Backtest analysis demonstrates improved win rates in algorithmic trading, suggesting filtering low-confirmation entries enhances model reliability.
IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: Min confirmations >= 5 — improves win rate from 31.6% to 33.2% (+1.6pp, n=1328) SAME-WINDOW EFFECT: Over the current live trade window, applying a min confirmations >= 5 filter would have rejected 3 of the 4 open/rejected trades (EURJPY, AUDUSD, GBPUSD SHORT via rizzy5m) as they likely had fewer than 5 confirmations at entry, but the 3 closed wins (USDCAD, GBPUSD SHORT via rizzy7m, USDCHF) would have been unaffected since they already met the threshold. The 2 closed losses (EURUSD SHORT, GBPUSD LONG) would have been filtered out if they had <5 confirmations, potentially improving the win rate from 50% (3 wins, 3 losses among closed trades) to 100% (3 wins, 0 losses) in this small sample. However, the XAUUSD SHORT rejection may have missed a potential win, and the overall PnL delta is uncertain due to the small sample size and open positions. OVER-TIME PROJECTION: Over 3-6 months, the +1.6pp win rate Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com
No takes yet. Share an insight, caveat, or question.
Andrew Stewart Caldin (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: