Topic: Can someone check this strategy?

Using generator I've came to this wonderful strategy, only to discover that it's a bug. Generator chose bar closing for opening and for closing points at the same time. If one chooses bar opening as opening point, the mistake becomes clear. Can someone check?

Post's attachments

NAS100USD 1d long.xml 11.22 kb, 25 downloads since 2013-11-11 

You don't have the permssions to download the attachments of this post.

2 (edited by footon 2013-11-11 23:39:50)

Re: Can someone check this strategy?

Does it open and close positions on the same bar if you look the chart?

P.S. I'm away from the platform to look it myself...

Re: Can someone check this strategy?

I think you have to see for yourself. I'm not sure if what I said is completely correct.

The strange thing is that if I choose "bar opening" as opening point the result is terrible, but if I choose "bar closing" the result is excellent. It's very strange for me!

I think this is related to the number of ambiguous bars, which is strangely much much higher when I select "bar opening". If someone could download the strategy and check it with care I would be very welcome!

Re: Can someone check this strategy?

I don't see anything. If you change logic conditions, then it is expected you won't see the same result. Where's the bug? Describe in detail.

About strangeness - bar opening does NOT equal to bar closing, otherwise where would be the point in this. No disrespect or offense intended, but at the moment you say like I'm trying my left shoe on my right foot, but it doesn't fit or feel right, strange! lol

Re: Can someone check this strategy?

Of course I know it's not exactly the same, but how can you justify the big difference? Your answer is generic, I would like someone to look at the bars and understand what's wrong.

Re: Can someone check this strategy?

Hmm, I don't have big difference. -10 vs -12.

Let me ask you this - how can you unjustify the difference between open and close of a bar? Does your data consists of doji bars and nothing else? Even more so, when you have break even, TP/SL all set up. Call it generic or whatever you like, but it doesn't change the fact that if you alter the starting point, you change the strategy. You are not dealing with one variable here.

Lastly, define what you mean by "look at the bars, understand what's wrong", do you realize how vague your statement is? You need to provide more explanations for someone helpful to come and see. You say bug, what is the bug, description, essence, ways of reproducing? It is not particularly humanly impossible to pick a trade, and trace it on the chart to understand what it is doing or what it's not doing, hell, you could even post a pic here and ask if got it right or not.

Good luck and trade well!
F

7 (edited by nquental 2013-11-12 20:22:33)

Re: Can someone check this strategy?

I post some graphs so you'll see what I'm talking about. In theory the difference of opening prices between "bar opening" and "bar closing" as opening point should be minor. If not, can you please explain why? This is not a metaphysical question. If it changes the strategy, in which way does it change it and how can it justify the enormous difference in results I've achieved? For me it's clear that something is wrong.

In my view this is due to the fact that with "bar closing" much less bars are ambiguous than with "bar opening" (disregard the specific numbers shown on the right because they are misleading, they refer to the bars with trades; concentrate instead on the graphs I post). Why there are more ambiguous bars with "bar opening" than with "bar closing"? If you could explain this to me I would understand what's going on with such a massive change than cannot otherwise be explained.

Post's attachments

bar closing 1.jpg 96.46 kb, file has never been downloaded. 

You don't have the permssions to download the attachments of this post.

Re: Can someone check this strategy?

2nd pic

Post's attachments

bar closing 2.jpg 88.66 kb, file has never been downloaded. 

You don't have the permssions to download the attachments of this post.

Re: Can someone check this strategy?

3rd pic

Post's attachments

bar opening 1.jpg 95.07 kb, file has never been downloaded. 

You don't have the permssions to download the attachments of this post.

Re: Can someone check this strategy?

4th pic and last

Post's attachments

bar opening 2.jpg 80.33 kb, file has never been downloaded. 

You don't have the permssions to download the attachments of this post.

Re: Can someone check this strategy?

nquental wrote:

I post some graphs so you'll see what I'm talking about. In theory the difference of opening prices between "bar opening" and "bar closing" as opening point should be minor. If not, can you please explain why? This is not a metaphysical question. If it changes the strategy, in which way does it change it and how can it justify the enormous difference in results I've achieved? For me it's clear that something is wrong.

