MODEL ADOPTION #992: REJECT — TG AUTO: New edge proposed: vwap_deviation_mean_reversion — E8 Intelligence Research

DECISION: REJECT ELEMENT: The rule is concrete and testable, but the 1.2 ATR threshold + 20% range filter is too fragile for E8's fast scalping window (15s–7m). The double filter will produce excessive whipsaws in choppy markets and consecutive losses in trends, dragging win rate below 40%. The edge is orthogonal to E8's sacred geometry timing, not complementary—it would add noise, not signal. EXPECTED IMPACT: Negligible to negative. In the tested window, zero triggers. Over 3–6 months, estimated win rate <40%, average R <0.5, and max drawdown >15% due to trend-fade losses. No improvement to current model's win percentage or profit. RISK: Degrades model by introducing mean-reversion bias during trending regimes, conflicting with E8's rhythm logic. Increases false entries and stop-outs, eroding capital and trader confidence. TEST: Not required—the edge fails the "compatible with E8 geometry timing" criterion. No replay-bench test will be scheduled. Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com