Who wants a strategy? Lots and for free) - page 20

 
Figar0 >> :

I was fascinated by the topic of replacing the trader's brain with a computer, and suddenly I remembered a long-awaited piece of software designed to replace the trader's grey matter. I was pleasantly surprised that the project is not dead and has made some steps forward. I was pleasantly surprised with development of Forex Strategy Builder http://forexsb.com/index.html, it is completely free, it is drawn by Bulgarian programmer, there are a lot of bugs, Russification is horrible, but it is harmless) program funny thing is in automatic TS generator. The program automatically goes through combinations of different indicators on the selected TF in search of the most productive combination. Primitive, but with taste. The result is given as a set of indicators and entry/exit conditions, and then it is easy to implement it in MQL or whatever.

Sometimes it's nice to substitute your own brain with computer ones. Of course, the adequacy of the generator must be tested, for example, by implementing the strategy in MQL and comparing the results. I could try it myself, but I have too much work to do now...

Help to get it up and running doesn't work.

 

I can't start ssb_1, it says "is not a Win32 application"... although it's the latest Java, Net Framework 2, the file is fully downloaded...

What am I doing wrong?

 
mpeugep >> :

I can't start ssb_1, it says "is not a Win32 application"... although it's the latest Java, Net Framework 2, the file is fully downloaded...

>> what am I doing wrong?

There was a glitch on site, instead of 610kb, site allowed me to download only a little over 200kb. As a result, it was an incomplete download. The glitch has been fixed, now the file is fully downloaded and runs.

 
Reshetov >> :

There was a glitch on the site, instead of 610 kilobytes, the site only allowed to download just over 200 kilobytes. This resulted in incomplete download. Now it has been resolved and file is downloaded and run completely.

>>Thank you, everything works!

 
One thing I can say... The Bulgarian programmer's programme works great. I have already implemented 4 strategies. You have to select the best, as the best will show the best results. In addition, the optimizer is always at your fingertips. The strategy sometimes needs a filter.
 
zfs писал(а) >>
>> I can say one thing... The program of the Bulgarian programmer works great. I already implemented 4 strategies. You have to select the best, as the best will show the best results. Besides, you always have the optimizer at your fingertips. The strategy sometimes needs a filter.

I like it, too, in principle! But only... in pessimistic fashion. :)))

And will you show the implemented strategies? At least the xml?

And preferably with reports from MT4... ;)

 
Reshetov писал(а) >>

... Someone will download broken and broken quotes and run tests on them and remove all profitable strategies. ...

Or maybe not to allow "users" to download quotes? Use the ones that are in the repository. And update them accordingly (to the administrator).
 
voltair >> :
How about not allowing "users" to download quotes? Use the ones in the repository. And update them accordingly (to the administrator).

It would be ideal if all brokerage companies would provide quotes from one place. And until they have their own quotes provider and their own settings of filters, these quotes, then we are just dreaming about the bright future.


And so, in principle, strategies will be sorted out based on a variety of different quotes from different servers and only the most "thick-skinned" will be naturally selected.

 
Reshetov писал(а) >>

It would be ideal if all brokerage companies provided quotes from a single centre. But as long as each of them has its own quotation provider and its own filter settings for these very quotations, one can only dream of a bright future.

As a wise man once said - "you do not have to eat the whole pan to taste borscht". :)

I.e., running quotes of unlimited number of brokerage companies is certainly good. But I suppose the computational resource of the repository is not unlimited, and not all brokerage companies are "interesting" and necessary. So, imho, we should choose only a limited number of brokerage companies (for example 10), the quotes loading for which could be automated (with correctness check). And the strategies should be based on them. And to select the brokerage companies in such a way, that quotes' characteristics significantly differ. Then if the strategy is really "thick-skinned", then it will show profit everywhere on this "continuum" of brokerage companies. And if a user wants to check "his" brokerage company - let him do it himself. And also we can include some brokerage companies into the quotes "continuum" and exclude some others.

Although, of course, there are brokerage companies where some strategies will work well, but will not work anywhere else. The question is whether such strategies should be kept in a depository?

 

At the request of those who suffer, I am printing the results of not very good strategies on our favourite analyser:

I am printing the results for 3 months (I am too lazy to wait, on longer periods the profit is smeared, but the analogy is the same, I still need to optimize more often :)

I have a small deposit, I don't have any money now, I've invested all my money in shares :)

All strategies so far for EURUSD

Based on the deviation from the typical price using the ATR MA Oscillator.

Bars in history1367Modelled ticks1093881Modeling quality69.23%
Chart mismatch errors67




Initial deposit1333.00



Net profit2631.09Total profit5161.32Total loss-2530.23
Profitability2.04Expected payoff51.59

Absolute drawdown195.06Maximum drawdown637.41 (20.29%)Relative drawdown21.45% (310.77)

Total trades51Short positions (% win)32 (53.13%)Long positions (% win)19 (52.63%)

Profitable trades (% of all)27 (52.94%)Loss trades (% of all)24 (47.06%)
Largestprofitable trade1337.98losing deal-109.29
Averageprofitable deal191.16losing trade-105.43
Maximum numbercontinuous wins (profit)4 (1819.64)Continuous losses (loss)4 (-417.69)
Maximumcontinuous profits (number of wins)1819.64 (4)Continuous loss (number of losses)-417.69 (4)
Averagecontinuous winnings2Continuous loss2

2.Bounce from round levels (will be finalising).

Bars in history1367Modelled ticks1093881Modeling quality69.23%
Chart mismatch errors67




Initial deposit1333.00



Net profit1225.45Total profit3613.25Total loss-2387.80
Profitability1.51Expected payoff47.13

Absolute drawdown148.29Maximum drawdown1030.80 (29.19%)Relative drawdown29.19% (1030.80)

Total trades26Short positions (% win)15 (53.33%)Long positions (% win)11 (45.45%)

Profitable trades (% of all)13 (50.00%)Loss trades (% of all)13 (50.00%)
Largestprofitable trade1149.96losing deal-193.58
Averageprofitable deal277.94losing deal-183.68
Maximum numbercontinuous wins (profit)5 (1805.59)Continuous losses (loss)3 (-556.54)
MaximumContinuous Profit (number of wins)1805.59 (5)Continuous loss (number of losses)-556.54 (3)
Averagecontinuous winnings2Continuous loss

2


3

Reason: