PAPER ONLY CANDIDATE ANALYSIS / RISK LABloading · —
Why candidates pass, fail, and what changes on PAPER
Auditable candidate evidence, edge distribution, research cost and read-only historical simulation.
Markets scanned today
—
Trade candidates today
—
FIRE today
—
REJECT today
—
Highest edge
—
Median edge
—
P90 edge
—
Current min edge
—
Research used / cap
—
Research OK
—
Research skipped
—
EDGE_TOO_LOW
—
Highest Edge: —
Required Edge: —
EDGE ELIGIBLE CANDIDATES EXIST
Edge distribution
Candidate edge buckets for the selected range.
Top candidates
Default sort: edge DESC · 0 matched.
| Timestamp | Market | Question | Price | p_model | p_yes | Edge | LLM | ILSA | HOLT | BRAM | Liquidity | Volume | Risk | Reason | Paper order |
|---|
Paper Risk Settings
DB overrides only. No `.env`, LIVE, secrets or Emergency Stop changes.
Research budget
Daily cap
$—
Used
$—
Remaining
$—
LLM calls
—
Skipped cap
—
What-if simulator
Read-only historical reclassification. It never changes settings or creates orders.
Candidates
—
Pass edge only
—
Reject at edge
—
FIRE after all rules
—
| Historical result | Overall | LLM | Math-only |
|---|---|---|---|
| Hypothetical trades | — | — | — |
| Wins | — | — | — |
| Losses | — | — | — |
| Win rate | — | — | — |
| Gross P&L | $— | $— | $— |
| Fees / slippage | $— | $— | $— |
| Net P&L | $— | $— | $— |
LLM effectiveness
| Metric | LLM | Math-only |
|---|---|---|
| Candidates | — | — |
| Median edge | — | — |
| P90 edge | — | — |
| Max edge | — | — |
| FIRE | — | — |
| Resolved win rate | — | — |
| Resolved P&L | $— | $— |