Classical Charting · Research Ledger
Every variation. Every disposition.
This page tracks the Classical Charting ideas we have tested, rejected, retained for correction, retired, or queued next. Preliminary testing is optimistic as we continue to fully test and validate performance—but an attractive historical number is never allowed to substitute for valid evidence.
Tracker summary
Prospective Evidence
Forward progress, without pretending waiting is performance.
The cards below come only from sanitized append-only or registered paper records. Historical simulations and retired tracks never enter this evidence lane.
Selected control
LoadingFixed-control paper account
- Equity
- —
- Registered decisions
- —
- Executions
- —
- Admitted
- —
- Closed
- —
- Drawdown
- —
Loading the authoritative paper-account summary.
Paired experiment
LoadingControl vs ADX10/W5 throttle
- Signals
- —
- Campaigns
- —
- Closed
- —
- Eligible
- —
—
Loading the registered comparison summary.
Former challengers
RetiredB / C / D are quarantined
- Tracks
- 3
- Sessions
- 0
- Trades
- 0
- Curves
- None
Retired before first accrual. Their initialization rows are preserved as invalidated provenance, not displayed as flat performance.
Named Study Ledger
The decision record behind the headline.
Filter the studies, then use the paired-delta map to see the historical tradeoff. Every plotted point compares only with its own study baseline; it does not make unlike tapes or accounting conventions directly comparable.
Loading studies…
Paired Historical Delta Map
Return change vs drawdown improvement
Correction / retest
Rejected
Historical / exploratory
Right is more CAGR. Up is a shallower maximum drawdown. The axes show percentage-point change against each study's own paired baseline. Daily-MTM and exit-event studies use different drawdown conventions and remain visibly labeled in the table.
No filtered studies have paired CAGR and drawdown metrics.
Browse the named-study table Filtered rows open in a compact, scrollable ledger.
| Stage | Metric basis | Candidate result | Evidence | View details |
|---|
Historical Configuration Archive
All 428 deduplicated configurations, on one map.
This is the broader Codex engineering archive through July 13, 2026. It is useful for seeing the shape of the search, not for choosing a winner: every point is preliminary, multiply tested, current-constituent historical research.
Not 428 equity curves.
The source contains 428 endpoint records. The map plots those exact endpoints; it never invents missing time series or treats ranking fields as approval.
Loading configurations…
Historical Risk / Return Map
Full-history CAGR vs maximum drawdown
Each dot is one exact endpoint from the frozen configuration catalog. Farther right is shallower drawdown; higher is greater CAGR. This is a map of past research, not validated expected performance.
No configurations match these filters.
| Public ID | Research source | Profit factor | Evidence cutoff | View details |
|---|
Page 1
How to Read the Tracker
Result and evidence are different questions.
Read the disposition first
A high historical CAGR can still be retired, rejected, PIT-gated, or marked for retest when its data, causal logic, or registration does not support the claim.
Compare only paired deltas
The named-study map uses change against each run's own baseline. Absolute endpoints from different tapes, allocations, windows, and drawdown conventions are not interchangeable.
Forward evidence starts at zero
Genesis, offered signals, and waiting orders are operational facts, not returns. Profit factor, win rate, and an equity curve remain absent until actual paper campaigns create them.
Status dictionary
- Reference
- A comparison anchor, not verified performance.
- Retest required
- The result cannot support its old claim until named defects are corrected.
- PIT-gated
- Interesting only if it survives point-in-time membership and delisting treatment.
- Rejected
- The tested candidate failed its decision bar; it remains visible so it is not rediscovered.
- Historical
- A development or exploratory result, not current promotion evidence.
- Retired
- A preserved track that cannot resume, catch up, backfill, or promote.