LAB #4814 NEUTRAL: LEDGER BENCH: Apply filter: Session = LONDON — improves win rate from 31.8% to 34.8% (+2.9pp, n=23) — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Session = LONDON — improves win rate from 31.8% to 34.8% (+2.9pp, n=23) SAME-WINDOW EFFECT: Of the 10 live trades in the window, the London-session filter would have excluded the 11:39–11:45+02:00 cluster (NZDUSD, GBPUSD, EURUSD rizzy7m — all REJECTED/open) and the 10:25–10:55 entries depending on exact London boundary definition, leaving roughly 3–5 trades. The two closed losers (XAUUSD 11:05, AUDUSD 10:38) sit inside London hours, so the filter would NOT have removed them — the +2.9pp claim is not demonstrated by this window. Net effect here is fewer trades, same losers, and a sample too small to confirm any win-rate delta. OVER-TIME PROJECTION: A +2.9pp edge on n=23 is statistically indistinguishable from noise (95% CI spans roughly −10pp to +16pp); adopting it live risks curve-fitting to a single session slice. Over 3–6 months the filter will cut trade count materially, which reduces compounding Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com