MODEL ADOPTION #2334: REJECT — LEDGER BENCH: Apply filter: Min confirmations >= 6 — improves win rate from 31.6% to 40.3% — E8 Intelligence Research

DECISION: REJECT ELEMENT: No single rule survives — the ≥6 confirmation filter is a frequency-killer with source-concentration risk, not a timing edge. EXPECTED IMPACT: Win-rate gain (+8.7pp) is real but illusory net of trade-frequency collapse (40–60% fewer trades) and source concentration (rizzy7m/rizzy only). Compounding loss and per-trade variance spike likely outweigh the modest PnL delta. The 75% live subset is n=4 — statistically meaningless. RISK: Filter lags in fast trends → missed entries on exactly the strong moves E8 rhythm targets. Source degradation (rizzy7m/rizzy) becomes single-point failure. Backtest n=380 is not robust to regime shift; Monte Carlo would likely show wide confidence intervals. TEST: Before any reconsideration, run a 30-day shadow with dual portfolios (all vs. ≥6 confirmations) AND a Monte Carlo bootstrap on the 380 trades — require the +8.7pp edge to survive with >80% probability of ≥+4pp, AND require trade frequency to stay above 70% of baselin Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com