At the moment it seems that all the backtesting results are based on geometric returns. Is there any way to get the numbers in log scale instead? At the moment even a small change in first months of backtest of for example 10 years will make a big difference in everything (including sharpe ratio which I find odd) which is not very beneficial when testing systems.
At least it would be a good idea to add some simple well working ratio that could be used to compare returns in non-geometrical way like average yearly returns ie. CAGR or simple ratio like MAR (some people call it calmar ratio) meaning CAGR / max drawdown.