The miracle of the grail... Myth or reality?! - page 11

 
khorosh: Just kidding, of course, there is still work to be done ... and the process is never-ending.

I think this is also the right approach. choose a good system and continually improve it, rather than jumping from one system to another.

 
multiplicator:

I think this is also the right approach. choose a good system and continually improve it, rather than jumping from one system to another.

And what is "improving" the system?

I have observed a constant "addition of crutches", an increase in the number of parameters that improve the system's performance on history, but worsen its performance on real time. As I see it, systems should not have many "degrees of freedom"; there should be as few parameters as possible.

 
multiplicator:

I think this is also the right approach. choose a good system and continually improve it, rather than jumping from one system to another.

I second that.

Ironically, the Grail is not alone in the market. There are an infinite number of them. And the thesis - "everyone has his own Grail" is quite true.

 
Georgiy Merts:

And what is the "improvement" of the system?

According to my observations, it is a constant "adding crutches", increasing the number of parameters - which improve the system performance on the history, but worsen the performance on the real site. It seems to me that systems should not have many "degrees of freedom", there should be as few parameters as possible.

I already said - for most TS, the only parameter that needs to be optimised is the size of the moving observation time window. Ideally, it should be self-tuning to the changing nature of the market. That's it.

 
Alexander_K2:

As I said before - for most TS, the only parameter that needs to be optimised is the size of the moving observation time window. Ideally, it should be self-tuning to the changing nature of the market. That's all.

Do you understand what you're saying? The market doesn't even care about what it was 5 minutes ago. It's thinking about the future.) I'd say it's gone into the future.)

 
Yuriy Asaulenko:

Did you understand what you said? The market doesn't even care about what happened five minutes ago. He's thinking about the future.) I'd say it's gone into the future.)

Nah. It's like in life - "here I remember, here I don't remember". Actually, the market is life.

 
Vitaly Muzichenko:

Sometimes you can trade pullbacks and sometimes you can trade reversals. The trend tends to reverse)

This is the same deal as the previous one, also by the system, and exit by the system as in the previous case, only there was a minus on entry, and here a plus, but both deals are simply excellent


What is not good in this system?

P.S. The stop was still the same ~7p

Not a bad entry and exit, author's TS?

 
Georgiy Merts:

And what is "improving" the system?

From my observation - it's a constant "adding crutches", increasing the number of parameters - which improve the performance of the system on the history, but worsen the performance on the real. It seems to me that systems should not have many "degrees of freedom", there should be as few parameters as possible.

This is where you confuse the concepts... :-)

The more degrees of freedom, the fewer the parameters...

More precisely, the fewer parameters, the more degrees of freedom the TS has

 
Alexander_K2:

Seconded.

Oddly enough, the Grail is not alone on the market. There are an infinite number of them. And the thesis - "everyone has his own Grail" is quite true.

Absolutely right!

 
Alexander_K2:

As I said before - for most TS, the only parameter that needs to be optimised is the size of the moving observation time window. Ideally, it should be self-tuning to the changing nature of the market. That's it.

+/- 20% of a reasonable value selection (based on daily calendar periods) does not make a difference.