Optimise an EA and get the best of the optimised ones. - page 38

 
Fast235:

the point is that you need 1-2 clearly substantiated indicators on the chart, rather than optimising a bunch

Erm... I don't quite get it.

I use only one indicator in any TS - either EMA or PriceChannel.

And the daily volatility is calculated as ATR(25) on D1.

Optimization of any TS is just to set the "speed" (the period of the indicator), and set the optimal indentation from the indicator readings for entry.

The "heap" is about the systems themselves, which operate on different principles, and my main idea is that no matter how the symbol behaves - some systems will definitely work in it. Optimisation is required in order to select the best settings for each system, but each one costs a little. The main "thing" about the League is precisely its "breadth" - there is always plenty to choose from.

Two months' experience shows that this is a correct assumption. At the moment I'm working out the principles of TC selection. The parameter "quality" reflects the "beauty" of the balance curve quite well, but it reflects the stability of the TS very poorly, and this is no less important criterion for the selection of a TS for trading.

 
Georgiy Merts:

Erm... I don't quite get it.

I have only one indicator in any TS - either EMA or PriceChannel.

And the daily volatility is calculated as ATR(25) on D1.

Optimization of any TS is just to set the "speed" (the period of the indicator), and set the optimal indentation from the indicator readings for entry.

The "heap" is about the systems themselves, which operate on different principles, and my main idea is that no matter how the symbol behaves - some systems will definitely work in it. Optimisation is required in order to select the best settings for each system, but each one costs a little. The main "thing" about the League is precisely its "breadth" - there is always something to choose from.

Two months experience shows that this is a correct assumption. At the moment I am working on the principles of selecting the TS. The parameter "quality" reflects the "beauty" of the balance curve quite well, but it reflects the stability of the TS very poorly, and this is no less important criterion for the selection of the TS for trading.

there is no working volatility indicator! example:

Flag example, forms quickly and clearly

 
Fast235:

there is no working volatility indicator! example:

Flag example, forms quickly and clearly

Well... I don't know, I'm quite happy with ATR, especially since I take it from a month's worth of days. I don't need "fast volatility readings" - I need long term volatility, and it's fairly stable and changes slowly.

 

But it reflects very poorly on the stability of the TS

everyone has this problem, you never know in 2 or 3 months, the backtest will average and the spread and slippage will kill

 
Georgiy Merts:

Well... I don't know, I'm quite happy with ATR, especially since I take it from the days of the month. I don't need "fast volatility readings" - I need long term volatility, and it is quite stable and changes slowly.

you can also take the MA monthly, it also takes a long time to change)


I can't offer anything new, 2-3 hour calculations I can help in the tester with my iron

 

I think from Monday - I will open a signal on the best TCs in the League.

The signal will be free for now.

I have not yet decided on the methods of selecting TS, and so far they are intuitive.

When these methods are worked out - probably, the signal will be paid, but by my estimation, 10 XML-files optimization - will be "worth" a monthly subscription.

 
Georgiy Merts:

I think from Monday - I will open a signal on the best TCs in the League.

The signal will be free for now.

I haven't decided on the methods of selecting the TS yet, and so far they are intuitive.

When these methods will be worked out - probably the signal will be paid, but I think 10 XML-files of optimization will be "worth" a monthly subscription.

This is great news!

I hope the signal will qualitatively change the attitude to this branch!

 
Aleksey Vyazmikin:

This is great news!

I hope the signal will qualitatively change the attitude of this branch!

Not sure... At least until I learn to choose those TS, which show good results...

By the way, it was an RTS-system of reverse trailing stop-loss... That's why I don't like these systems... They are good on the flat, but the first good trend - they just explode...


But we'll see.

 
Georgiy Merts:

Not sure... At least until I learn to choose those TS, which show good results...

Otherwise - you see yourself, now entered the slump - due to just one unsuccessful TS. By the way, it is exactly the RTS-system of reverse trailing... That's why I don't like these systems... At flat they are good, but the first good trend - they just explode...


But we'll see.

In any case, it's more interesting to follow the movement in real time!

And the rest is up to us.

 

Hmmm... Only signals from demo accounts can be free...

And I, alas, have all the "slots" on the UPU already occupied, and I can't put another terminal for a demo signal...

It seems that arranging a free subscription for those involved in over-optimization - won't work yet either.

I'll have to think about how to solve this problem...