[WARNING CLOSED] UmnickTrader Adaptive EA - page 10

 
VictorArt: Why so many words? Suffice it to say that you cannot create an EA that passes the OOS test of 9 years.

To be honest, I don't fucking need one with parameters like yours. I don't fight windmills, so I don't create such monsters ))))
 
VictorArt: passed the OOS test at 9 years of age.
An OOS test at 9 years does not guarantee future earnings, nor does an OOS test at 1 month. So why fight for something you don't understand when you can do more realistic things?
 
VictorArt:

Why so many words? Suffice it to say that you cannot create an EA that passes the OOS 9 year test.
It's not addressed to me. But it's not personal, so I'll respond. I have no idea whether I can create an EA, which would have passed the test OOS 9 years. Because I don't know why I need it.
 
LeoV:
An AOS test at 9 years of age does not guarantee future earnings, nor does an AOS test at 1 month of age. So why fight for something you don't understand when you can do more realistic things?


What do you mean by "more real"? With a guarantee of future earnings? :)

And it is quite clear what to fight for - basic science, without which the applied science is "neither here nor there" :)

 
tara:
It is not addressed to me. I have no idea what to do with it. I have no idea whether I'll be able to create an Expert Advisor, which would have passed the test OOS 9 years. Because I do not understand why I need it.


The more OOS, the more likely that the EA won't stop working right at the moment when you want to make some money. OOS is what? It's just a test. The more any product is tested - the better for the consumer.

But here we still need to see what this OOS will look like. There is an EA of 5 lines, which without any training (one solid OOS) is not sinking on the major forex pairs since the flood, and in some places even gives a noticeable profit.

I also have more serious EAs which, with proper settings, are likely to pass this test 1 to 9. But this will most likely be achieved at the expense of the possible efficiency of the EA, and I am not going to trade it with such settings. After all, we need much at once, don't we?) Why not as a test? It has the right to live.

 
I can write you an EA that will pass the test in 20 years ! with minimal drawdown and wild profits ! ( Only on the real will lose the deposit in a week ... )
 
Figar0: The bigger the SOC, the more likely it is that the Expert Advisor will not stop working the moment you want to make money with it.
That's not a fact. The more OOS, the more likely it is that it will fail on the real account.)))
 
VictorArt: "More real" - how do you mean? With a guarantee of future earnings? :)

They only give guarantees in the morgue. And here, it is much more realistic to find a working market model on a shorter timeframe than on a longer timeframe with unclear parameters ))))
 
atik:
I can write you an EA that will pass the test even in 20 years ! And with minimal drawdown and wild profits ! ( Only on the real will lose the deposit in a week ... )


Using the features of a tester is not serious.

For starters, just write one that shows the best results, and we'll all take it apart together :)

 
LeoV:

Guarantees are only given in the morgue. And here, it is much more realistic to find a working market model on a shorter time horizon than on a longer one, but with incomprehensible parameters ))))


The future cannot be predicted - this is an axiom (for me) from which I draw all other conclusions.

So your phrase, from my point of view, is absurd, i.e. you will inevitably lose more in the long term than you will gain in the short term.