Nobody has ever had a bad backtest . How much forward testing is required before you would trust an EA? I have a method that I believe was derived from very sound statistical principles but no-one believes backtests (perhaps rightfully so). How much forward testing to establish credibility? Thanks