Yes and the only way to test this is to do it manually. We need to basically pick multiple random segments of data (all of the same duration)
I can’t get my head around why it has to be random / manual?
If we are trying to simulate what we would do in Live running then 10k contiguous chunks of data (as using the WF tool) is as near as we can get to Live running (other than Live running)? Live running / re-optimising during the last 50k bars would have used the exact same 50k bars that we can now use via the WF tool.
I’ve been running a Walk Forward with 3 variables over 50k bars for the past 3 hours and it has only just started on it’s 4th stage! WF slowness saps one’s enthusiasm for doing studies like we are attempting??
