I’ve created a scoring system so as to easily rate each set of robustness test results.
The calculation is:
100-(((H-L)/H))*100)
Basically an inverted percentage of the difference as a percentage of the best result.
H = the best % of winners rate
L = the worst % of winners rate
Replace % of winners with average gain per trade if scoring that. A high score is a good score and a low score is a bad score.
I have set up an Excel book so that I can just clear the contents and drag in my latest test results and the graphs just appear and the calculations for the score are presented. The attached image is for QTY and RANDOM settings of 10 and 5. If I want to test other settings then I will just copy the Excel sheet to another tab and alter the chart data selection and then have a sheet for those settings all ready to use in the future. I’ve included graphs of trade numbers so it is easy to see if there has been plenty of them across the test.
The Excel book is attached – just right click and clear the contents and then drag in your average gain per trade results to the top left box and then drag the % of winners directly under it leaving a one line gap.
