Closing reconciliation

Tested three new market-complexity measures to see whether each says something the shipped features do not. One was clean everywhere, one borderline, one a near-duplicate — and a closing matrix showed the three also overlap each other in a handful of extreme cells.

What it measured

Quantity	Value	Meaning
Test grid	3 candidates x 210 tests (21 symbol-threshold cells x 10 regime slices)	the verdict is the worst cell, not an average
Statistical Complexity worst cell	max \|rho\| 0.97 vs bar_petrosian_fd; 140/65/5 cell split	overlaps an existing feature in 5 cells
CECP velocity worst cell	max \|rho\| 0.67, max R2 0.12; 210/0/0 split	clean in every one of 210 cells
Binary CECP worst cell	max \|rho\| 0.94 vs bar_petrosian_fd; 201/9/0 split	borderline in 9 cells
Cross-candidate matrix worst case	\|rho\| 0.942, pairwise R2 0.963, leave-one-out R2 0.970 over 209 valid cells	the three overlap each other at the extremes
Breach localisation	3 of 209 cells (1.4%), all BTCUSDT at 750 dbps	redundancy confined to sparse high-threshold cells

How it ran

Firing	What it did
01	Statistical Complexity (MPR) screened against the live panel across 210 cells
02	CECP velocity — same screen, same grid
03	Binary CECP — same screen, same grid
Finish	Cross-candidate matrix over all 10 slices (crosscheck.html)

What it produced

Three candidate pages plus the Finish matrix crosscheck.html (generated 2026-06-10 05:07 UTC) with per-pair medians and leave-one-out R2
Underlying results/cross_candidate_slice*.jsonl; registry entries in orthogonal_candidates_v2.py
All three ride the already-ratified PR #513 ordinal embedding (m=3, tau=1) — no new project-wide constant

Carries forward

Promotion to production columns is a separate decision outside this loop. Open caveat on candidate #20: confirm Binary-CECP's sign-binarization is faithful to Pinto 2025 (arXiv:2504.01974) before any promotion.

Why it is closed

Slate exhausted AND the Finish step ran: '3 Candidates 1 PASS 1 WATCH 1 BAN 0 Pending' plus crosscheck.html generated 2026-06-10 05:07 UTC. (LOOP_LEDGER.json _status_evidence)

Lifecycle status is process state, not a judgement of the findings. Results are stated as numbers with their uncertainty. Generated between CLOSING-RECONCILIATION markers — regenerate rather than edit.

crypto · ODB pipeline · nasimubd · overnight orthogonality loop · 2026-06-10 03:08 UTC

Crypto CECP · Statistical-Complexity — 3-Candidate Orthogonality Evaluation

Testing 3 complexity-entropy (CECP) candidate features for whether each adds information orthogonal to the live production panel — one candidate per firing, read-only across 10 regime slices × 21 (symbol, threshold) cells = 210 tests on ClickHouse. Gate: ‖ρ‖≤0.85 ∧ R²≤0.85 ∧ h_norm≥0.05 (worst-valid-cell). Scope: orthogonality (Axis-1) only — these three already PASS Axis-2 (parameterless) via PR#513 reuse. Resource-capped to 2 cores @ 200% / 2 GB, spare capacity only.

Candidates

PASS

WATCH

BAN

Pending

How to read this dashboard

What the loop is doing: checking, one feature at a time, whether a new market-complexity measure tells us something the features we already ship don't. A PASS means it's genuinely new everywhere we tested; WATCH means it's borderline-redundant somewhere; BAN means it duplicates an existing feature.
What "grounded" means here: a PASS is earned on the worst of 210 real-data cells, not an average — a feature has to be orthogonal in every symbol, threshold, and regime.
Each candidate page explains in plain terms what the feature measures, how it was tested, what every number means, how its verdict was reached, what resources it used, and anything worth flagging to you.
Scope: orthogonality only — no profit/signal claim is made here. Promotion to a production column is a separate operator decision after the cross-candidate Finish.

Candidates: #16 statistical-complexity (MPR) · #19 cecp-velocity · #20 binary-cecp. All three ride the ratified PR#513 ordinal embedding (m=3, τ=1) — no new project-wide constant is promoted by this loop. Empirical sibling-redundancy among them is resolved by the cross-candidate matrix (all 10 slices) at Finish.

Candidate roster

#	Candidate	Verdict	Max ‖ρ‖	Valid cells	Cost
01	🧭 Statistical Complexity (MPR) statistical-complexity · ordpy	BAN	0.97	210	fast
02	🛰️ CECP velocity statistical-complexity · ordpy	PASS	0.67	210	fast
03	🔳 Binary CECP statistical-complexity · custom	WATCH	0.94	210	fast