Holy Graily Bob 'n Dotty 'n Super Slope 'n Flying Buddha

[TraderJoeForex]

Hi Everyone,

I'm going to close down my demo today. Running out of space , but just wanted to share how it has been behaving since 13/3, so almost 7 weeks. It has been pretty quiet and only this week, it ran into major drawdown.
This was using TJF's setfile from page 2 on H1

Hi Louise,

Thanks for running your test.. and that was the first incarnation of what is now my Grid + Sixths set.

... I think the addition of breakeven + jumping SL along with other modifications I gradually made have helped make the set file more robust although drawdown is still a big factor in considering whether or not to use a tight grid. I'll wait until tomorrow to post the collective results for my current tests but here's a snapshot of this one since it may be relevant for your comparison (if you filter just GBPJPY results out of yours for just April)..

Bear in mind I have 10 tests running on the same account so the trading size here is 0.01 per 10,000 rather than 1000 per pair. That means the drawdown for comparisson for just GBPJPY for April with this set file would have been 29% not 2.9% if that makes sense.

[TraderJoeForex]

Something I have been contemplating, is using Steve's dash board ea set up for the W1D1H4 pivots to recognize daily trend direction or we can even set it up for D1H4H1 pivots for immediate direction to open a dedicated chart template set up for HGBnDottynSSnFb.

Obviously I will only have the pairs that I want the Ea to trade on the Dash Ea. So we don't have to have so many charts open.

If the Ea opens a trade against the daily trend direction I can make an informed decision if I want to close my trades for that particular pair. The good thing is that the dashboard Ea wont open charts for pairs that are ranging.

It simply has to work in my opinion

Cheers
Leon

Nice idea and would be great if this can reduce drawdown on the grid sets somewhat. I will give this a go multipair with the winner from my April collective test and run another account without. Good that is May's supertest sorted!

[TraderJoeForex]

...week 4 for my test which brings me to the end of this first instalment:

Screenshot 2017-04-28 19.46.09.png

Below are the individual results for the top 3 green pips and what's interesting is that in terms of profit vs drawdown the default offsetting control set offered almost the same positive ratio as my modified sixths + grid all be it with much smaller numbers but this would be a huge plus on a small account. These are the only two sets that have managed to deliver more profit than drawdown.

It's been interesting watching this all on one pair. Next step for me is to look at the default offsetting set and my grid + sixths set on multiple pairs and I will test this for May alongside Leon's latest hybrid idea a few posts above.

Have a great weekend everyone.

[TraderJoeForex]

Keeping the default 1682 candles for Sixths in my control sets in April has saved me from my initial thinking that 720 (1month on H1) would be better so both the grid + sixths and default offsetting sets that I am continuing my test with now have 1682 for Sixths as below. Continuing with the same test account and myfxbook link for anyone following I now have these two sets on the majors + GJ and will run this for May.

[rapple]

Hi all. As always, thank you to all those developing this EA for all their time and assistance

Just wanted to report my 1 month data for HGBDSSFB which I am trading a bit different to others

I have set it up so that it only takes single trades when the 1W & 1D CSS are in agreement and there is a FB signal on H1. {HGBFB CSS 1D1W_1HFB.set}

I then use MPTM to set the TP and SL on each individual pair at 2.5ATR (Daily) {MPTM 3SL basket.set}

I also have MPTM set so that the Basket TP is 100% and the SL is 3%.

The idea is that the SL and TP on each pair are wide enough to give the trade room to move and it only closes if it hits the 2.5ATR TP or SL (Note: it takes quite a few days close each trade).

The 3% basket SL is there so that if a number of positions reverse all trades are closed out. I am hoping if this is the case then the 1D or 1W CSS would have reversed so there will be no more trades opened till the trend is established again (to date I have not had a basket closure so not sure about this setting yet - 3% maybe to high on a $10k demo account)).

