MODEL ADOPTION #4037: REJECT — LEDGER BENCH: Apply filter: Session = OFF_SESSION — improves win rate from 32.5% to 42.9% — E8 Intelligence Research

DECISION: REJECT ELEMENT: No single rule survives — the OFF_SESSION filter is a session-time heuristic, not an E8 geometry timing edge, and it would disable wkjudas entirely without proving that strategy has no profitable session. EXPECTED IMPACT: Win-rate gain is likely overfit to n=126 and low-volatility regime; even if +10pp holds, it sacrifices high-volatility breakout winners, and PnL gain is modest (skipping 4 losses vs losing potential outsized wins). Net profit impact uncertain, likely negative in trending regimes. RISK: Broker time/DST drift makes OFF_SESSION unstable; wkjudas may have a profitable session (e.g., London open) that this filter blindly kills; also, the filter is not derived from E8 node/Gods-Clock rhythm, so it violates model geometry coherence. TEST: Must pass a 12-month replay-bench with ≥300 post-filter trades, split by strategy and session sub-buckets, AND require that the filter's win-rate improvement holds when conditioned on E8 phase (e.g., only O Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com