β€ΊNavigation

Iteration 1 β€” Tier-A Aggregate: Metric-Scale Verdict CONDITIONAL GO

Question: do the forward-matrix output scales discriminate well enough to ever support an "orthogonal-tomorrow" declaration?

Substrate: BTCUSDT@250 Γ— 10 crypto regimes (16 features) + EURUSD@5 Γ— 10 forex slices (141 features). 20 cells, real + shuffled-null per cell. Gates: G1 spread Β· G2 anchor-separation Β· G3 threshold-ability Β· G4 cross-regime signal Β· G5 null-collapse (blocking).

Scorecard

MetricCryptoForexReading
adv_auc (residual drift)USEFUL [10111]USEFUL [10111]Best performer β€” spread 0.96/0.88, regime-range spread 0.77/0.60, KS vs null 0.55/0.32
wold_R (fresh vs echo)4/5 (fails anchor only)USEFUL [10111]Massive spread (0.999), strongest null separation (KS 0.90/0.55)
granger_echo_p (set→f echo)USEFUL [10111]DEGENERATESplit verdict — works on crypto; forex set-factor degenerates, needs rework
redundancy (max|ρ| vs full set)input-onlyinput-onlyAnchor fix VERIFIED (duplicates read 1.000 both sides). Fails G5 by construction (rank-correlation is permutation-invariant) β†’ stays the INPUT series, not a forward gate
stab_freqMISCALIBRATEDMISCALIBRATEDWeak null separation (KS 0.09/0.01)
bootci_wMISCALIBRATEDMISCALIBRATEDRides on redundancy; tiny scale

Grounded

Verdict: CONDITIONAL GO β€” 2 validated gates + 1 validated input β‰  the pre-registered β‰₯5-USEFUL full-GO bar. On the right track with named laggards. Full-grid replication (all 40 symΓ—thr combos) in progress decides robustness.

Artifacts: bigblack ~/tierA_v2/tierA_aggregate.json Β· cells/*.json Β· audit folder TIER-A-METRIC-SCALE-SPIKE.md

Campaign: 2026-06-12-forward-orthogonality-matrix-scale-evaluation Β· generated 2026-06-12 08:27:05 local Β· read-only ClickHouse Β· cap 5c/5GB