Also note that I have removed most of the NZD and CHF pairs

Cheers Richard

[TraderJoeForex]

Hi all. As always, thank you to all those developing this EA for all their time and assistance

Just wanted to report my 1 month data for HGBDSSFB which I am trading a bit different to others

I have set it up so that it only takes single trades when the 1W & 1D CSS are in agreement and there is a FB signal on H1. {HGBFB CSS 1D1W_1HFB.set}

I then use MPTM to set the TP and SL on each individual pair at 2.5ATR (Daily) {MPTM 3SL basket.set}

I also have MPTM set so that the Basket TP is 100% and the SL is 3%.

The idea is that the SL and TP on each pair are wide enough to give the trade room to move and it only closes if it hits the 2.5ATR TP or SL (Note: it takes quite a few days close each trade).

The 3% basket SL is there so that if a number of positions reverse all trades are closed out. I am hoping if this is the case then the 1D or 1W CSS would have reversed so there will be no more trades opened till the trend is established again (to date I have not had a basket closure so not sure about this setting yet - 3% maybe to high on a $10k demo account)).

Also note that I have removed most of the NZD and CHF pairs

Cheers Richard

A solid and logical approach, well done.

[SteveHopwood]

The latest core library update is in post 1 at http://www.stevehopwoodforex.com/phpBB3 ... =87&t=5077.

This is for the Diad traders. I noticed the relevant lines were disappearing; the code needed functions to replace it when we restart the platform/change time frames etc.

Coders, I found something that will affect any EA needing to know when was the earliest trade in a group/basket. It probably affect my other shells as well, although I have not checked.

I have declared a constant at the top of the file: #define million 1000000; I use this as a starting point e.g.
double lowestPrice = million;
if (OrderOpenPrice() < lowestPrice) lowestPrice = OrderOpenPrice();

I used million incorrectly when finding the earliest order open time. Do a search for this line in CoundOpenTrades():
EarliestBuyTradeTime=million;

EarliestBuyTime is a datetime variable, so this construct from further down the function fails:

Code: Select all

            //Furthest back in time
            if (OrderOpenTime() < EarliestBuyTradeTime)
            {
               EarliestBuyTradeTime = OrderOpenTime();
               EarliestBuyPrice = OrderOpenPrice();
               EarliestBuyTicketNo = OrderTicket();
            }//if (OrderOpenTime() < EarliestBuyTradeTime)

You can see that it affects a number of variable values. The code needs to be:
EarliestBuyTradeTime=TimeCurrent();

You will need to check that all the Earliest...TradeTime variables are initialised correctly at the start of CountOpenTrades()

[trader689]

looks great thanks for the share Rapple

Hi all. As always, thank you to all those developing this EA for all their time and assistance

Just wanted to report my 1 month data for HGBDSSFB which I am trading a bit different to others

I have set it up so that it only takes single trades when the 1W & 1D CSS are in agreement and there is a FB signal on H1. {HGBFB CSS 1D1W_1HFB.set}

I then use MPTM to set the TP and SL on each individual pair at 2.5ATR (Daily) {MPTM 3SL basket.set}

I also have MPTM set so that the Basket TP is 100% and the SL is 3%.

The idea is that the SL and TP on each pair are wide enough to give the trade room to move and it only closes if it hits the 2.5ATR TP or SL (Note: it takes quite a few days close each trade).

The 3% basket SL is there so that if a number of positions reverse all trades are closed out. I am hoping if this is the case then the 1D or 1W CSS would have reversed so there will be no more trades opened till the trend is established again (to date I have not had a basket closure so not sure about this setting yet - 3% maybe to high on a $10k demo account)).

Also note that I have removed most of the NZD and CHF pairs

Cheers Richard

[SteveHopwood]

I have DIAD enabled and noticed this morning that whenever it took a sell, it instantly took out two buy trades. Initial buy trades behaved as expected. The problem was in the function that replaces missing lines; it was drawing them in buy line style when it should have been drawing them in sell line style.

The well dragged up little bot obediently discovered the buy lines, correctly spotted that the market had already crossed them and sent the buys. I have remarked before that my bots are impeccably polite and cooperative; just dumber than a box of rocks.

Fixed in post 1 at http://www.stevehopwoodforex.com/phpBB3 ... =87&t=5077.

