LAB #2584 NEUTRAL: LEDGER BENCH: Apply filter: XAUUSD LONG @ NY_OPEN with 5 conf — improves win rate from 31.0% to 76.2 — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: XAUUSD LONG @ NY_OPEN with 5 conf — improves win rate from 31.0% to 76.2% (+45.2pp, n=21) SAME-WINDOW EFFECT: Over the current live window, the filter would have blocked the XAUUSD SHORT entries on 2026-08-04 and 2026-07-31 (both losses), while allowing the XAUUSD LONG on 2026-08-06 (win). That would have removed 2 losses and kept 1 win, improving the window's raw win rate from 6/10 to 7/8 (70% → 87.5%) and adding roughly +2R net (assuming ~1R per loss avoided, ~2R win retained). However, the filter would not have touched any GBPUSD, USDCHF, AUDUSD, USDCAD, or USDJPY trades, so the overall PnL delta is modest but positive. The n=21 backtest is small, and the 45pp jump is suspiciously large — likely overfit to a specific session/conf threshold. OVER-TIME PROJECTION: Over 3-6 months, this filter could reduce XAUUSD drawdowns if the edge is real, but the risk is high: NY_OPEN + 5 conf is a narrow slice Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com