Statistics, optimisation and "lucky coin" .... - page 3
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
....
It always amazes me when you say that. I can find you anything in history, in any quantity.
Where did I say "on history"? Humbly, humbly, geniuses.
...... what does this have to do with trading systems that use patterns, i.e. the exact opposite of randomness?
Topikstarter. If a TS with a stable positive expected payoff is found, the purpose of optimization is to find the optimal set of parameters (and not to find the only profitable set for some abstract coin on this piece of history). The purpose of a set of statistics is to find the system itself, which is worth optimizing. In the Trade Idea - Statistics Set (idea verification) - Finding Optimal Parameters cycle, you release the most essential part, the first. The second and third make sense only if it is available.
Question #1: Please decipher the term "trading systems that use patterns" .....
Question #2: A trading idea - when the price reaches an extremum of N bars - to enter through (.... or on the rebound). We checked it in the tester and got a conclusion that somewhere it is better to enter through, somewhere it is better to bounce. We start to optimize: change the number of bars for identification of extrema, add a couple of indicators to improve accuracy, etc. Essentially, we obtain some system with a set of parameters. If the set of these parameters is large, the probability of getting good statistics in the test (for XXXX bars, for AAA symbols, on MMM timeframes) will be very high. In other words: we can always obtain positive results in the tester on history for any trading idea when optimizing it. The problem is: will this trading idea maintain its positive results in the future?
NB: I am not discussing the essence of what you call a "Trading Idea" ..... Let it be the simple X pips entry I suggested, something from Turtles, channels, Forks..., whatever...
Regarding question 1: What's not clear?
On question 2: It won't, of course.
I see, you've given up hope of seeing a pattern.
I can show you if you like. Effective now and since it has been identified.
Can you show me?
Here is the moment of recognition of the current level of 1.3028. 5 April, in real time.
By the way, the way the pattern is detected is also shown :)
The amount of nonsense in the new branches is steadily increasing. Has spring really arrived...?
Azarus, you seem to be one of those people who heard something somewhere, but did not understand what it means and what it can be applied to. Your first thesis begins with the words "There is a certain TS: "TP - X pips, SL - Y pips; ... The direction of the entry - determined by flipping a coin... "You can not read further, because what you propose is not a TS, but a random number generator. It's really senseless to optimize it - it's a fact. But to conclude on this basis that it also makes no sense to optimize any TS is either deliberate demagogy or a logical error of finding a connection where it does not exist, i.e. between TS and PRNG.
For a start, I don't quite understand the reason why you're so excited....
From what I wrote in the first post, you didn't understand anything (perhaps I wrote too much, well, sorry...) . I won't retell it, but I will explain for you personally, that my idea is that the price (as a certain set of numbers, presented in the form of certain levels and as indications) used in the optimization process is no better than the same set of numbers, obtained from the LFO. Since any TS built on random matching on history (RNG coin) does not give "confidence" in its performance in the future, the TS built on some detected relation of prices to history does not give such confidence either....
If you have the desire/capacity to identify demagogy/logical error in the above statement, it would be extremely interesting....
I see, you've given up hope of seeing a pattern.
I can show you if you like. Effective now and since it has been identified.
I am interested in your understanding of the term "Regularity".
And in words......:)
Question #1: Define the term "trading systems that exploit patterns" .....
This is where one can give a mathematically very precise definition.
A regularity as applied to the price series X(t), is any function
F(t, X(t), X(t-1), ..... where t is time and Z is the vector of "external" in relation to X-values (i.e. such values for which the mutual information I(Z,X(t))=0 for any t),
for which at any time t it is possible to specify one or more moments Ti>t at which the expectation
E{F(Ti, X(Ti), X(Ti-1), ..... , Z)} > 0.
It does not seem to require any special education to understand this definition. In simple words, a pattern is any dependence of transaction results on the moment and conditions of entry, which allows you to earn not only yesterday, but also tomorrow. This is in such a narrow sense, someone may outline wider, calling any dependence in a number of quotations as a pattern at all.
Question #2: Trading idea - when the price reaches an extremum of N bars - buy entry (.... or rebound). We checked it in the tester and got that somewhere it is better to enter on breakout and somewhere it is better to enter on the rebound. We start to optimize: change the number of bars for identification of extrema, add a couple of indicators to improve accuracy, etc. Essentially, we obtain some system with a set of parameters. If the set of these parameters is large, the probability of getting good statistics in the test (for XXXX bars, for AAA symbols, on MMM timeframes) will be very high. In other words: we can always obtain positive results in the tester on history for any trading idea when optimizing it. The problem is: will this trading idea maintain its positive results in the future?
This one won't save, the other one exploiting profiteering patterns (see answer to question #1) will save.