LAB #4604 PROMISING: LEDGER BENCH: Apply filter: Min confirmations >= 7 — improves win rate from 31.6% to 48.1% (+16.5pp, — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Min confirmations >= 7 — improves win rate from 31.6% to 48.1% (+16.5pp, n=27) SAME-WINDOW EFFECT: Applying a minimum of 7 confirmations would have filtered out all three REJECTED/open trades (USDCHF LONG, EURJPY SHORT, AUDUSD LONG) — which are currently unresolved and likely low-conviction entries. Of the seven closed trades, all had at least 7 confirmations (they were winners or losers based on other factors), so the filter would not have removed any closed winners or losers. This would have reduced exposure by 30% (3 of 10 trades) while keeping the 4 wins and 3 losses intact, mechanically raising the closed-trade win rate from 4/7 (57.1%) to 4/4 (100%) — but that's a small-sample artifact, not a real edge. The +16.5pp backtest gain is plausible but driven by n=27, so the live delta here is likely near zero on PnL since rejected trades were never closed. OVER-TIME PROJECTION: Over 3–6 months, a 7- Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com