Advice on not using the MetaTrader 4 Strategy Tester - page 6

New comment
 
Renat:
You forgot the more conceptual rule of thumb (for those who have actually spent many years testing): moving to the smallest timeframes leads to imminent failure.

Many beginners look for a solution to their problems in micro analysis, trying to get into tick levels and engaging in pipsing. As a result their strategies fail due to deviations of 1-2 pips. This happens all the time. No one can resist trying to analyse the noise :-)

And those who have gone through several rounds of losses understand that it's insanity to analyze on ticks and minutes. And they go in the opposite direction - to analyze at higher and higher periods, where the pip noise is almost irrelevant.


We are well aware of everyone's desire to step on the rake of the pipsqueak microanalysis. They really should be stepped on.
Probably no one doubts that you and your team are experienced and competent programmers.
But you sometimes make statements that can only be uttered by an experienced trader... An example is in the above quote.
OK, if your beliefs were only your own business, but they indirectly affect your work. Example (lest you get banned for being unsubstantiated :-((( ) ), everyone has had the opportunity to test their Expert Advisors on history since January 2006. My Expert Advisor prepared for trading on one-minute charts (what a crazy thing!!!) was given a history for testing only since August 15, 2006. Although my real account at Alpari has a history on the 1-minute chart since December 14, 2005. What prevented you from making a normal history? Probably your neglect of trading on the 1 minute chart. But you don't know the basis of that trading, but you have the conviction.
Then you keep repeating about some kind of noise. Please show me what it is.
 
To clarify the topic of the topic: Inside Minutes Simulation vs Potik's History

Unfortunately, you made the wrong conclusion "What prevented you from making a normal history? Probably your neglect of trading on the minute chart." due to the fact that you simply didn't do the technical calculations. And if you did, everything would immediately fall into place. We've been describing all the technicalities of the lack of a regular deep detailed history many times over in the metaquotes.ru forum for years.

It's strange that you have overlooked our MetaTrader History Center project, in which we will provide a clean and normalized M1 history for many instruments over 5-7-10 years. The beta version is scheduled to launch in October 2006.

By the way, I've never been against minutiae. I'm against "burrowing into the tick stream within a minute thinking - this is accurate testing" and unsubstantiated statements "MetaTrader tester is bad, because it doesn't use tick data". On the contrary, I try to encourage traders to conduct their own research, so that they would get away from speculation and conduct practical testing themselves and see that the testing is very accurate. And be sure to honestly publish their research in the forums, rather than stopping with an "why should I do research - it's clear as it is!" grudge.

Those traders who have spent a fair amount of time testing have already done their tests and are satisfied with the quality of the testing and the tolerance for error. It is a pity that few publish their studies.
 
Renat писал (а):
To clarify the topic of the topic: Modelling Inside Minutes vs Potik's Story

<skip>.
Weird. Actually, the topic of the topic of the topic is somewhat different. Let's not wishful thinking. Mandor has only outlined the problem with the example of M1. As it turned out, including in my case, something is not clear to everyone with other TFs as well. That is why we are trying to find answers and possible solutions. This is quite normal. I realized that everyone is afraid that MetaTrader History Center in M1 will become some kind of substitute for reality again. It's like starting to study nuclear physics at the level of molecules. Or am I missing something again? Well, then I'll be silent altogether and only read clever thoughts. I can only say that from my experience that having summary data you can never get source data, but on the contrary you can. Miracles do not happen.

Regarding research. I have already written that I try not to get into minutes or tics unless absolutely necessary. I just needed to understand the reason of differences in results of my program's operation in the Strategy Tester and on the demo account. I have already made certain conclusions. I.e. I can say the goal is achieved. However, I cannot say the same about the result. Generally speaking, the Championship will show everything, including the consistency of the concept of MT4 autonomous testing. If no one wins, the developers will have to think about it. Frankly speaking, I cannot believe that a program tested only in real time will win. It's very complicated and time-consuming. So time will judge. We don't have long to wait.
 
The topic of the thread is exactly that - re-read the essence of the posts, please. All your other questions are secondary.

You clearly point to the topic (give me ticks, don't make up modelling!), get an answer to it, but abruptly forget about the question you asked and pretend that the question is about something else. But even in your last post you declare "MetaTrader History Center with M1 will become again some kind of reality substitute" - it's no good. So, you have a complete misunderstanding.

Actually, the statement "MetaTrader History Center with M1 is going to be another kind of reality substitute" puts a full stop. It should be framed.
 
Renat писал (а):
The topic of the thread is exactly that - re-read the essence of the posts, please. All your other questions are secondary.

You clearly point to the topic (give me ticks, don't make up modelling!), get an answer to it, but abruptly forget about the question you asked and pretend that the question is about something else. But even in your last post you declare "MetaTrader History Center with M1 will become again some kind of reality substitute" - it's no good. So, you have a complete misunderstanding.

Actually, the statement "MetaTrader History Center with M1 is going to be another kind of reality substitute" puts a full stop. It should be framed.

I'm done talking :)
If no one else shows up, it means that everything is OK. We are the only ones who are like that with Andriukha...
 
The instigator of the whole mess was banned, but six pages were added. At the very least, it's clear again that we need to write more in-depth articles on the subject.

In the meantime, we invite everyone to the Championship - it has already opened.
 
>> Those traders who have spent a fair amount of time testing have already carried out their tests and are satisfied with the quality of the testing and the tolerance for error. It's a pity that not many of them publish their research.

I want to make a comment on MY research.

I have a written Expert Advisor. It is on M1. It does not seem to be in scalping class (expected payoff is 3-4 and higher, it's stable). Does not steal from broker, making ~2 trades a day. He does not steal needles, spikes, gaps.

BUT

We test demo and real (2 accounts). Condition of position opening - Bid - MA_main > THAT. MA_main is a mouving on Open, so it does not dance inside a bar.
When the Expert Advisor opens a position, it writes in the log - Bid-MA_main (it should be about 11 points).

The real broker, being an honest broker, opens the position exactly at that moment when Bid-MA_main = 11 points (no requotes, slippage in OrderSend = 0).
But the tester, on the other hand, skips a quote when Bid-MA_main = 11 and gives the next one when Bid-MA_main = 12 (it opens at this price). It ends up being one point more optimistic.

And here's a concrete example - me. I sit and wonder - is it such a slight 1 pip noise that sometimes plays in my direction (when the tester gives a quote and the dealer misses it), or a drawback in the modelling algorithm? If I knew that I had a tick story in front of me, I would feel more relaxed.
 
Am I being ignored?
 
kniff писал (а):
Am I being ignored?

i don't know what to say.
the EA does not fall under pips and claims accuracy of ONE pip.
 
Yes, but, sorry, I find the difference of 10% between real and tester to be INCRITICAL. This means that the minimum expectation should be 10!!!! It's one thing to say, don't writereal with 0.25 expected payoff, and quite another to say, write only with 10!

All this is even multiplied by the fact that slipaddings may occur by 2!...points (and in both directions). So, instead of understanding the real tick history, we have been collecting statistics for a month already and instead of doing our job :)))))
12345678
New comment