LAB #4777 NEUTRAL: LEDGER BENCH: Apply filter: Pool density HIGH (>=90) — improves win rate from 31.8% to 33.3% (+1.5pp — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Pool density HIGH (>=90) — improves win rate from 31.8% to 33.3% (+1.5pp, n=1942) SAME-WINDOW EFFECT: The filter would only bite on trades where pool density was below 90 at entry — and the live window gives us no density readings, so I can't confirm which of the 10 trades it would have blocked. The two closed losses (XAUUSD, AUDUSD) and the two closed wins (USDCHF, USDCAD) are the only PnL-relevant rows; if the filter had excluded either loss it's a small positive, if it excluded either win it's a wash-to-negative. Net effect on this 10-trade window is likely zero to ±1 trade, which is statistically meaningless at n=4 closed. OVER-TIME PROJECTION: +1.5pp on a 31.8% base is a ~4.7% relative improvement — real but thin, and the n=1942 sample means the 95% CI on that delta is roughly ±2pp, so the true effect could plausibly be zero or negative. The bigger risk is selection drift: density is correlated Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com