LAB #3867 PROMISING: LEDGER BENCH: Apply filter: Session = OFF_SESSION — improves win rate from 32.5% to 43.5% (+11.0pp, — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Session = OFF_SESSION — improves win rate from 32.5% to 43.5% (+11.0pp, n=124) SAME-WINDOW EFFECT: Applying the OFF_SESSION filter would have excluded all trades entered outside the defined session. In the current window, the only trade that would survive is the GBPUSD SHORT via rizzy5m (11:05) and possibly the EURUSD LONG via rizzy7m (10:20) — the rest (entries at 09:15, 09:25, 09:26, 09:52, 03:57, 10:11) would be blocked. That would have avoided 2 losses and 1 open/rejected, but also removed 1 win (EURUSD LONG). Net effect on this tiny sample: win rate would improve from 2/8 closed (25%) to 1/2 closed (50%), but PnL delta is ambiguous — you lose the EURUSD win and avoid two losses, roughly a wash on expectancy. OVER-TIME PROJECTION: Over 3–6 months, the +11pp win-rate improvement (n=124) suggests a genuine regime bias — OFF_SESSION entries likely catch stale momentum or low-liquidity noise that re Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com