Hi, I've been using tickstory now very longtime and recently I'm getting very confused and I don't know what is a reliable result or not.
I 've developped a strategy and after optimizing and backtesting I had great results on XAUUSD. working several weeks on this project.
This week, I've downloaded fresh tickdata and I have totally different results.
I did a new installation afterwards, and again everything was different.
I'm using windows 10.
I also tried this on another computer with windows vista and here also the results are different from the windows 10 version.
I switched from 950 to 910 also.
In the "tester" directory I removed logs and cash very often to be sure that my optimising is starting clean.
My questions :
1. How can I be sure that my downloaded data and the metatrader are reliable ?
Is there some kind of test-procedure that I can use to check it out ? How can I be sure that there is a good synchronization between the history data and the FXT data ?
2. Which version of metatrader can I trust: 950 or 910 ?
3. Do I have to work offline or not (connected to the server); or is this not a problem anymore ?
I hope you can help me out.
Thx in advance !
Difference in results after new tickdata upload
Re: Difference in results after new tickdata upload
Hi Danny_pips,
Regarding your questions:
1. Tickstory always checks to see if you have already downloaded data for a specific period. If you have any data missing or corrupt, then it will automatically retrieve it from Dukascopy (if it fails to do so, it will generate an error). There is a chance that Dukascopy has updated the historical data (such as provided additional information when there was none) which means there is potential for two data sets downloaded at different times to be different. If you believe this may be the case, we suggest you simply create a new "Dukascopy data folder" (or delete the existing one) and re-download the period again.
2. Both Build 910 and 950 should yield the same results. Be sure you have set the spread correctly in Build 950 as detailed in our release notes: https://tickstory.com/articles/tickstor ... -released/. See also this thread of another use having the same question: viewtopic.php?f=2&t=987.
3. It is still recommended to work offline since historical data taken from the broker can cause your Tickstory data to be overwritten by the MT4 terminal.
Hope this helps.
Regarding your questions:
1. Tickstory always checks to see if you have already downloaded data for a specific period. If you have any data missing or corrupt, then it will automatically retrieve it from Dukascopy (if it fails to do so, it will generate an error). There is a chance that Dukascopy has updated the historical data (such as provided additional information when there was none) which means there is potential for two data sets downloaded at different times to be different. If you believe this may be the case, we suggest you simply create a new "Dukascopy data folder" (or delete the existing one) and re-download the period again.
2. Both Build 910 and 950 should yield the same results. Be sure you have set the spread correctly in Build 950 as detailed in our release notes: https://tickstory.com/articles/tickstor ... -released/. See also this thread of another use having the same question: viewtopic.php?f=2&t=987.
3. It is still recommended to work offline since historical data taken from the broker can cause your Tickstory data to be overwritten by the MT4 terminal.
Hope this helps.
Re: Difference in results after new tickdata upload
Hi Danny_pip,
You can just login with blank/incorrect credentials and/or ensure that MT4 is not saving your login details. You will end up with a terminal that reports "No connection" when it starts.
Regards,
You can just login with blank/incorrect credentials and/or ensure that MT4 is not saving your login details. You will end up with a terminal that reports "No connection" when it starts.
Regards,