Optimisation results differ from single tests on them - page 7

[fxsaber]

should be added as a line in the log. Such a solution would still help with feedback to find the problem.

[Михалыч Трейдинг]

I've already posted in the latest version thread, but since I found an old thread on this topic, I'll repeat it here as well - this problem has also affected me.

I run a single test from a fresh optimization window and get completely different results than the optimizer gave me. The difference is not "trivial" but "purely specific". At the same time some of the single tests coincide with the optimizer's ones. Its behavior is unpredictable - now and then I like it both at different optimizations and at different results of one optimization.

I don't use clouds, I've cleaned the cache, orders are modified on bar opening...

[Vladimir Karputov]

Михалыч Трейдинг:

I've already posted in the latest version thread, but since I found an old thread on this topic, I'll repeat it here as well - this problem has also affected me.

I run a single test from a fresh optimization window and get completely different results than the optimizer gave me. The difference is not "trivial" but "purely specific". At the same time some of the single tests coincide with the optimizer's ones. Its behavior is unpredictable - now and then I like it both at different optimizations and at different results of one optimization.

I don't use clouds, I've cleaned the cache, orders are modified on bar opening...

You should first define: Are you using MQL5 or MQL4 for testing? Your starting to flood all the sections and both terminals with no proof. This is a technical forum - without a minimum description - you'll be completely ignored. In severe cases - there will be a reading mode.

[Nikolay Kuznetsov]

Vladimir Karputov:

You first determine: Are you running tests on MQL5 or MQL4? Otherwise, you start to flood all the sections and both terminals with no proof. This is a technical forum - without a minimum description - you will be completely ignored. In severe cases - there will be read mode.

I have spent today searching for an error in my EA code - same problem, then the same result of single test coincides with the optimization result, then it is different. Moreover, the amount of CPU load during optimization is now much higher than in the previous build. Of course, we are talking about MT5.

Ps got to this thread through Yandex search and only then I found out that the thread is about mt4...

[Михалыч Трейдинг]

Nikolay Kuznetsov:

Ps Got to this thread through a Yandex search and only then did I find out that the thread is about mt4...

And in the top post it says " New version of MetaTrader 5 platform "

The title of the thread doesn't specify the terminal - so anyone could easily get screwed, not to kill for it.