I'm learning to write an advisor...

New comment
 

What do you think of the test results?

I know you can "draw" anything in the tester, but I'm not a programmer and I didn't expect that I could do that))

I tested EURUSD, GBPUSD, GBPJPY from August 2007 till November 2008 (1 year)

Entry on a level breakout, stop static from 5 to 9 depending on the pair, static profit on all pairs 100.

-First three pairs figs 0.1 lot

-Second three figures with increasing lot depending on the balance: (thanks to KimIV, I found his way somewhere...)

      if ( ab>=20 && ab<40)     Lot=0.01;
      if ( ab>=40 && ab<60)     Lot=0.02;
      if ( ab>=60 && ab<80)     Lot=0.03;
      if ( ab>=80 && ab<100)    Lot=0.04;
      if ( ab>=100 && ab<120)   Lot=0.05;
      if ( ab>=120 && ab<140)   Lot=0.06;
      if ( ab>=140 && ab<160)   Lot=0.07;
      if ( ab>=160 && ab<180)   Lot=0.08;
      if ( ab>=180 && ab<200)   Lot=0.09;
      if ( ab>=200 && ab<400)   Lot=0.1;
      if ( ab>=400 && ab<600)   Lot=0.2;
      if ( ab>=600 && ab<800)   Lot=0.3;
      if ( ab>=800 && ab<1000)  Lot=0.4;
      if ( ab>=1000 && ab<1200) Lot=0.5;
      if ( ab>=1200 && ab<1400) Lot=0.6;
      if ( ab>=1400 && ab<1600) Lot=0.7;
      if ( ab>=1600 && ab<1800) Lot=0.8;
      if ( ab>=1800 && ab<2000) Lot=0.9;
      if ( ab>=2000 && ab<4000) Lot=1;
      if ( ab>=4000 && ab<6000) Lot=2;
      if ( ab>=6000 && ab<8000) Lot=3;
      if ( ab>=8000 && ab<10000) Lot=4;
      if ( ab>=10000 && ab<12000) Lot=5;
      if ( ab>=12000 && ab<14000) Lot=6;
      if ( ab>=14000 && ab<16000) Lot=7;
      if ( ab>=16000 && ab<18000) Lot=8;
      if ( ab>=18000 && ab<20000) Lot=9;
      if ( ab>=20000           )  Lot=10;



EUR 200$ 0.1 lot


GBP 200$ 0.1 lot


GBPJPY 200$ 0.1 lot


EUR 200$ 0.1 to 10 lots


GBPJPY 200$ 0.1 to 10 lots


GBPJPY 200$ 0.1 to 10 lots

 
Lot 10? Well, that's cool, I guess.
 

It's not about awesomeness, you can test any lots...

The first 3 rice 0.1 lots...

 

Take=5/9, stop=100

Do you realize that 95% of the time in real trading your position will be in the loss zone, waiting out the "misery". And with a real account you yourself 95% of the time, will constantly be in a state of gloomy depression, looking at the monitor.

Uncomfortable trading.

And of course, in the best case the profit on deals will not be +5/+9, but rather -1/-10.

Unfortunately, the dealers are not fools either...

 
ALex2008 писал(а) >>

I know you can "draw" anything in the tester, but I'm not a programmer and I didn't expect it myself...)

....

Entry on level breakout, stop static from 5 to 9 depending on the pair, profit static on all pairs 100.

-For non-programmers in the tester, the optimizer draws.

-SL 5-9 TP 100 - unrealistic parameters. How many parameters are optimized in total, what is the frame?

-Tighten your simulation quality. n/a and 30% indicate a problematic "history". It may not be crucial, but it may be important for a novice non-programmer.

-Fewer attention in the test should be paid to increasing the lot. Don't you care how many Strategy Tester quid an Expert Advisor earns in absolute value? But the test results are washed out.

-Your pictures in this case say very little. If you want something from us, except "What good you are", post the full results of testing with a constant lot.

 
rid писал(а) >>

Take=5/9, stop=100

Hey, the man's got it backwards)

 
rid >> :

.. And with a real account yourself, 95% of the time, you will be in a state of gloomy depression, looking at the monitor all the time.

Uncomfortable trading...

That's what advisers are for...)

 
rid >> :

Take=5/9, Stop=100

Read the 1st post carefully!

 
Figar0 >> :

-For non-programmers, the optimiser draws in the tester.

-SL 5-9 TP 100 - unrealistic parameters. How many parameters are optimised in total, what frame?

Optimised both take and stop, frame H4

 
Figar0 >> :

-Tighten up the quality of the modelling. n/a and 30% suggest a problematic 'story'.

To be honest, I don't even know what it is))

If you need anything from us other than "How good you are"

No one has said that yet)

..post the full test result in a permanent lot.


Please...


EUR


GBP


GBPJPY

 

In general, after analyzing trades made by the Expert Advisor with all its profitability, I see that there are many moments when stops are fulfilled by "a lot" and as a result the price does not go to profit zone. This is a minus! But since stops are small comparing to TP, the next TP after the series of stops overlaps them and even earns more...

The only thing I would like to try is to cut the number of stops and introduce their limitation in the code when the signal triggers. Example: we get a buy signal - trade, 5 or 10 stops, wait for the next signal... But I don't know how to implement it yet...

1234567
New comment
Reason: