I agree with Kevind, and add that I find it wise to add a significant Slippage when backtesting in an attempt to match real-time trading - presently I use Slippage = 5 ticks (I use MedianRenko charts), but after some time testing in sim in real time I think perhaps 8 is more accurate.