Workflow Server

i read very carefully all topics in SQ forum and all bugs here and there and testing the users strategies and mostly found out, that they dont even know what they are doing

they even dont know the basic - that they must prepare the data

X1) A workflow generate 5000 strategies for period A1and the portfolio test for unseen data produces profitfaktor P1 >1.1

X2) A workflow generate 5000 strategies for period A2 (period shifted by z days in the past) and the portfolio test for unseen data produces profitfakator P2  >1.1

..

Xn) A workflow generate 5000 strategies for period An (period shifted by z days in the past) and the portfolio test for unseen data produces profitfaktor Pn>1.1

averageProfitfaktorOfAll=(P1+P2+... Pn)/n

A good workflow have a good averageProfitfkator for example.

This is my test for good portfolios. It is very difficult to find workflows which are good for all time periods. But some workflows are better than other.

I can find portfolios with averageProfitfkatorOfAll=1.15 for example. This is not so much. But If you see the results for every periods of unseen data, there are market phases where the portfolios are good.

You can find with this mechanism very fast bad workflows.

You can compare this mechnism with a walk forward analysis. Every period of unseen data should be good.

If we collect the good workflows we will get a big pool of knowledge.