Counsellors to whom. Lots of them and for free! - page 3

 
I've also had EAs come in that I didn't like. I just don't save them, but I've taken a couple of them to my liking. I came to the conclusion that the strategy generated by SSB should show profit on history on several pairs. Plus we can combine these strategies in one strategy, dividing the risk at the same time. What do you think of this method?
 
rider >> :
..........
4. Why do different MACD period values appear when selecting strategies for different symbols and timeframes? ..... neither in the code, nor in the same-name parameter file (set), does it seem that they can be changed?

Thanks for the answers, special "thank you" for the idea ...... one question remains unanswered, if you can clarify, plz.....

It is clear that all of us have our own understanding of optimization, as a process..... not to spite you, but as "mine" - I do not like that positions, when optimizing, are opened by a signal from a zero bar, and at the price of its opening....

One more important thing: if your Creator could be changed a little bit "for your own needs", then you can get almost ready code of Expert Advisor without any add-ins, with possibility to choose the best strategy out of plenty of ones generated, to test it with OOS, both before and after optimization...... and only to cut off unnecessary things ))

Once again: the idea, approach and implementation in the code is great!

 
rider >> :

There is one more important thing: if you change your Creator a little bit "for yourself", then you can get almost ready code of Expert Advisor without any add-ons, with possibility to choose best of many generated strategies, to test OOS, both before and after optimization......


I will repeat for especially gifted ones and in language, which is clear to those gifted ones: After selection of strategy with changed Creator, you will get practically ready-made code of nonsense. It is easy to check in the tester. It won't give out errors, but it won't give the results that were demonstrated before saving either.
 
Reshetov >> :
I will repeat for talented ones: after selecting strategy with changed Creator you will get practically ready-made code of nonsense. It is easy to check in the tester. It will not give out errors, but it will not give the results that were demonstrated before saving either.

This is for "especially gifted" ..... and I am not talking about code generation - I refused..... it is much more convenient to have a ready-made code at once, with the possibility of meaningful choice of both strategies and OOS periods, without any manipulations with history. Repository is an interesting thing, of course, but on a good computer all plenty of strategies in all TFs and symbols, with possibility to add indicators according to one's "taste and colour" is computed within a day - then comes routine, not automated yet: forwards, best selection, optimization, forwards again, best selection from optimized ones and only then long-awaited DEMO :) ......

Agree that there will always be some element of distrust of "someone else's" testing, and the code that your add-in gives, even on demo without rework, cannot be put. But the idea is good, thanks again and....... this is the second time you have walked away from my fourth question :)

 
danja писал(а) >>
I've been getting EAs I didn't like either. I just don't save them, but I've taken a couple of them to my liking. I came to the conclusion that the strategy generated by SSB should show profit on history on several pairs. Plus we can combine these strategies in one strategy, dividing the risk at the same time. What do you think of this method?

Yes, the idea lies on the surface. I have already implemented such a thing, but with my own strategies. Yesterday or today one such Expert Advisor has made +3 deposit on demo. :)

But it is dangerous to implement it on real account anyway, imho.

As for EAs that are not liked - this is not a reproach to SSB, I'm just interested in the selection / elimination criteria. Perhaps they remain because they make profit on other TFs / symbols, and on the "main" one they simply don't plummet...

 
rider >> :


Agree that there will always be some element of mistrust in "other people's" testing, and the code that your add-in gives, you can't even put on a demo without rework.

What you all time speak nonsense? What kind of testing, where do you see it? You seem to be paranoid in a way.


Only the ranking of strategies in the repository database is performed by distributed computations. What the end user receives is the results of testing on his/her computer and on his/her quotes.


If you do not trust your computer, who is stopping you from double-checking the obtained Expert Advisor in the Strategy Tester?

 
Free Cheetah 2.5 here!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
 
sv-show >> :
Free Cheetah 2.5 here!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Ah there he is!!! Don't show it to anyone!!!

 
sv-show >> :
Free Cheetah

In a crisis, less than an elephant or at least a hippopotamus is not an option! :)

 
Reshetov >> :

Why are you always talking nonsense? What kind of testing, where do you see it? You seem to be paranoid.


Distributed computing performs only the ranking of strategies in the repository's database. What the end user receives is the results of testing on his/her computer and on his/her quotes.


If you do not trust your computer, who is stopping you from double-checking the Expert Advisor in the Strategy Tester?

Do you talk to everyone like that? ..... as if you were the ultimate truth.

The question is rhetorical, you don't have to answer it.