Technical analysis. Truth or fiction? - page 11

 
Usual_Trader:

there is no need to rip it out of the cont. 5m to select more optimal points. Not used to throwing away a couple of hundred pips, you know.


So. At 5 minutes any EA will fail, if you've written it of course. And an inverted one will be losing. For a trivial reason: movements are too small compared to commissions. And not because TA does not work.

So Omirbek is Omirbek, but you have to tie up the camel.

 
paukas:

Here goes. You'll have any Expert Advisor on 5 minutes, as long as you've written it, of course.


Why the irony? Do you think that writing in µl is much more difficult than, say, writing the firmware of some real-time device controller? Believe me, no, not if you're not familiar with this area of programming. It's much harder to find and formalise an idea.

Once again, since you liked the idea of lower timeframes - if you haven't noticed, I work on hourlies, with an optimized input-output on lower timeframes.

 
Usual_Trader:

... It is much harder to find and formalise an idea.

...

It is easier to formalise than to find.
 
Usual_Trader:

Why the irony?


It's not irony. I asked you to introduce an advisor who has been losing out on TA. At least one.

You said flat earth, free city and other bullshit.

So what's the conclusion?

Either he is not the ta, or you have not written any advisor, or one of the two.

 
paukas:

It's not irony. I asked you to introduce an Expert Advisor that was losing money on TA. At least one.

You responded with flat earth, free city and other nonsense.

So what's the conclusion?

Either he is not the ta, or you have not written any advisor, or one of the two.


Well, I will throw out some code, on the intersection of stochastics or masks, or on a mcd, amplified by RSI boundaries and bolinger bars, one of many experiments, what will change from that? Just to give someone a reason to troll for style of writing, lack of comments, etc.?

There's tons of them in the kodobase, take your pick. That's not what this is about.

I've already said my opinion - I use EAs with TA analysis in some timeframes, predicting, or rather trying to predict the market situation based on other market indicators and using several EAs with different algorithms.

 
moskitman:
It is easier to formalise than to find.


I agree, which is why I put "find" in the first place. But the truth is, it depends on whom - I have a friend who is a fairly successful and stable trader, and he cannot formalise his sense of the market in any way. It doesn't work, he's not a mathematician, he's a creative person.
 
Usual_Trader:

Well, I will throw out some code, on the intersection of stochastics or masks, or on a mcd, amplified by RSI boundaries and bolinger bars, one of many experiments, what will change from that? Just to give someone a reason to troll for style of writing, lack of comments, etc.?

There's tons of them in the kodobase, take your pick. That's not what this is about.

I've already said my opinion - I use EAs with TA analysis in some timeframes, predicting, or rather trying to predict the market situation based on other market indicators and using several EAs with different algorithms.


You can make a profitable one out of any of your faulty EAs if you add TA element and throw out the bullshit.
 
Usual_Trader:

I agree, which is why I put "find" in the first place. However, it depends on whom - I have a friend who is a fairly successful and stable trader, and he cannot formalise his sense of the market in any way. He cannot, he is not a mathematician, he is a creative person.
You can't formalise your gut, that's a fact.
 
paukas:

You can make a profitable one out of any of your draining EAs if you add a TA element and discard the bullshit.

OK, prove it in practice. Take any of the failed EAs from the base based on TA and make a profitable one, based on the results of several thousand trades.
 
Usual_Trader:

OK, prove it in practice. Take any drainable TA-based EA from the base and make a profitable one, based on the results of several thousand trades

I need yours. Not from kodobase. This is essential.