DIY fix could not be easier. Copy this over the top of the existing void ReplaceMissingLines() function:

Code: Select all

void ReplaceMissingLines()
{
//if (Symbol() == "USDCHF") Alert(EarliestTradeTicketNo);
   if (!BetterOrderSelect(EarliestTradeTicketNo, SELECT_BY_TICKET, MODE_TRADES) )
      return;//Trade was closed for some reason
//if (Symbol() == "USDCHF") Alert("Y");
    
   //Calculate the line prices
   Trade1Price = OrderOpenPrice();   
   if (OrderType() == OP_BUY)
   {
      Trade2Price = Trade1Price + (Trade2At / factor);
      Be1Price = (Trade1Price + Trade2Price) / 2 + (DiadBreakevenProfit / factor);
      Trade3Price = Trade1Price + (Trade3At / factor);
      Be2Price = (Trade2Price + Trade3Price) / 2 + (DiadBreakevenProfit / factor);
      TpPrice = Trade1Price + (TakeProfit / factor);   
   }//if (OrderType() == OP_BUY)
         
   if (OrderType() == OP_SELL)
   {
      Trade2Price = Trade1Price - (Trade2At / factor);
      Be1Price = (Trade1Price + Trade2Price) / 2 - (DiadBreakevenProfit / factor);
      Trade3Price = Trade1Price - (Trade3At / factor);
      Be2Price = (Trade2Price + Trade3Price) / 2 - (DiadBreakevenProfit / factor);
      TpPrice = Trade1Price - (TakeProfit / factor);      
   }//if (OrderType() == OP_SELL)

   //Replace any missing lines
   if (OpenTrades == 1)
   {
      if (ObjectFind(TradeLine2Name) == -1)
      {
         if (OrderType() == OP_BUY || OrderType() == OP_BUYLIMIT || OrderType() == OP_BUYSTOP)
            DrawHorizontalLine(TradeLine2Name, Trade2Price, Trade2LineColour, Trade2BuyLineStyle, 0);
         else
            DrawHorizontalLine(TradeLine2Name, Trade2Price, Trade2LineColour, Trade2SellLineStyle, 0);
      }//if (ObjectFind(TradeLine2Name) == -1)
      
      if (ObjectFind(BeLine1Name) == -1)
      {
         DrawHorizontalLine(BeLine1Name, Be1Price, BreakEven1LineColour, BeLineStyle, 0);
      }//if (ObjectFind(BeLine1Name) == -1)
         
      if (ObjectFind(TradeLine3Name) == -1)
      {
         if (OrderType() == OP_BUY || OrderType() == OP_BUYLIMIT || OrderType() == OP_BUYSTOP)
            DrawHorizontalLine(TradeLine3Name, Trade3Price, Trade3LineColour, Trade3BuyLineStyle, 0);
         else
            DrawHorizontalLine(TradeLine3Name, Trade3Price, Trade3LineColour, Trade3SellLineStyle, 0);
      }//if (ObjectFind(TradeLine3Name) == -1)
         
      if (ObjectFind(BeLine2Name) == -1)
      {
         DrawHorizontalLine(BeLine2Name, Be2Price, BreakEven2LineColour, BeLineStyle, 0);
      }//if (ObjectFind(BeLine2Name) == -1)
            
      if (ObjectFind(TakeProfitLineName) == -1)
      {
         DrawHorizontalLine(TakeProfitLineName, TpPrice, TakeProfitLineColour, STYLE_SOLID, 0);
      }//if (ObjectFind(TakeProfitLineName) == -1)
            
   }//if (OpenTrades == 1)
   


}//End void ReplaceMissingLines()

You can remove the greyed out Alerts - they were part of the debugging and not needed.

[SteveHopwood]

Ye Gods alive, dead and all points in between or outwith, but the sodding Diad functions have driven me to distraction today. I have been using them to code a suite of manual DFC helper EA's and coding errors have been emerging.

The fixed and finally working version of the bloody core library is in post one in the usual place. Don't even have bad dreams about asking "Where is that" if you want to live.