FOREX - Trends, forecasts and implications 2016 - page 726

 

http://ec-analysis.livejournal.com/65556.html

«КРИЗИС, Однозначно!» ©Миллиардеры
«КРИЗИС, Однозначно!» ©Миллиардеры
  • 2016.08.18
  • wallet.livejournal.com
Сегодня я опубликую предпоследний, но самый интересный пост про грядущий глобальный финансовый кризис. Завтра заключительный пост касательно ситуации по России. Но сегодня самый увлекательный пост для меня. А дальше я закрываю тему «Кризис в мире». Про грядущий кризис я писал почти все предыдущее время – полгода, достаточно почитать мои...
 
new-rena:

In short.

The indicator visually displays the operation of a formula and allows me to test its performance on the price history. After that, if it works successfully, I only need a formula for my robot.

EXAMPLE:

FIBO - % of current price

MA = sum of N ticks/N

This is how I keep the indices, but usually only 1% of all the good ones I test. There are no FIBOs or MAs in the collection.

What do you mean? What is it? Like a mathematical tester?
 
azfaraon:
What do you mean? What is it? Like a mathematical tester?
I take an indicator and create an Expert Advisor for it. Then I run it in the strategy tester and change the parameters.
 
new-rena:
I take an indicator and make an Expert Advisor for it. Then I run it in a strategy tester and change its parameters.
You won't believe it - that's how everyone does it
 
new-rena:
I take an indicator and make an Expert Advisor for it. Then I run it through the strategy tester and change its parameters.
Question, what tester report results tell you if an indicator is good or not? There are no clear grounds for sifting out Expert Advisors and Inductors so far.
 
My Expert Advisors in the Strategy Tester have earned thousands and tens of thousands of% over a long period of many years, and what's the use, they were losing in real trading, so if you make a robot, you should test it only on a real chart.
 
Dimmerd:
My Expert Advisors in the Strategy Tester have earned thousands and tens of thousands of% over a long period of many years, and what's the use, they were losing in real trading, so if you do a robot, you should only test it on a real chart.
They were busy fitting it, of course...
 
azfaraon:
Because they were busy adjusting, of course...
I mean that I don't see the point in the tester at all, if you do it, it's long and tedious to observe the bot on a real chart, analyze and correct it, and I think this process is endless, but you have to be able to write the code yourself.
 
Dimmerd:
I'm saying that I don't see the point in the tester at all, if you practice, it's long and tedious to watch the bot on a real chart, analyze and correct it, and I think this process is endless, but you have to be able to write the code yourself.
The tester is like a training apparatus where you can practice variations of techniques as an athlete, but of course it's not the fact that you'll end up being an athlete, and especially a champion on the tatami.
 
Dimmerd:
I mean that I don't see the point in the tester at all, if you do it, it's long and tedious to observe the bot on a real chart, analyze and correct it, and I think this process is endless, but you have to be able to write the code yourself.
The tester is designed primarily to verify the correctness of the algorithm and detect errors, not for estimating potential profit. Of course, there are nuances when the result of testing almost coincides with real ones, minus slippage, delays in execution and spread widening.
Reason: