Not the Grail, just a regular one - Bablokos!!! - page 285
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
Yes.
Joker, hi. I'll ask a question about the famous example of Necolla. Since we started to discuss (maybe some new thoughts will appear).
If we are in the last line when increasing the lot to 5, we get -5000? It is 50/50 in the end. And our MO is zero again.
Necolle's model with lots increase (martini) is understandable, I've read and experimented a lot in excel,
But to win ALWAYS on a random series...
/***********************************************************************/
Look Shirsha - MM apply: Mlyn...Profit is some...
1 -48054,6 49942,6
1 -24024,5 23985,3
0,1 -1301,33 1099,09
0,1 -1000,63 299,58
0,2 -1202,1 799,72
0,4 -801,4 1599,16
0,8 -800 797,76
1,6 -1600 0
5 0 5000 5000 (!!!!!!! -that's random)
4738,65 -78784,56 83523,21 4738,65_
=====
/***********************************************************************/
RIGHT ... TO THE RIGHT VIEW.... I and want to roughly find synthetic through pattern comparison with synthetic, but I can't find a tool to compare. And about regression, isn't it linear? We are far from linear ... although if we split the series into parts we get several linear ones....
linearity of regression should not be confused with linearity of the independent function
As a rule, from the practical point of view, only a linear regression is needed
We cannot trade with square or logarithmic lots anyway
And the non-linearity of the independent variable can be arbitrary (under the problem)
searching for the synthetic by pattern is relatively easy
write the pattern in the array: for(j=0; j<points; j++) MODEL[j]=/* here we write the pattern or function */;
shift the function to zero: double zero_shift=-MODEL[0]; if(zero_shift!=0) for(j=0; j<points; j++) MODEL[j]+=zero_shift;
if regression has a free term (aka intercept), this action need not be performed
feed the variables (pre-calculated equities) into the matrix: for(i=0; i<variables; i++) for(j=0; j<points; j++) MATRIX[j].Set(i,EQUITY[j,i]);
put the model into matrix: for(j=0; j<points; j++) MATRIX[j].Set(variables,MODEL[j]);
count regression: CAlglib::LRBuildZ(MATRIX,points,variables,info,LM,AR);
extract roots:::LRUnpack(LM,ROOTS,variables);
remember to scale and round lots
-- came in by accident and ran away fast --
linearity of regression should not be confused with linearity of the independent function
As a rule, from the practical point of view, only a linear regression is needed
We cannot trade with square or logarithmic lots anyway
And the non-linearity of the independent variable can be arbitrary (under the problem)
searching for the synthetic by pattern is relatively easy
write the pattern in the array: for(j=0; j<points; j++) MODEL[j]=/* here we write the pattern or function */;
shift the function to zero: double zero_shift=-MODEL[0]; if(zero_shift!=0) for(j=0; j<points; j++) MODEL[j]+=zero_shift;
if regression has a free term (aka intercept), this action need not be performed
feed the variables (pre-calculated equities) into the matrix: for(i=0; i<variables; i++) for(j=0; j<points; j++) MATRIX[j].Set(i,EQUITY[j,i]);
put the model into matrix: for(j=0; j<points; j++) MATRIX[j].Set(variables,MODEL[j]);
count regression: CAlglib::LRBuildZ(MATRIX,points,variables,info,LM,AR);
extract roots:::LRUnpack(LM,ROOTS,variables);
remember to scale and round lots
-- came in by accident and quickly escaped --
put the pattern into an array: for(j=0; j<points; j++) MODEL[j]=//* here you write the pattern or function */;
Here is an example, if you don't mind.
How to write a template correctly and how to write a function correctly ?
Joker, hi. I have a question about the famous example of Necolla. Since we started to discuss (maybe some new thoughts will appear).
If we are in the last line when we increase the lot to 5, we get -5000? It is 50/50 in the end. And our MO is zero again.
Necolla's model with lot build-up (martini) is clear, read and experimented a lot in excel,
but so that on a random series ALWAYS win...
/***********************************************************************/
see Shirsha - MM apply: Ml...Profit is some...
1 -48054.6 49942.6
1 -24024.5 23985.3
0.1 -1301.33 1099.09
0.1 -1000.63 299.58
0.2 -1202.1 799.72
0.4 -801.4 1599.16
0.8 -800 797.76
1.6 -1600 0
5 0 5000 5000 (!!!!!!! is that random)
4738.65 -78784.56 83523.21 4738.65_
=====
/***********************************************************************/
Greetings.
I think his strategy has the right to life and here's why. The betting system that he applies is actually the length of this betting combination may reflect the size of the market cycle, where in this case the bet size is the statistical probability that a decision is true at one point or another. In fact the size of his bet is the coefficient of importance or truthfulness ( truthfulness ) of decision making. The current performance of a betting system is the size of the profit which this betting system has made at this moment. My assumption is that he chooses 4 maximum betting systems (for 4 instruments) from a set of 7 majors, based on two-week statistics or something like this....
His strategy is as follows for example:
1. he opens in the direction of the previous candle's close.
For a single instrument at each moment the optimal betting system for example for a candlestick depth may be
EURUSD 0.11 0.42 0.11 0.31 ( max. profit e.g. 5000 USD )
USDCHF 0.25 0.66 ( max. profit e.g. 3000 CU )
...
USDJPY .... ( max profit e.g. 25 CU )
This table and the betting system are constantly changing over time. Every betting system is the most effective for every single instrument at the given moment.
Probably he chooses 4 with maximal indexes thus increasing his probability of the successful total trades completion.
The direction to enter the trade - for example to enter in the direction of the previous candle's close or something like that.
To some extent this is a special case and a kind of regression model, only through game theory.
GerbertX:
Aleksander, что действительно можно зарабатывать на случайных числах, или ты троллишь?
paukas:
I can do it.Taki, what do I hear! Rebbe! Rebbe, bring us a bottle of your best. Don't be stingy, I said the best! Such a day! Our Moisha's all grown up!
Has anyone managed to replicate Joker's system?