Disliked{quote} All is well and good, but I was just trying to adapt the existing program of GoldenEA to the ideas of Paracelsus. He has posted many results here and I'm just trying to reproduce it. For new ideas, please make evaluations and post charts. If I program had no time to think ...Ignored
MGH System (Martingale Grid & Hedging)
{quote} good morning! and great job! only perform the test with the EU, it would be interesting to test other pairs? AU, GU, EJ, GJ, EA ...?
________________________
Just a visual update about what has happened during the news:
After a meditation about hands clap with one hand, I finished the version 2.3. The EA is still in work. There are still some errors in it, but the arrows look beautiful ... - The entrys are made for new bars. - I tested charts on M15, the trend is determined on H1 with two SMA 8 for typical prices, normal and with shift 1. (John Ehlers used this trend indicator) You can also take the EMA 200 trend. {image} {file}
Just little question about last version 2.3 on the picture
and next question
Thank you
- #808
- Nov 7, 2014 10:58am Nov 7, 2014 10:58am
{quote} Good point! I was already thinking of an emergency stop when Batch5 is in play, so instead of waiting for 5 USD Basket Profit, the EA would accept a -10% SL. Let's keep on thinking together guys!
extern double EquityLossPercent = 10;
then
if (AccountEquity()<(AccountBalance()*(1-(EquityLossPercent/100))))
and
int CloseOrders(int Magic)
{
int total=OrdersTotal();
for (int cnt=total-1;cnt>=0;cnt--)
{
OrderSelect(cnt,SELECT_BY_POS,MODE_TRADES);
if(OrderMagicNumber()==Magic)
{
int type=OrderType();
bool result=false;
switch(type)
{
case OP_BUY : result=OrderClose(OrderTicket(),OrderLots(),MarketInfo(OrderSymbol(),MODE_BID),5,Red);
break;
case OP_SELL : result=OrderClose(OrderTicket(),OrderLots(),MarketInfo(OrderSymbol(),MODE_ASK),5,Red);
break;
case OP_BUYLIMIT :
case OP_BUYSTOP :
case OP_SELLLIMIT :
case OP_SELLSTOP : result=OrderDelete( OrderTicket() );
}
if(result==false)
{
Alert("Order ",OrderTicket()," failed to close. Error:",GetLastError());
Sleep(3000);
}
}
}
return(0);
}
But if that occurs many times it's bad....
{quote} You are right but we try here to get an EA that works somewhat like the original idea of KFX with the small changes of Paracelsus. Everything else I think belongs in an other new thread.
Does anybody can show a solid backtest with Swingman 2.3 EA ? Mine is really bad for this year.
So the base of the strategy is good, but not enough in itself. Needed some modifications, improvements.
{quote} The strategy, as it is in the EA's founded in this thread, doesn't have good backtest at all. I did a lot of backtests with these EAs with many settings, on more pairs, and there aren't good results. So the base of the strategy is good, but not enough in itself. Needed some modifications, improvements.
Namely: Finding an extremely unlikely (or at best: never seen) "Death Trail" price movement pattern and to build the strategy around it.
_________________________________
Bluepanther has summarized the mathematical side of similar approaches very well. Thank you, Bluepanther, and please stay with us here!
Finding an extremely unlikely - or "never seen" - "Death Trail" pattern is theoretically possible, even if the profits are not too impressive. Once this unlikely or never seen "Death Trail" is found, the rest is just "paperwork".
Just "off-hand" I can give an extreme example of a never seen price pattern: After "London Open" the market does not move at least 2 pips in any direction within the first 30 minutes. What are the chances of it (0.0000000000000001%)? Can anybody show a day in history when the price did this idling "price pattern" after London Open? I doubt it. So - theoretically - a "Holy Grail" could be constructed around this "never seen price pattern", namely that the price will break out of a 2 pips range with 99.999999999999999999% certainty within the first 30 minutes after "London Open". Again, it was just a theoretical exercise, but similarly "never seen price patterns" can be the basis of a successful MGH strategy.
In my approach of tweaking MGH I have followed this path and I ask fellow "MGH fans" to do the same.
I am positive that by reverse engineering an unlikely (or better: never seen) "Death Trail", MGH has a good fighting chance. Why am I so sure about it? Beacuse market movements are not random "mathematical" phenomena. (Lucky for us: they are driven by human emotions, i.e. mainly greed and fear.)
And this simple fact can be our "EDGE" to perfect MGH.