LAB #2078 NEUTRAL: LEDGER BENCH: Apply filter: Min confirmations >= 6 — improves win rate from 31.9% to 42.0% (+10.0pp, — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Min confirmations >= 6 — improves win rate from 31.9% to 42.0% (+10.0pp, n=355) SAME-WINDOW EFFECT: Applying a minimum-6-confirmation filter would have eliminated the 5 rizzy5m entries (XAUUSD SHORT x3, AUDUSD LONG, USDCAD SHORT, GBPUSD SHORT) — all of which entered on 5m confirmations. Of those, 2 were wins (AUDUSD, USDCAD) and 3 were losses (XAUUSD x2, GBPUSD). Net PnL impact: removing 2 wins and 3 losses would have shifted the window from 5W/4L/1O to 3W/1L/1O — a raw win-rate improvement from ~55% to ~75% on closed trades, but a PnL delta likely negative because the two removed wins (AUDUSD, USDCAD) probably outweighed the three small losses. The filter would not have touched the wkjudas or sqz entries, so those remain. OVER-TIME PROJECTION: Over 3–6 months, the +10pp win-rate gain (31.9%→42.0%) comes at the cost of fewer trades (n drops from ~355 to ~250–280). If the backtest is honest, this is Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com