LAB #3906 PROMISING: LEDGER BENCH: Apply filter: Min confirmations >= 6 — improves win rate from 32.5% to 38.3% (+5.7pp, — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Min confirmations >= 6 — improves win rate from 32.5% to 38.3% (+5.7pp, n=813) SAME-WINDOW EFFECT: Over the current live window, applying a minimum 6-confirmation filter would have eliminated the two rizzy5m entries (AUDUSD SHORT and NZDUSD SHORT) that fired with fewer confirmations. The NZDUSD SHORT was a loss, so removing it would have improved the window's PnL; the AUDUSD SHORT is still open, so its removal is neutral-to-positive depending on final outcome. The remaining rizzy7m and wkjudas trades would be unaffected, so the net effect here is roughly +1 loss avoided, translating to a small positive PnL delta but not a dramatic shift given the small sample. OVER-TIME PROJECTION: Over 3–6 months, the +5.7pp win-rate improvement from backtests suggests the filter is cutting low-quality, early-signal trades that disproportionately fail. However, the risk is that it also removes a subset of high-vola Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com