Scoring System
The initial implementation uses deterministic scores. Scores are research triage aids, not trade signals.
Continuation Quality Score
continuation_quality_score ranges from 0 to 100.
| Component | Intent |
|---|---|
| Catalyst quality | Durable revaluation catalysts score higher than no-news pops |
| Liquidity | Higher dollar volume improves execution realism and institutional relevance |
| Volume shock | Volume expansion indicates unusual attention and participation |
| Close quality | Strong closes receive higher continuation score for upside events |
| Setup context | Base breakout and gap-and-go patterns score higher than weak or failed moves |
| 52-week context | Upside events near 52-week highs receive additional score |
| Data quality | Missing history, missing volume, and split-like moves reduce confidence |
| Extension risk | Extreme prior extension reduces quality score |
Reversal Risk Score
reversal_risk_score ranges from 0 to 100.
Risk increases when:
- Catalyst is
NO_CLEAR_NEWS,LOW_FLOAT_SPECULATION, orCAPITAL_STRUCTURE - The event is extremely extended
- Upside event closes weakly
- Dollar volume is below the liquidity floor
- Data quality flags are present
Study Priority Score
study_priority_score ranks which events deserve manual review. A high-priority event is not necessarily a trade; it may be a high-quality example, a dangerous failure, or a useful negative case.
Suggested Human Ratings
| Condition | Human-facing label |
|---|---|
| Quality >= 75 and catalyst is durable | A_REVALUATION_OR_HIGH_QUALITY_EVENT |
| Quality >= 60 | B_MOMENTUM_EVENT |
| Reversal risk >= 65 | D_LOW_QUALITY_OR_REVERSAL_RISK |
| Otherwise | C_STUDY_ONLY |
Handoff Flags
| Flag | Meaning |
|---|---|
stockbee_episodic_pivot |
Event resembles a Day 1 EP / gap-and-go / base breakout candidate |
pead_screener |
Earnings or guidance catalyst may warrant PEAD-style research |
theme_detector |
Theme or high-priority event may deserve cluster review |
edge_candidate_agent |
Event is strong enough to export as an edge research prompt |
parabolic_short_watch |
Reversal-risk profile is high enough for study-only exhaustion review |