Retraining - page 5

 
Yousufkhodja Sultonov:
If the fitting or re-fitting is done on all the historical data, it only improves the quality of the TS.
Deception (self-deception) because it occurred on all data does not stop it from being a deception. Even worse. Because in order to expose it, it must exist as long in the future as it does in history.
 
Youri Tarshecki:
Deception (self-deception) because it occurred on all the data does not stop being a deception. It is even worse. Because in order to expose it you have to exist as long in the future as in history.
I am optimising from early 1973 to the present on TF D1, which is about 11000 bars of history, do you think there will be an event in the near future that hasn't happened in the previous 43 years? I think you are wrong.
 
Yousufkhodja Sultonov:
I am optimising from early 1973 to the present, do you think there will be an event in the near future that the previous 43 years have not seen? I think you are mistaken.

No, you are the one who is wrong.

One should model the future, not be sure of it.

Optimise for 40 years and then run it through the remaining 3 years. Did the error match? If the error matched, that's fine.

 
СанСаныч Фоменко:

No, you are the one who is wrong.

One should model the future, not be sure of it.

Optimise for 40 years and then run it through the remaining 3 years. Did the error match? If the error matched, it's fine.

You have to live forever and have unlimited capital to invest on a horizon of 3 years and still have the right to make a mistake. I thinkYousufkhodja Sultonov, this is a sincere delusion, just really does not understand what it means to check on a non-optimised area and why it is necessary
 
СанСаныч Фоменко:

No, you are the one who is wrong.

One should model the future, not be sure of it.

We optimise for 40 years and then run it through the remaining 3 years. Did the error match? If the error matched, that's fine.

Period from 1973 to 2013:

Period from 2013 to 2016, the settings are the same:


 
Yousufkhodja Sultonov:

The period from 1973 to 2013:

Period from 2013 to 2016, the settings are the same:


I tell you-it's genuinely misunderstood.
 
Youri Tarshecki:
In order to invest on a horizon of 3 years and still have the right to make a mistake one has to live forever and have unlimited capital. It seems to me thatYousufkhodja Sultonov, who is sincerely deluded, just does not really understand what it means to check on a non-optimized site and why it is necessary
Of course, I understand much less than you do, but it is enough to state that you are sadly mistaken.
 
Youri Tarshecki:
I'm telling you - it's a sincere misunderstanding.
What is the "misunderstanding"? Is it that I am convinced that it is impossible to make a stable profit on Forex in the short term? Then you're right. You can continue to live in a world of illusion.
 
Yousufkhodja Sultonov:
What is the "misunderstanding"? Is it that I am convinced that it is impossible to make a stable profit in Forex in the short term? Then you are right. You can continue to live in a world of illusion.
The misunderstanding is that you're not saying - I've had these setups for 40 years and lo and behold, they work on the last non-optimised three as well. You talk about the same settings. But you don't say how you got them. Judging by your products and posts - you just optimised the whole story. And this is the fitting. Just do a wolfing forward of your EA at least in 10 year optimisation - 5 year check mode and then you will see the sad result yourself.
 
Youri Tarshecki:
The misunderstanding is that you're not saying - I've had these settings for 40 years and lo and behold, they work on the last unoptimised three as well. You are talking about the same settings. But you don't know how you got them. Judging by your products - you just optimised the whole story. And this is a fitting.
What is your problem with this kind of "fitting"? For 40 years you got MO = 6.02 points/day, profitability = 2.06. For the last 3+ years, MO = 7.26 and Pr = 2.66. Figures of the same order of magnitude.
Reason: