Wrong backrest results!!??
Wrong backrest results!!??
Today I backtest the same ea I tested in the past, but now it has a huge drawdown. Why is that ? Is it because I have auto upgraded to latest terminal version? What I mean is, the result is very different from what I used to see.
Re: Wrong backrest results!!??
Please see viewtopic.php?f=4&t=1120 and see if this applies. Thanks.
Re: Wrong backrest results!!??
Hello,
I updated to Tickstory 1.8.2 (+Build 988) and have a backtest-problem too.
Before I updated, I used version 920 with a Tickstory 1.7.3 for my backtests.
My current problem is that I get different results between a backtest in the optimization-mode and if I run it without optimization.
Maybe I found a small hint that could help in that case. If I run it with optimization, than I get the following message in the strategy-tester journal:
"2016.07.06 13:50:07.245 TestGenerator: actual tick file "C:\Users\Administrator\AppData\Roaming\MetaQuotes\Terminal\85D92BEAF0C0FC647A9D2D6AB1712DC5\tester\history\CADCHF15_0.fxt" found"
If I use the same settings and only disable the "Optimization"-Checkbox, than I get the following message:
"2016.07.06 13:49:28.260 TestGenerator: file "C:\Users\Administrator\AppData\Roaming\MetaQuotes\Terminal\85D92BEAF0C0FC647A9D2D6AB1712DC5\tester\history\CADCHF15_0.fxt" is read-only"
I think before the update, I always saw the first message, no matter if I used the optimization or not.
Any idea how to solve this?
I updated to Tickstory 1.8.2 (+Build 988) and have a backtest-problem too.
Before I updated, I used version 920 with a Tickstory 1.7.3 for my backtests.
My current problem is that I get different results between a backtest in the optimization-mode and if I run it without optimization.
Maybe I found a small hint that could help in that case. If I run it with optimization, than I get the following message in the strategy-tester journal:
"2016.07.06 13:50:07.245 TestGenerator: actual tick file "C:\Users\Administrator\AppData\Roaming\MetaQuotes\Terminal\85D92BEAF0C0FC647A9D2D6AB1712DC5\tester\history\CADCHF15_0.fxt" found"
If I use the same settings and only disable the "Optimization"-Checkbox, than I get the following message:
"2016.07.06 13:49:28.260 TestGenerator: file "C:\Users\Administrator\AppData\Roaming\MetaQuotes\Terminal\85D92BEAF0C0FC647A9D2D6AB1712DC5\tester\history\CADCHF15_0.fxt" is read-only"
I think before the update, I always saw the first message, no matter if I used the optimization or not.
Any idea how to solve this?
Re: Wrong backrest results!!??
Hi Xeres,
Do you know the last Build that the optimisation was working for you? There have been a flurry of recent updates so if you are able to identify the breaking version then this may help. Have you mentioned this on the MQL forums as this might be the best place to help identify/resolve the issue.
Thanks.
Do you know the last Build that the optimisation was working for you? There have been a flurry of recent updates so if you are able to identify the breaking version then this may help. Have you mentioned this on the MQL forums as this might be the best place to help identify/resolve the issue.
Thanks.
Re: Wrong backrest results!!??
I'm afraid I have the same problem. When I used tickstory 1.7 everything was fine. When I upgraded ro v1.8 I had a big drawdown on my GBP algo back-testing. The DD was on Jun 23-24 (Brexit day). It didn't happen on v1.7 and it didn't happen in real trading (I use this ALGO on a live account).
Today I upgraded to v 1.8.2 and I still get this huge DD, and I also get different results vs. the previous version of tickstory.
I use MT4 build 950.
Any ideas?
TX
Today I upgraded to v 1.8.2 and I still get this huge DD, and I also get different results vs. the previous version of tickstory.
I use MT4 build 950.
Any ideas?
TX
Re: Wrong backrest results!!??
Hi Adibi83,
Do you have the specific of Tickstory v1.7 that you were using? Can you see any information in the change log that may apply to you? Between v1.7.5 and v1.8.2 the changes only relate to launcher support which should not affect your results. If the differences can be isolated, then it should be easier to identify why you are getting your drawdown,
If you are able to isolate the issue to the "MACD Sample" EA and show you're getting different results between the two versions, then it would be possible to investigate further. There are obviously lots of variables that could be causing your changed results otherwise.
Thanks.
Do you have the specific of Tickstory v1.7 that you were using? Can you see any information in the change log that may apply to you? Between v1.7.5 and v1.8.2 the changes only relate to launcher support which should not affect your results. If the differences can be isolated, then it should be easier to identify why you are getting your drawdown,
If you are able to isolate the issue to the "MACD Sample" EA and show you're getting different results between the two versions, then it would be possible to investigate further. There are obviously lots of variables that could be causing your changed results otherwise.
Thanks.