For those who are convinced that all EAs with a martin are losing out. - page 10

 
tol64:
By the way, how long does the test last (one pass during optimisation) ?
I haven't timed it, but it feels like 15 min.
 
khorosh:
Didn't time it, but it feels about 15 min.

So you should go straight to the cloud. )) Don't you want to try to harness the power of cloud computing?

Why such a long test? Do you use pending orders (grid)?

 
tol64:

So you should go straight to the cloud. )) Don't you want to try to harness the power of cloud computing?

Why such a long test? Are pending orders (grid) being used?

They are used and the computer is weak (netbook). I do not want to try it because I see no reason to optimize the Expert Advisor on the whole history. If the Expert Advisor works only on the optimized area, it is of little value. I think 3 (or less) years is enough for optimization, and if the Expert Advisor does not work on the entire history after such optimization, then it's better to discard it.
 
khorosh:
They are used. I do not want to try it because I do not see the point in optimizing the Expert Advisor on the entire history. If the Expert Advisor works only on the optimized part, it is of no value. I believe 3 (or less) years is enough for optimization, and if the Expert Advisor does not work on the entire history after such optimization, it should be discarded.

So we were talking about optimizing parameters without the included block with martin, and that only if the result is negative.

Here is another interesting test, try the same parameters (even with martin), but on a different symbol. For example, AUDUSD or GBPUSD. Also on the entire history.

 
tol64:

So we were talking about optimizing parameters without the included block with martin, and that only if the result is negative.

Here is another interesting test, try the same parameters (even with martin), but on a different symbol. For example, AUDUSD or GBPUSD. Also on the whole history.


I tried the same parameters with GBPUSD, on the entire history there were 2 or 3 points of loss. The duration of optimization does not depend much on the lot size (it depends on the number of orders).
 
khorosh:

I tried it with the same parameters onGBPUSD, on the whole history there were 2 or 3 points of loss.

Here we have already 2-3 points. Of course it would be desirable to look at the same set of parameters on many symbols. Then the picture will be more complete.

The optimization duration does not depend much on the lot size (it depends on the number of orders).

I did not write that in relation to the duration of optimisation, but in relation to the result without martin. But I agree that this is too much to ask. )) So it's just interesting to see the result with the same parameters, but without martin.

 
khorosh:

There's nothing for you to do in this thread, you've grown out of your kid's trousers and you know everything about martin perfectly well).

Although, if you haven't told me everything yet, go ahead. I will listen to you carefully.

I also know a lot about loki. Would you like some?
 
khorosh:

I tried it with the same parameters on GBPUSD, and on the whole history there were 2 or 3 points of loss. The duration of optimization depends only slightly on the lot size (it depends on the number of orders).
For an intelligent person, which we all consider you to be, this should be quite enough to understand the futility.
 
DhP:
To an intelligent person, which we all consider you to be, this should be enough to understand the futility.

And that an earning expert should earn on all symbols without changing parameters? Have you seen such a thing anywhere?
 
moskitman:
I also know a lot about loki. Would you like some?


Proloc is bullshit, it's all about the epilocke.
Reason: