Signs of a REAL system - page 4
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
))) In that case...
In principle, it makes no difference. The essence of the message here is a significant difference in the result. Hence, TC is extremely sensitive - bad.
I disagree.
Firstly, there was no mention of an extreme difference. Hence, there is no reason to claim EXTREMELY sensitive either.
Secondly, the pattern detected at the optimization site could have been more frequent at OOS.
So I consider the better results at the OOS site to be a good sign.
All IMHO.
I think that these are the signs of TS intended for sale, a kind of universal machine for making money.
The main thing is not to be in a hurry and not to be greedy.
Betting few positions? But it is reliable, lots more will be taken. The Expert Advisor knows 2*2=4 and waits for these 2*2.
TS and SL are fixed, how else could they be? As soon as you open an order and the computer crashes, you are not around for 2 days.
I agree that there should not be much optimisation, and only for a particular pair.
Sta2066 писал(а) >>
... For example, it works well in the flat, but it is losing in the trend: So let it work, there is another one for the trend. The main thing is not to rush and not to be greedy....
You do not know when the flat ends / trend starts and vice versa.
Then you don't need any TS.
I think that these are the signs of TS intended for sale, a kind of universal machine for making money.
The main thing is not to be in a hurry and not to be greedy.
Betting few positions? But it is reliable, lots more will be taken. The Expert Advisor knows 2*2=4 and waits for these 2*2.
TS and SL are fixed, how else could they be? As soon as you open an order and the computer crashes, you are not around for 2 days.
I agree, there should not be much optimisation, and only for a particular pair.
It is my understanding that the TC is not considered a fit if the owner clearly knows the limits of its use, and does not test beyond those limits.
I disagree.
Firstly, there is no mention of strong differences. So there is no reason to claim extreme sensitivity either.
Secondly, the pattern identified at the optimization site may have been more common at OOS.
So I think the better results at the OOS site are a good sign.
Everything is IMPOrTANT.
goldtrader wrote >>
Firstly, it was mentioned.
Secondly, I agree: better is not worse. Provided we are not comparing God's gift with an egg. I.e. see first.
TS and SL are fixed?
Fixed == value is set by the user in pips via external parameters, not calculated by an expert based on some market data.
I agree that there should not be much optimisation, and only for a particular pair.
goldtrader wrote >>
Firstly, it was mentioned.
I mentioned strong differences for the worse, joo asked about the differences ("strong" was not mentioned, see previous page) for the better.
Not the point in general. Any difference on the OOS for the better is only positive and not a fitting.
goldtrader: результаты на другом ТФ сильно отличаются (в худшую сторону) относительно базовых,
I don't get it. explain the connection with the fit.
Roughly and briefly, if the TS successfully operates a certain pattern, for example on H1, then the same pattern should (probably a little worse) work on M30, and on H4.
Let us take for example a classical TA. Not a single textbook says that figures, lines and levels work only in a certain TF. Suppose the lower TF is more noisy and the TS works worse there,
But on neighbouring TFs compared to the base one (where the pattern has been found) the TS should at least not lose.
I mentioned strong differences for the worse, joo asked about the differences ("strong" was not the case, see the previous page) for the better.
Not the point in general. Any difference on OOS for the better I consider only positive and I don't attribute it to fitting.
Not the point, but the question about differences for the "better" was addressed to you and your post, which was about "strong differences". Meaningful inheritance. OK, really not the point.
I disagree with "any" though. In fact, imagine that the test with the best result you did earlier than the one with the worst. Observation time has nothing to do with it - that's what I mean.
And what was good and showed better somewhere is, well, okay. Or vice versa. Most importantly, not by orders of magnitude - it makes you wary.