.g., 22 dbps on a 100 dbps bar)
• Clusters of micro-bars appear at a single timestamp
• Diagnosing whether anomalies are data bugs vs market microstructure
• Investigating specific time windows flagged by the Flowsurface chart UI

Data Sources