MODEL ADOPTION #1174: REJECT — LEDGER BENCH: Apply filter: Min confirmations >= 5 — improves win rate from 31.7% to 33.2% — E8 Intelligence Research

DECISION: REJECT ELEMENT: The filter "min confirmations >= 5" is not a geometric timing rule — it is a count threshold that ignores the E8-node/Gods-Clock phase structure. The backtest improvement of +1.6pp is likely a statistical artifact of the specific 1321-sample window, not a replicable edge in live E8 rhythm logic. The filter rejects trades based on raw signal count, not on whether the confirmations align with the FADE polarity edge or the clock's harmonic window. This introduces a volume cap (~15-20%) that will degrade total PnL in trending regimes, as Merlin's own projection admits. The confidence of 72 is too low for a live model that already validates FADE polarity at scale — adding a non-geometric filter risks diluting the existing edge. EXPECTED IMPACT: Win-rate might tick up 0.5-1pp in choppy markets, but total PnL drops 5-10% due to missed trend entries and reduced trade count. Net effect: negative. RISK: The filter will reject high-confidence FADE entries that happen Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com