Signs of a REAL system - page 20
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
If it's all on a 'chat' level, then yes :)
If it's at the level of advice, then it's definitely not good for you.
... If at the level of recommendations, then it is unequivocally harmful.
What do you think is useful at the level of recommendations?
faa1947, you believe that there are no(theory) constant patterns in the market. Therefore any strategy is a fit for a particular period.
The probability (not mathematical, but subjective) of a strategy succeeding in the future is considered.
There are two methods of strategy creation:
- A haphazard one. For example, by selection of indicators and their parameters a profitable Expert Advisor is made.
- systemic. A system is written based on the identified patterns of a certain time interval.
I.e., it is possible to come to the strategy from two sides. The non-systemic way (without evidence) seems to be the wrong approach. Mr. Reshetov has written an excellent constructor for creating haphazard strategies. The results of these strategies after creation are more inclined to believe that the haphazard approach is wrong. The systemic one is not perfect either, but it is not worse.
If we follow only theory, it is inevitable (because of the spread), but theory allows certain regularities to work during any finite time interval. And if that interval is measured in years, then why not...
What do you think is useful on a recommendation level?
They don't have any recommendations. Except to shoot the participants in the discussion. As pests.
faa1947, you believe that there are no(theory) constant patterns in the market.
I don't think so. I will reiterate my understanding.
There are patterns in the market (the crossing of two averages, for example). There are a great number of these patterns. TS should be able to recognize this pattern. The NS recognize some patterns that are impossible to draw or describe. A tester gives statistics about the occurrence of such a pattern in the past, but gives no guarantee of occurrence in the future.
There are several basic questions for a pattern:
- how common it is (e.g. for longs, for shorts, for longs and shorts) ;
- how to recognize the occurrence of a pattern in time;
- how to timely detect when this pattern is dying, which is a must, as we can see by losing trades during testing;
- whether this pattern can be adapted to a different section of BP.
The question of adaptation is the most difficult one. In particular, as an adaptation we can use reoptimization of the TS in a new section of BP - not to be confused with a forward test.
All this is obvious if we do not forget for a moment that the RT we are dealing with is a non-linear, dynamic, non-stationary system with memory. If we constantly relate our decisions to this understanding, and trust in the test results to this understanding as well, many mistakes will be avoided.
P/S. Neuron specifically defined over-optimisation, taking it from the NS. As for testing requirements, again, this has been discussed many times on this forum and in articles. So one should read before opening a thread - reading in general is a very useful thing to do.
They have no guidelines. Except to shoot the participants in the discussion. Like pests.
And it all started so well :))))))))
What do you think is useful on a recommendation level?
It is impossible to answer one way.
Look, for example, as a reference point, at the book "Guidelines for the design of automatic control systems".
The problem statement, the search for solution methods, the style of presentation.
They have no guidelines. Except to shoot the participants in the discussion. As pests.
That's supposed to be very funny?
Well, guys, I'm just crazy about all of you.
....
Regarding testing criteria, the author of this thread would do well to familiarise himself with the forum in which he opens the thread. This issue has been discussed many times, although I, in my opinion, am the first who does not recognise fitting as such and all talk about it I consider harmful.
why am i suddenly not allowed to raise an issue that concerns me at the moment and make my suggestions as to what i think about it? :)
read my very first sentence carefully:
I have tried to systematize my criteria for assessing the correctness of a trading system and an Expert Advisor built using it. Here is a brief summary of my rules that came to my mind.
This is my opinion, joined by quite a few participants who think the same way as I do. we have all benefited from this discussion.
Now you are free to create your own discussion about what counts as fitting in and what doesn't, just as I do, based on another thread (in this case this one). I think those who are affected by it will participate (I will definitely participate).
A one-sided answer is impossible.
But it is possible to reject all suggestions at once, isn't it?
Look, for example, as a reference point, at the book "Guidelines for the design of automatic control systems".
Problem setting, search for solution methods, presentation style.
Do you think people present on this forum have not read books?
Here people stated their own experience obtained with blood and sweat which cannot be replaced with books.
Or, with a couple of bullet points, mark the most useful conclusions from this book in your opinion.