Optimise an EA and get the best of the optimised ones. - page 31
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
Outgoing TCs in need of re-optimisation at the moment:
I will put the latest CADCHF ChnTrendRTS to optimize - so that CADCHF has a complete set of TS.
And the question of choice is still very relevant.
Havethe selectioncriteria been defined? What are they?
If you have the passwords for investments from both demo and real accounts, Alexey ! Have a look.
The result across the league? Of course it's negative - because the losing periods for any TS are longer than the earning periods, and all the TS work there, regardless of their results.
The result of the best ones I have chosen ? Slightly positive. I think if it goes like this - I will open a signal in a week.
I think it's not just me who's interested in this question, especially given your mindset that I read in the posts...
I hope there will be a signal, so that the result of this activity can be seen immediately.
And the question of choice is still very relevant.
I've been working a lot with counter-trend systems, I automate the selection there.
If the main indicator for counter-trend or flat systems is a recovery factor, then for other systems (where there is no risk of over averaging) the most important indicator is profitability. I'm also looking at Ksharp - in general it gives a good sample.
CADCHF ChnFlatSP
Taken into account.
33 regcodes.
Havethe selectioncriteria been defined? What are they?
There are no criteria! That is also the problem.
An algorithm for assessing the quality of trade has now been found and accepted for use. I think it is quite adequate.
But let us take two TS with similar quality parameters. Suppose AUDCHF EMAFlatRTS and AUDCAD EMATrendSP.
Both of them have made 15 trades on Demo. And they both showed similar results, quality 130%, balance $10.
But, I suspect that their resilience will be radically different.
The problem is exactly in the word "suspect". The choice between these TSs is completely intuitive for me now, and I don't like it.
Here, let's say another TS on the real has died. I have to replace it. The best on the demo these two TS. Which one to choose and why ?
And what concerns trend or flat systems - here it is a bit more complicated, if for contrend system the most important indicator is recovery factor, then for other systems (where there is no risk of overaverage) the most important indicator is profitability. And I look at Ksharp - in general it gives a good sample.
Both recovery factor and profitability are considered in the "quality index". And, of course, if the quality of systems is different, then the one having the better trading quality will be used. But I have so many systems, that even the best ones, who have been working on demo for some time and have good results, have significantly different systems. So the question is about choosing between such systems, which show approximately the same quality, but are based on completely different algorithms and work on different symbols.
I'm betting on my overoptimisation:
Another TC for optimisation:
The recovery factor and profitability are both taken into account in the "quality indicator". And, of course, if the quality of the systems is different, then the one with the higher quality of trade will go into business. But I have so many systems, that even among the best ones, who have been working on demo for some time and show good results, the systems are essentially different. So the question is about choosing between such systems, which show approximately the same quality, but are based on completely different algorithms and work on different symbols.
Then we should just divide the lots between them, and leave them all to trade.
There are no criteria! That is the problem.
Now an algorithm for assessing the quality of trade has been found and adopted for use. In my opinion, it is quite adequate.
But let's take two TS with similar quality parameters. Suppose AUDCHF EMAFlatRTS and AUDCAD EMATrendSP.
Both of them have made 15 trades on Demo. And they both showed similar results, quality 130%, balance $10.
But, I suspect that their resilience will be radically different.
The problem is exactly in the word "suspect". My choice between these TCs is now completely intuitive, and I don't like it.
Here, let's say another TS on the real has died. We must replace it. The best on the demo these two TS. Which one to choose, and why?
So we need to introduce a measure of this very resilience that (to begin with) would at least approximate what the intuition selects.