Yoghurt systems and canned systems or The relationship between trading tactics and the reliability of historical test results - page 9
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
Well, three years for days is fine, but a month or so for hourly bars would be "not enough". And the whole thing is 750 bars.
I take more than three years... Three years is a better fit...
I've been taking over three years...
For what TF?
For which TF?
M1 to H1... I don't go any higher...
M1 to H1... I don't go any higher...
I wonder what is the result? How long does it work in real time without over-optimisation?
"The search for patterns" is an eternal and inexhaustible topic, for which no answer will ever be found (in the perpetuum mobile sense). And testing and evaluation of TC is quite a practical problem that can be solved for a given TC, even if the allegedly found patterns are not logically understood or completely unknown. It seemed to me that the topicstarter's question was asked precisely about how to adequately assess the behaviour of the TS in the future...
The search for patterns is meaningless if we cannot justify with any degree of reliability that the TS built on the found pattern will behave acceptably in the future.
I believe that without knowing the nature of the pattern it is premature to talk about the degree of reliability. I think the concept is simple - we look for a pattern, study it and justify the degree of reliability. It is from the nature and the very properties of the pattern that we can judge how the pattern will behave in the future, and I see no great problem in justifying the degree of reliability from the nature of the pattern. I'm sure the topicstarter's question would remain unanswered if the pattern were taken out of brackets.
I may be getting the wrong impression, but I see a tendency to justify the degree of reliability of the TS without having any idea of the pattern behind the TS. That, to me, is exactly the pointless thing to do. Lack of a pattern or understanding of what kind of pattern is being exploited indicates trivial fitting. What is the point of justifying the reliability of the fit?
That's how I see it, Mathemat. If you don't have a pattern in your hand, then you've been fitting, so you can't justify it, but if you have a pattern in your hand, the justification is straightforward.
Then I want to ask - is trading more art or mathematics? If it is mathematics, then maybe it cannot be justified mathematically, but if it is art, then maybe one can "feel" the market and "thirdly" understand what will work in the future?
Tough question, LeoV. Probably both. But I myself would like to change the ratio of these components towards mathematics. Of course, there will never be complete mathematical reasoning, and so every system will be a disaster.
I wonder what the result is? How much works in real life without over-optimisation?
:-) The result is very modest :-) I'm even embarrassed.
Last time I optimized during May holidays (2nd May) on interval from 01.01.2001 till 31.12.2007. I checked on four months of 2008. Several different sets of parameters worked on 4 real accounts.
:-) The result is very modest :-) I'm even embarrassed...
Last time I optimised in May holidays (2nd May) on the interval from 01.01.2001 to 31.12.2007. I checked on four months of 2008. Several different sets of parameters worked on 4 real accounts.
I'm not asking for profit percentage))). I'm asking how long may this TS, optimized for 7 years with OOS in 4 months, work in your experience?
I'm asking how long in your experience can such a TC, optimised for 7 years with an OOS of 4 months, work?
I don't know... as long as it works. I have a criterion by which I clearly identify that the system has stopped working and I will stop all EAs on all accounts.
Well how can I tell you - a trend can essentially be taken as a pattern, and so can pullbacks on the tern - but how can you determine the duration of the trend, and whether the pullback is the start of a new opposite trend?
A trend is not a pattern. A trend is a condition in which different patterns can manifest themselves. For example, we got ahead of an inflection point and the market moved away from this extremum by a certain value. If we open a position in this situation, we can detect a pattern, for example, that in a certain percentage of cases the market continues to move by more than N points. Or come up with other conditions and run them, maybe they will. But the trend itself is not what I'm talking about.