MODEL ADOPTION #1993: REJECT — LEDGER BENCH: Apply filter: Pool density HIGH (>=90) — improves win rate from 31.9% to 33. — E8 Intelligence Research

DECISION: REJECT ELEMENT: No single rule survives — pool-density filter (≥90) is a statistical artifact, not an E8-geometry-compatible edge. EXPECTED IMPACT: +0.0pp win-rate, +0.0R/month. The +1.6pp gain is within noise for n=1350 (σ≈1.3pp), and the live "3/5 vs 3/8" is a 3-trade coincidence. Filtering ~30% of trades reduces sample size and compounds slower, while the claimed benefit is not tied to any phase/rhythm node in Gods-Clock logic. RISK: Regime-dependence — high pool density often coincides with low-volatility consolidation; in trending markets (where E8 fade edges are strongest), density drops, and this filter would cut the exact high-volatility winners that drive profit. Also, threshold 90 is untested against 85/95, and the backtest window may be a single favorable regime. TEST: Replay-bench must show: (1) On the full 3-year tick history, split by volatility tercile (low/mid/high), the filter improves win-rate AND profit factor in ALL terciles, not just the aggregate Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com