LAB #3894 NEUTRAL: LEDGER BENCH: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 32.5% to 33.3% ( — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Quarterly phase = MANIPULATION — improves win rate from 32.5% to 33.3% (+0.8pp, n=1445) SAME-WINDOW EFFECT: Applying the MANIPULATION-phase filter would have blocked the three SHORT entries on AUDUSD, NZDUSD, and XAUUSD taken on 2026-08-27 (all rizzy5m). Of those, two closed as losses and one remains open/rejected — so the filter would have removed 2 confirmed losses and 1 unresolved trade. The remaining 7 trades would have been unaffected, yielding a net improvement of +2 wins avoided (no PnL gain, but loss avoidance). Expected win-rate delta on this small sample: from 2/9 (22%) to 2/7 (28.6%) — a +6.6pp improvement, but with n=7 the confidence is near zero. OVER-TIME PROJECTION: Over 3-6 months, a +0.8pp win-rate gain (n=1445) is statistically marginal (z≈1.1, p≈0.27) — not significant at 95% confidence. The risk is overfitting to a quarterly phase label that may be arbitrary or regime-dependent; Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com