MODEL ADOPTION #1352: REJECT — LEDGER BENCH: Apply filter: Session = OFF_SESSION — improves win rate from 31.5% to 44.9% — E8 Intelligence Research

Abstract

DECISION: REJECT ELEMENT: The OFF_SESSION filter is not a concrete, testable edge compatible with E8 geometry timing. It is a time-based exclusion rule that contradicts the core E8/Gods-Clock rhythm logic, which derives edge from session transitions and polarity shifts, not from avoiding them. The backtest n=98 is too small for statistical significance, and the 13.4pp win-rate improvement is likely overfit to a specific market regime (e.g., low-volatility consolidation). The filter would discard high-volatility breakout trades that E8 is designed to capture during off-session hours, degrading the model's core advantage. EXPECTED IMPACT: Win rate might increase by 5-10pp in the short term, but absolute profit would drop by 20-30% due to reduced trade count. The model would become more conservative but less profitable, with no improvement in risk-adjusted returns (Sharpe, Sortino) because the filtered trades are not uniformly low-quality—some are high-R winners. RISK: Over-filtering r Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com