Sensation! A profitable strategy for playing beagle has been found! - page 10

New comment
 

Counsellor. In the download.

It's kind of like a pure Eagle's Nest. It's flailing with pauses.

Dax, m1 from May 1 till now. Don't forget that the spread on futures is not taken into account by the tester. Only commission.

Lot initial=0.01. Subsequent doubling of each step.

//--------------------------------------------------------
BroCo-San-Francisco (Build 221)
Symbol FDAXM9 (DAX (eurex) (08:00 - 22:00) Exp 19/06/2009)
Period 1 Minute (M1) 2009.05.04 08:00 - 2009.06.09 21:59 (2009.05.01 - 2009.06.10)
Model All ticks (the most accurate method based on the smallest available timeframes)
Parameters pips=280; profitpips=240; Lots=0.01; time=0; starttime=7; stoptime=17;

History bars 23285 Modelled ticks 344646 Modeling quality 25.00%
Chart mismatch errors 0

Initial deposit 3000.00
Net profit 1132.27 Total profit 4569.06 Total loss -3436.79
Profitability 1.33 Expected payoff 4.66
Absolute drawdown 42.05 Maximum drawdown 336.06 (9.80%) Relative drawdown 9.80% (336.06)

Total trades 243

Short positions (% wins) 120 (64.17%)

Long positions (% of all wins) 123 (65.04%)
Profitable trades (% of all) 157 (64.61%)

Loss trades (% of all) 86 (35.39%)
The most profitable trade is 544.91 (loss trade -588.26)
Average profitable trade 29.10 losing trade -39.96
Maximum number of continuous wins (profit) 12 (133.92)

Continuous losses (loss) 4 (-833.55)




Files:
hlopmaster.ex4  13 kb
 

It seems reasonable to add a virtual trade algorithm to the COUNTER code and allow entry after the next significant virtual loss. I think this can significantly reduce the drawdown.

Analogue - Filter based on trading history

 
rid писал(а) >>

It seems reasonable to add a virtual trade algorithm to the COUNTER code and allow entry after the next significant virtual loss. I think this way you can significantly reduce the drawdown.

yes... we need to add an algorithm... I am in the process of finalizing it... I added a lot system to martin's EA... I'll post the results soon...
 

Кстати, а откуда такая уверенность, что используется какой то стандартный сишный генератор, а не какой нибудь tcl - оаский? В документации ничего не нашел, это какая то инсайдерская информация?

One person tested my algorithm on a classical C compiler. The result is the same. => Generator code was borrowed from there. In addition, the architecture and syntax of MQL is very similar to C, so it can be assumed that it was the C compiler that was used.
 
HideYourRichess >>: What is the general solution? The general solution of what?

Well, I went a little overboard with the "general" part. No big deal, the idea is the same as yours. It was just interesting to generate Bernoulli scheme for arbitrary p. I have to compare it with rounded 32767*p.

 
Mathemat >> :

Well, I went a little overboard with the "general" part. No big deal, the idea is the same as yours. It was just interesting to generate Bernoulli scheme for arbitrary p. We have to compare it with rounded 32767*p.

One could convert an integer to a real number and then compare it to p. But that's a bit tricky.

 
Good day! If you think about the fact that from Friday to Sunday these generators kind of rest, and neither on Saturday nor on Sunday life does not stop, the exchange offices work, imagine the mast of the beloved English company near in Spain, how many ATMs toil, and on Monday Gap gives out a generator. (he's been injected, and he's gapped-correcting error) and DEJAVU again... But I rather like how on the news these generators generate an illogical sequence, with a successive issuing of a series of only Tails for example...
1...345678910
New comment