Absolute courses - page 62

 
Dr.F.:

The system is simple: open trades with TP=SL. This and only this. The algorithmic core can be anything. Including the one you are demonstrating now with the Expert Advisor.

If your EA will show profit in THIS experiment geometry, then it's good. If not - sorry, it will not show profit in any other experiment geometry either (stably meaning).

Afftar, justify it. I do not believe it. I'm an unbeliever.

 
By the way, Dr.F, it's just that with sl=tp it's somehow more beautiful (because you can compare it to a coin) and that's all. And if I have 3*sl=1*tp and P(sl)=P(tp), is it already bullshit?
 
I've added a signal, so I'll familiarise myself with the service.https://www.mql5.com/ru/signals/4881
 
DYN:
Yes, by the way! Dr.F, it's just that it's prettier (because you can compare it to a coin), that's all.
It's not about prettier. TP=SL is a good and simple assessment of TC inputs and nothing more... And the topicstarter, in his usual way, makes unchildish generalizations and far-reaching conclusions.
 
Figar0:
And the topic-starter, in his usual manner, makes unchildish generalisations and far-reaching conclusions.
This is one way of making the wishful thinking real! However, the power of "word and thought" must not be childish...
 
Figar0:
It is not about beauty. TP = SL is a good and simple estimate of TC inputs and no more... And the topicstarter, in his usual manner, makes unchildish generalizations and far-reaching conclusions.

Such an assessment? ? kn of profitable / kn of losing trades ? What's wrong with (number of profitable *tp )/ (number of losing trades *sl)? - Same thing = profitability with a fixed lot. (slippage is neglected).

And the Recovery Factor, Mo and other characteristics of the TS adequately show the situation with whatever ratio of slop/bottom.

 
DYN:

Such an assessment? ? kNo profitable / kNo losing trades ? What's wrong with (number of profitable trades *tn )/ (number of losing trades *sl)? - Same thing = profitability with a fixed lot. (slippage is neglected).

And the Recovery Factor, Mo and other characteristics of the TS adequately show the situation with whatever ratio of slop/bottom.

TP=10P, SL=1000P. there are 1000 trades, all with TP, but each trade was first at 900P loss, and then closed at TP. A good entry? If we had enters in the opposite direction, with the same TP and SL, all 1000 trades would have closed on TP in the same way. What are you going to calculate with formulas like this?

And FS, MO, PF are rather characteristics of TS than evaluation of its input.

 
Figar0:

TP=10P, SL=1000P. There are 1000 trades, all with TP, but each trade first went to a loss of 900P and then closed at TP. A good entry? If we had enters in the opposite direction, with the same TP and SL, all 1000 trades would have closed on TP in the same way. What are you going to calculate with formulas like this?

And TP, SL, FF are rather characteristics of TS, than evaluating its input.

For TP=SL we need the least amount of trades for statistical validity. If TP=2SL or 2TP=SL, then we need 2 times more trades for the same level of statistical validity.

Of course, the ratio TP and SL is determined by system logic, and not assigned from above.)

P.S., although it still depends on what stat.indicator system evaluation is taken into account. For some we need more trades, for others less.

 
Doctar F, do not take the advice to run in the tester (although it's not easy) as a reproach for the robustness of your algorithm, it's not a reproach that it's bad, or rather your inputs or whatever, the tester is not only to check robustness, it is also mm, how will you tweak mm without having a normal depth of trade statistics. And typing it in live will make your hair turn gray. I mean a long time.
 

I decided to look at the performance of the system with SL=TP and number of positive inputs =65%.

on 11 crosses the picture is more cheerful.

P.S I will try at my leisure to run this circuit with my kernel, which show the ratio of positive inputs to negative inputs = 85.