Rinse 'n Repeat

[WhoKnows]

My live account however, running #1 is in $2.035,51 DD as I'm writing this. Profit last week was $394,14
No idea why the huge difference in DD, didn't had a chance to look into it yet.

Hi Gertje, did you already find what caused this 2K DD?

[Gertje]

My live account however, running #1 is in $2.035,51 DD as I'm writing this. Profit last week was $394,14
No idea why the huge difference in DD, didn't had a chance to look into it yet.

Hi Gertje, did you already find what caused this 2K DD?

The live-account had a bunch of buy trades open that the demo's didn't have because they started later.
The buy-basket of the demo's was closed last week during the high volatility, where at the live-account the trades didn't reach the closing-point and stayed open.
These open buy's [~80] caused the DD.

CSSO could have solved this partially, but earlier testing showed me that CSSO 'off' is 50% more profitable than CSSO 'on'.

To be clear, I'm NOT surprised about the 2K DD because I know this happened before and will happen again.
I was surprised by the huge difference in DD between the live- and demo accounts.

[WhoKnows]

The live-account had a bunch of buy trades open that the demo's didn't have because they started later. ...............
I was surprised by the huge difference in DD between the live- and demo accounts.

Thanks, very interesting.
In my testing I start each batch of test setfiles at the same time, and have 1 reference setfile that goes into each batch.
Comparing results I am getting the suspicion that the starting time of the EA with the same setfile is often more different than the different setfiles.
E.g. I have a reference setfile running for weeks that is making lot of profit, while two identical RnRsetfiles started in a different batch just keeps losing.
I guess I have to do some tests starting multiple identical RnR set-files at exactly the same time, and some other identical ones at some intervals later. Maybe that will tell me more about the 'natural variation' between identical RnRsetfiles.

[pivotter]

The RnR EXP 1D has a problem with unknown ticket.The attached version RnR EXP 1E.patched was patched with code suggestions from the DottyBot threat, the unknown ticket problem is gone.

Only the bool MopUpTradeClosureFailures() code tightening update on http://www.stevehopwoodforex.com/phpBB3 ... &start=210 was skipped as the new code contained an error.


2016.08.05 12:35:20.176 Rinse-n-Repeat-EXP.1d GBPJPY,M5: unknown ticket 139001359 for OrderCloseBy function
001550

you can search the attached file for //patch to see what is patched.
file:Rinse-n-Repeat - EXPv1E.patched.mq4

Rinse-n-Repeat - EXPv1E.patched.mq4

RnR EXP 1E.patched and RnR EXP 1E are currently running in demo started on the same tick till the end of next week.

pivotter

[WhoKnows]

@pivotter. Thanks, I will include some patched ones in my next batch.

[TraderJoeForex]

Interesting results. I think the MA filter may work best on it's own i.e. with recovery mode OFF so not using the recovery set file I posted as there may be a conflict there between all the filters.

.... from testing this week I have found using a 10LWMA filter on M1 to be a revelation and much more accurate than M5 or M15. I have adopted this with 5 and 50pip grids in my TrendRangeCombo thread.

Cheers again Bruster and Gertje for this addition.

[Gertje]

.... from testing this week I have found using a 10LWMA filter on M1 to be a revelation and much more accurate than M5 or M15. I have adopted this with 5 and 50pip grids in my TrendRangeCombo thread.

Cheers again Bruster and Gertje for this addition.

But.... Does it help with fighting DD.......?

image.png

[TraderJoeForex]

.... from testing this week I have found using a 10LWMA filter on M1 to be a revelation and much more accurate than M5 or M15. I have adopted this with 5 and 50pip grids in my TrendRangeCombo thread.

Cheers again Bruster and Gertje for this addition.

But.... Does it help with fighting DD.......?

Yes!

[pivotter]

The RnR EXP 1D has a problem with unknown ticket.The attached version RnR EXP 1E.patched was patched with code suggestions from the DottyBot threat, the unknown ticket problem is gone.

Only the bool MopUpTradeClosureFailures() code tightening update on http://www.stevehopwoodforex.com/phpBB3 ... &start=210 was skipped as the new code contained an error.


2016.08.05 12:35:20.176 Rinse-n-Repeat-EXP.1d GBPJPY,M5: unknown ticket 139001359 for OrderCloseBy function
001550

you can search the attached file for //patch to see what is patched.
file:Rinse-n-Repeat - EXPv1E.patched.mq4



RnR EXP 1E.patched and RnR EXP 1E are currently running in demo started on the same tick till the end of next week.

pivotter

update
After traveling today to another location R'nR-exp and R'nR-exp-patched OOTB got restarted 10 hours and started both with a massive close in profit.
There was no high spread.
Rinse-n-Repeat - EXPv1E had 29 unknown tickets on 44 orders, the patched version had no unknown tickets and closed 40+ orders as it should.

[WhoKnows]

Gertje, is that DD from RnR? Just asking, as your the EA in your screenshot is from Dotty.
Apparently you had some real bad luck with having open buys just before the market dropped.

Thanks TJF for the tip on using a 1 min TF for the MA.
My tests with 5 min TF were unsuccessful, I will try the 1M next week too.

I have been running about 30 tests last 3 weeks. But as internet lately has been so crappy that the modem regularly dropped out, unfortunately I don't know how much the results I found were influenced by this.
What concerns me is that I see a big (>$300) difference between identical setfiles only started some hours from each other, with the DD determined by having more or less positions just before the market moved.
Within every week I have seen differences in the results from identical setfiles, that were as big as differences between between the result of different setfiles.

I made me realize I should know more about the 'natural variation' of identical setfiles. It might be that differences that I saw between the results of different setfiles is due to random variation, rather than due to different settings in the setfiles.
Will test this again next week from a VPS with multiple identical setfiles started the same time and some more identical ones started at 4 hour intervals.