thank you for your feedback.
A 'model quality of 0.0%' simply indicates that *no* modeling and optimization was performed.
MetaTrader allows to optimize your input parameters during a backtest and terms this brute-force search for best parameters 'modeling'. However, the problem is the following: If your EA is highly depending on input parameters it is very likely that it is well fitted for the past, but will likely fail in the future if trends/markets change. Therefore, my goal was to build an EA which is as parameter-invariant as possible. In addition, I did not performed any optimization (which is not possible for 10 years anyways without super-computer power). The result is an EA which is *not* the best with regard to maximum win, but shows on average a stable positive result and a low risk profile.