Goosey

[SWG123]

None of which makes the ea profitable. It has lost money on my demo slowly but steadily.

I agree, left to run with the default settings it seems it will drain your account progressively.

But one thing I've noticed, if you drop down to M5 and only allow entries during the first hour after LO, it starts to look a bit more interesting...

[patmontes]

None of which makes the ea profitable. It has lost money on my demo slowly but steadily.

I agree, left to run with the default settings it seems it will drain your account progressively.

But one thing I've noticed, if you drop down to M5 and only allow entries during the first hour after LO, it starts to look a bit more interesting...

Hmmm might be a good test for next week!

[AnotherBrian]

Steve - this function is called in the EA but the indicator "Wilder's DMI book alerts nmc 2" is not included in the first post.

Edit: I found it here http://www.stevehopwoodforex.com/phpBB3 ... =16&t=4432
Inside the rar file.

Code: Select all

double GetBookDmi(string symbol, int tf, int length, double level, double plusminus, int buffer, int shift)
{

   double val = 0;
   
   val = iCustom(symbol, tf, "Wilder's DMI book alerts nmc 2", length, true, false,
                 level, plusminus, " ", 
                 false, false, false, false, false, 
                 false, false, false, false, false, 
                 " ", buffer, shift);
   
   return(val);

}//double GetBookDmi(string symbol, int tf, int length, double level, double plusminus, int buffer, int shift)

[SteveHopwood]

Steve - this function is called in the EA but the indicator "Wilder's DMI book alerts nmc 2" is not included in the first post.

Edit: I found it here http://www.stevehopwoodforex.com/phpBB3 ... =16&t=4432
Inside the rar file.

Code: Select all

double GetBookDmi(string symbol, int tf, int length, double level, double plusminus, int buffer, int shift)
{

   double val = 0;
   
   val = iCustom(symbol, tf, "Wilder's DMI book alerts nmc 2", length, true, false,
                 level, plusminus, " ", 
                 false, false, false, false, false, 
                 false, false, false, false, false, 
                 " ", buffer, shift);
   
   return(val);

}//double GetBookDmi(string symbol, int tf, int length, double level, double plusminus, int buffer, int shift)

Copied from post 1 here:

Go to celte83's thread at http://www.stevehopwoodforex.com/phpBB3 ... =16&t=4432 to read the trading method, download the indi's and template etc.

Did I ever mention that I am not best pleased to find that people cannot be bothered to read post 1 in my threads? And that I tend to put quite a bit of effort into them? And that I loathe writing them? And that I would infinitely prefer to tell people to sort everything out for themselves?

I am moderately sure I mentioned this somewhere before. I could be wrong of course. The memory does play tricks.

[AnotherBrian]

OK, I confess - I didn't read the first post assuming that all the required gear would be attached in the rar as a tidy package. Should never ASSume!

Here's a new discovery for me. I was playing with the code and discovered this. When the data window shows nothing for a buffer value you would think that it is returning EMPTY_VALUE but I found that the Wilder DMI indi is returning NULL. This screws up the code logic decision if we use EMPTY_VALUE (I don't know if the CloseEnough function would deal with this or not).
I did the same test for HGI and it does return EMPTY_VALUE.
So I now use EMPTY_VALUE || NULL.

I stuck this in under the call to to indi - val[1] being the result of the call to whatever buffer
If(val[1]=EMPTY_VALUE)Print("its empty");
If(val[1]=NULL)Print("its NULL");

Empty4 creates all kinds of challenges.

[SteveHopwood]

OK, I confess - I didn't read the first post assuming that all the required gear would be attached in the rar as a tidy package. Should never ASSume!

Here's a new discovery for me. I was playing with the code and discovered this. When the data window shows nothing for a buffer value you would think that it is returning EMPTY_VALUE but I found that the Wilder DMI indi is returning NULL. This screws up the code logic decision if we use EMPTY_VALUE (I don't know if the CloseEnough function would deal with this or not).
I did the same test for HGI and it does return EMPTY_VALUE.
So I now use EMPTY_VALUE || NULL.

I stuck this in under the call to to indi - val[1] being the result of the call to whatever buffer
If(val[1]=EMPTY_VALUE)Print("its empty");
If(val[1]=NULL)Print("its NULL");

Empty4 creates all kinds of challenges.

I have been coding these things for quite a long time Brian. The first thing I do is check whether an indi returns 0 or EMPTY_VALUE by calling the relevant buffers via a script. Wilder returns 0. You will find no instance of EMPTY_VALUE in the indi code.

So Goosey looks for non-zero values, not EMPTY_VALUE. An integer or double set to NULL returns 0.

[SteveHopwood]

I turned CSS on yesterday, on the H4 time frame. So far today up +257 pips - the first time the bot has made pips instead of losing them steadily.

Early days. It might be worth adding some of Bob's filters into the mix; we shall see.

[patmontes]

I turned CSS on yesterday, on the H4 time frame. So far today up +257 pips - the first time the bot has made pips instead of losing them steadily.

Early days. It might be worth adding some of Bob's filters into the mix; we shall see.

Thanks Steve! I will try out the same

[trader689]

Hey Steve

Sounds promising. do you need to put CSS timeframe as 240 or leave at 0?

thanks

EDIT: i think i read that wrong; meaning you leave the trading timeframe at m15 but turn on CSS at 240. I thought you meant switching the charts to H4

I turned CSS on yesterday, on the H4 time frame. So far today up +257 pips - the first time the bot has made pips instead of losing them steadily.

Early days. It might be worth adding some of Bob's filters into the mix; we shall see.

[SteveHopwood]

Encouraging results from turning on CSS last week.

Up +467 pips but down -$1.22 in cash. The week was about $10 in profit until some late trades on Friday closed out at losses that wiped out all the gains.

I used mptm to close half the trade at +15, so quite a lot of those +467 pips will actually have been counted twice when a trade carried on to hit tp. sl 20; to = 40. There were 10 trades that reached full tp, so the profits from the full lot size reaching tp would have left some excellent profits.

The downside of closing half the trade at +15 is that this creates an effective take profit of 20 pips compared to a trade going to full stop loss, so I am going to try locking in 2 pips at BE and see if the trades running to tp overcome the losses.

TradeReport is brilliant for highlighting the best and worst pairs. Best by far were UC, GChf and EG at 137, 107 and 99 pips respectively. Biggest losers were AChf and ACad at -25 and -33 respectively.

28 pairs traded with 50/50 winners and losers.