MODE_TICKVALUE -- LIES!!!! :) - page 6

New comment
 
SProgrammer >>:

Комбат - Вы по расчетам по формуле получили сколько ? восемь с копейками, а МТ вернул одинадцать. 8 и 11 разница есть? Восем правильная а 11 не правильная.



From the post on the third page:

2010.01.13 12:14:42 2008.01.02 08:01 OTestExpert3 USDJPY,M1: MODE_ASK=111.70900000

2010.01.13 12:14:42 2008.01.02 08:01 OTestExpert3 USDJPY,M1: MODE_BID=111.69000000

so the calculation and result are correct,

why those exact prices are there is a tenth question.

 
kombat >>:

Из поста на третьей странице:

так что расчёт и результат правильные,

а почему уж там именно эти цены, вопрос десятый.




How right are they?


:) MT returns 11.

How do you get 11 out of 100/112 ?

:)

 
Posted here.
 
getch >>:
Поместил сюда.

Shit, I've found the "terrible" mistake in MT again.


:) All right, the previous ones were "especially ugly" but nevertheless, I want to apologise in case :) and ask - :) Did everybody realize how much the results obtained all these years were meaningless?

And especially during the optimization. I mean continued, when the results have already been calculated.


But on the whole it's not so bad - it's just that the value of profit in the tester is an abstract thing. :)

But you can't compare one tester run with another by profit! :)

 

This is not an error ( TickValue calculation in the tester).

You can test strategies on a USD account, e.g. GBPJPY, without having history on USDJPY. This has advantages and disadvantages.

In order to correctly calculate TickValue in the tester, it should work on the tick history of several trading instruments.

 
getch >>:

Это не ошибка (расчет TickValue в тестере).

Вы имеете возможность тестировать стратегии на USD-счете, например, на GBPJPY, не имея истории на USDJPY. В этом есть свои плюсы и минусы.

Чтобы корректно считать TickValue в тестере, необходимо, чтобы он работал по тиковой истории сразу нескольких торговых инструментов.


You still don't seem to understand what the mistake is :)

 
You have been given all the arguments and reasons, what comes from where and why.
 
getch >>:
Вам привел все аргументы и причины, что, откуда и почему берется.

I already knew and understood that. I brought it up just to show it was a mistake. :)


:) But what does it have to do with the fact that you probably never understood that it's not what it's supposed to be?

 
And how should the tester from your presentation be for GBPJPY in a USD account, for example?
 

Can I ask a counter question - but is it possible in your idea, to show one BID via marketinfo, and count something there by your own, who knows what and also by the way, variable, but another BID? :) It's a funny way to calculate.


As to your question... well :) 100% it is obvious, that we have to use data of GBPUSD at the moment of test´s start :)

1234567
New comment