Fibotrend » Sat Apr 09, 2016 5:39 am wrote: Hi Remon,

I'm looking forward to seeing your new strategy.

After running many scenarios on the simulator I think M1 is more reliable than M2.

As previously posted you can see the reason in this scenario (no spreads/all settings tested)

Runup: 111001110011100111001110011100111001110011100111001110011100


The difference is that M1 is always hedged to cover losses on any retracements. M2 is only hedged on pullbacks and it needs at least two consecutive winning levels above the retracement to score a profit.

That's why M1 is so clever, you are making money by pulling off the oldest trade and by keeping the grid hedged all the time. It's literally a sowing-and-reaping system, I really hope the EA will be developed around this idea.

Keep on rolling mate نستمر على هذا النحو

Simihouse » Sat Apr 09, 2016 11:39 am wrote:Sorry if I annoy you again Mr. Hopwood, but I think that the logic of the last EA is far from the original Remon's idea.
The operations from line 2035 to 2115 follows exactly the strategy, is the winning side closure logic and is ok. But the logic from 2118 is entirely different. Following the original idea we no need to find some equally profitable trading to close the losing positions. We need to create a virtual box, the Slush Fund (as per the first EA), and saving inside that the half of gains of the winning sides closures. After, as soon as the losing side start to retrace for a value of the grid pips, we must check if we have enough funds in the Slush Fund box to close the bigger loser along with the last winner(s) of the losing side. In this way until we no have enough Slush funds the trades are anyway hedged so the DD should be controlled, and most important, we no leave abandoned trades on the losing side.
The first EA was more similar to the original trade logic and indeed works almost perfectly, but only for the BUY operations...the instructions for the SELL operations are missing and moreover the first EA not open the symmetric grid (BUY limit orders are missing and the first order is a SELL instead a BUY-SELL). I think start again from there is the better solution.

SteveHopwood » Sat Apr 09, 2016 11:16 am wrote:I need fresh pairs of coders' eyes and more logical brains to look at this.

Run it thorough a quick and dirty M1 Open Prices craptest and watch the graph. You will see the balance rising as the equity falls. After the crash at the end, go to the bottom results tab and you will see a trade from early on that never got closed.

Go to line 2037 which is the start of the block that looks for a simple offset of the biggest winner against the loser. Comment out lines 2037 to 2116, recompile and rerun the craptest. The hedge/basket closure is working, as is offsetting x winners on one side against the biggest loser on the other.

I have fully commented the block from 2037 to 2116, so can you see what is wrong?

Arkangela » Sat Apr 09, 2016 9:56 pm wrote:
with the block between 2037 and 2116 line and the block between 2250 and 2357 trades are closed only if there are more than 2 buy or 2 sell or 2 buy and 2 sell trades. if you have just one or two sell or buy trades they will be closed only with basket closure.
Example : if price go down 25 pips it will open 2 sell trades after that if price go up 60 pips the two sell trades are not closed

Anyone wants to cause trouble? Fuck off and find a different forum.

1111111111111111100000001110000011111111110000111111000001111100001111000011110000111000011110001110000110

remon78eg » 07 Apr 2016, 05:57 wrote:
Method 2 + Source Code !

This Is The Update Of Method1 And Method2
With The Source Code!!

So, Programmers Can Copy The Code To Metaeditor
And Convert It To MQL4 With The Exact Same Way ,

The EA Can Open Orders In (Memory Like Simulator)
And Copy Trades From Memory To Real Trades.
By Looking To Deals.Deal() Array
And Open Orders When It Is Have Orders.

PeterUK » Sun Apr 10, 2016 10:18 am wrote:Guys,

If you're getting a runtime error '5' when running the Methed 2 simulator:



Download the latest version posted by remon78eg on 7th April (http://www.stevehopwoodforex.com/phpBB3 ... 32#p138432)

remon7878eg: Query for you: should something be showing in the 'heart area' when running the Method 2 simulator (highlighted orange)?



Numbers are showing when I run the Method 1 simulator, and also when I use Method 2 settings in Method 1 Simulator.

Using the following input:

Code: Select all

1111111111111111100000001110000011111111110000111111000001111100001111000011110000111000011110001110000110

Running Method 2 settings in Method 1 Simulator gives different results than just running Method 2 simulator OOTB. Is there a reason? unless both simulators are coded slightly differently...

Method 1 Simulator with Method 2 settings:


Method 2 Simulator:


Many thanks for the simulators, it has helped me understand the ideas and concepts behind the implementation and development of 'Rinse 'n Repeats', and we can assist Steve when possible. I will demo the EA next week

Thanks,

Peter

remon78eg » 07 Apr 2016, 05:57 wrote:
Method 2 + Source Code !

This Is The Update Of Method1 And Method2
With The Source Code!!

So, Programmers Can Copy The Code To Metaeditor
And Convert It To MQL4 With The Exact Same Way ,

The EA Can Open Orders In (Memory Like Simulator)
And Copy Trades From Memory To Real Trades.
By Looking To Deals.Deal() Array
And Open Orders When It Is Have Orders.