MODEL ADOPTION #2128: REJECT — LEDGER BENCH: Apply filter: Min confirmations >= 5 — improves win rate from 32.0% to 33.6% — E8 Intelligence Research

DECISION: REJECT ELEMENT: No single rule survives — "min confirmations >=5" is a statistical artifact, not an E8-geometry timing edge. EXPECTED IMPACT: +1.6pp win-rate on n=1354 is within noise (σ≈1.3pp for that sample). The 10-trade window "improvement" is cherry-picked survivorship bias — you cannot assume filtered losses wouldn't be replaced by other entries in live flow. Real effect likely ≤0.5pp win-rate, with 5-10% trade-frequency loss → net PnL flat or negative. RISK: Overfitting to confirmation count will systematically delay entries during fast-trend regimes (Gods-Clock acceleration phases), causing missed breakouts that are the model's primary profit source. Also, the filter correlates with lower volatility — it will bias the book toward slow, mean-reverting trades, degrading the FADE polarity edge that depends on early counter-trend timing. TEST: Before any reconsideration, run a 12-month historical replay with strict out-of-sample walk-forward (6-month train, 6-mont Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com