LAB #2816 PROMISING: LEDGER BENCH: Apply filter: Min confirmations >= 6 — improves win rate from 31.1% to 37.9% (+6.8pp, — E8 Intelligence Research

Abstract

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Min confirmations >= 6 — improves win rate from 31.1% to 37.9% (+6.8pp, n=409) SAME-WINDOW EFFECT: Applying a minimum 6-confirmation filter would have blocked the two XAUUSD SHORT losses (rizzy5m entries on 2026-08-04 and 2026-07-31) and the GBPUSD SHORT loss from 2026-08-03, assuming those entries had fewer than 6 confirmations at trigger. That removes 3 losses from the 10-trade window, converting the record from 5W/5L to 5W/2L (71.4% win rate) — a dramatic improvement. However, it would also have blocked the USDJPY SHORT loss (wkjudas), which is already a loss, so net PnL delta is positive but not as clean as the raw win-rate suggests. The live window is too small (n=10) to separate signal from noise, but the direction aligns with the backtest. OVER-TIME PROJECTION: Over 3–6 months, a 6-confirmation floor should reduce false entries from low-conviction rizzy5m signals, likely lifting win rate towa Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com