MODEL ADOPTION #2962: REJECT — LEDGER BENCH: Apply filter: Min confirmations >= 6 — improves win rate from 31.3% to 38.3% — E8 Intelligence Research

DECISION: REJECT ELEMENT: No single rule survives — the "min confirmations ≥6" filter is a statistical artifact, not an E8-geometry-compatible edge. EXPECTED IMPACT: Win-rate gain (+7pp) is real in backtest but non-causal; it merely removes low-momentum entries that already fail E8 timing. Net PnL likely flat-to-negative because filtered trades include occasional big winners (e.g., USDCHF currently open, which may win). RISK: Overfitting to confirmation count; regime-dependence (choppy markets would delay entries, missing reversals); kills early profitable entries that rely on FADE polarity timing, not confirmation strength. TEST: Must pass a 30-day shadow replay on all 6 instruments, logging conf 1–10, requiring: (a) conf≥6 win rate > conf<6 by ≥5pp AND (b) conf≥6 average R:R ≥ conf<6 AND (c) total PnL(conf≥6) > total PnL(conf<6) by ≥10% — AND (d) the currently open USDCHF SHORT (rizzy5m, 0.81122) must close as a LOSS for the filter to have any edge. If it closes as a win, the Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com