Raw Data
~2.4B rows/day raw feed, reduced into study-usable sessions after parsing and normalization.
Function Walkthrough
How a study is built.
A step-by-step look at how raw intraday data becomes a conditional probability study. Example: LF → IB High Break.
1
Raw Data
Ingest tick, 1-second, or bar data for all sessions.
2
Session Segmentation
Split data by session and timezone rules.
3
Event Detection
Find the first break satisfying the event definition.
4
Rule Evaluation
Apply context and qualifying filters.
5
Outcome Measurement
Measure target, stop, and excursion behavior.
6
Aggregation
Compile all events into summary statistics.
7
Final Probability
Publish the study with counts and limitations.
Example Study
LF → IB High Break → RTH +10 pts
Condition: first break above opening range high during a Low-First session.
Outcome: price trades at least +10.0 points above IB High at any time during RTH.
Study Summary
| Probability | 63.4% |
| Sample Size | 742 days |
| Expectancy | +1.42 pts |
| Median MFE | +11.8 pts |
| Pullback Rate (≥ 10 pts) | 54.2% |
Session Segmentation
Trading day segmented into LF, RTH, and ETH windows based on a fixed exchange calendar.
Event Detection
Identify instances where the first valid break above the IB High occurs within the LF session.
Rule Evaluation
Require the event to occur within the study window and reject duplicates or invalid timing.
Outcome Measurement
Track whether price reaches +10 pts above IB High before the RTH session ends.
Aggregation
Aggregate across all valid sessions and compute probability, distribution, and R-results.
Integrity Checks
| Check | Purpose | Status |
|---|---|---|
| Duplicate Removal | Remove exact duplicate timestamps and prices. | Pass |
| Timestamp Ordering | Ensure strict chronological order within each instrument. | Pass |
| Lookahead Protection | All rules use only data available at event time. | Pass |
| Session Boundary Validation | Verify session windows align with exchange calendar. | Pass |
| Sample Threshold Check | Ensure minimum event count for reliability. | 742 days ≥ 200 |
From Study to Insight
Raw Outputs
Event logs, outcomes, MAE/MFE paths, distributions.
Aggregation Engine
Compute probabilities, confidence intervals, and robustness metrics.
Result Summary
Present the concise probability, expectancy, and context.
Decision Context
Compare across studies, sessions, and regimes.
Related Studies