LAB #4821 NEUTRAL: LEDGER BENCH: Apply filter: Min confirmations >= 6 — improves win rate from 31.8% to 36.2% (+4.4pp, — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Min confirmations >= 6 — improves win rate from 31.8% to 36.2% (+4.4pp, n=945) SAME-WINDOW EFFECT: The filter would have removed the NZDUSD, GBPUSD, EURUSD(rizzy7m), USDCHF(0.81152), and EURJPY entries — all currently sitting in REJECTED/open limbo — leaving the closed-book record untouched at 2W/3L, since none of the five closed trades (EURUSD, XAUUSD, AUDUSD, USDCHF, USDCAD) show confirmation counts in the provided data. Net effect on realized PnL over this window: zero. The only observable change is a smaller pending book, which is a selection effect, not a performance one — and the +4.4pp claim rests on n=945 from a different sample, not this window. OVER-TIME PROJECTION: A 4.4pp win-rate lift on a 31.8% base is a ~14% relative improvement, which is meaningful if it survives out-of-sample — but the GTX 30-day tape (153 TP2 vs 305 LOSS vs 16 BE, ~32.3% win rate) is consistent with the *unfiltered Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com