In my view this is due to the fact that with "bar closing" much less bars are ambiguous than with "bar opening" (disregard the specific numbers shown on the right because they are misleading, they refer to the bars with trades; concentrate instead on the graphs I post). Why there are more ambiguous bars with "bar opening" than with "bar closing"? If you could explain this to me I would understand what's going on with such a massive change than cannot otherwise be explained.

You are on daily timeframe. So opening position at the bar opening and at the bar ending
seems to be 24 hour difference in trade. You do not open position at 11/13/2013 0:01 am. Instead you open postion at 11/13/2013 11:59 pm
That's a huge diffrence, don't you think.

Re: Can someone check this strategy?

nquental wrote:

Using generator I've came to this wonderful strategy, only to discover that it's a bug. Generator chose bar closing for opening and for closing points at the same time. If one chooses bar opening as opening point, the mistake becomes clear. Can someone check?

From What I found in FSB.
Regardless the opening logic (enter at bar opening or bar closing) when you use exit at bar close logic on daily timeframe the positions are closed imediatelly.
Bar Opening Time    Direction
1/2/2008 0:00    Long
1/3/2008 0:00    Closed
As the price did not move as well it seems so.
Transaction    Order Price
Open    1.4715
Close    1.47151

Re: Can someone check this strategy?

togr wrote:
nquental wrote:

I post some graphs so you'll see what I'm talking about. In theory the difference of opening prices between "bar opening" and "bar closing" as opening point should be minor. If not, can you please explain why? This is not a metaphysical question. If it changes the strategy, in which way does it change it and how can it justify the enormous difference in results I've achieved? For me it's clear that something is wrong.

In my view this is due to the fact that with "bar closing" much less bars are ambiguous than with "bar opening" (disregard the specific numbers shown on the right because they are misleading, they refer to the bars with trades; concentrate instead on the graphs I post). Why there are more ambiguous bars with "bar opening" than with "bar closing"? If you could explain this to me I would understand what's going on with such a massive change than cannot otherwise be explained.

You are on daily timeframe. So opening position at the bar opening and at the bar ending
seems to be 24 hour difference in trade. You do not open position at 11/13/2013 0:01 am. Instead you open postion at 11/13/2013 11:59 pm
That's a huge diffrence, don't you think.

It's a continuous market, normally the difference should not be very large.

Re: Can someone check this strategy?

togr wrote:
nquental wrote:

Using generator I've came to this wonderful strategy, only to discover that it's a bug. Generator chose bar closing for opening and for closing points at the same time. If one chooses bar opening as opening point, the mistake becomes clear. Can someone check?

From What I found in FSB.
Regardless the opening logic (enter at bar opening or bar closing) when you use exit at bar close logic on daily timeframe the positions are closed imediatelly.
Bar Opening Time    Direction
1/2/2008 0:00    Long
1/3/2008 0:00    Closed
As the price did not move as well it seems so.
Transaction    Order Price
Open    1.4715
Close    1.47151

The positions are closed at the end of the day, very far from immediately.

Re: Can someone check this strategy?

nquental wrote:

In theory the difference of opening prices between "bar opening" and "bar closing" as opening point should be minor. If not, can you please explain why? This is not a metaphysical question. If it changes the strategy, in which way does it change it and how can it justify the enormous difference in results I've achieved? For me it's clear that something is wrong.

In my view this is due to the fact that with "bar closing" much less bars are ambiguous than with "bar opening" (disregard the specific numbers shown on the right because they are misleading, they refer to the bars with trades; concentrate instead on the graphs I post). Why there are more ambiguous bars with "bar opening" than with "bar closing"? If you could explain this to me I would understand what's going on with such a massive change than cannot otherwise be explained.

I see now your problem - you think of current bar opening and previous bar closing. In this sense, yes, there would be slight difference expected. BUT this is not how FSB works! We are talking about only current bar in regard to position opening. This means: 1) open at current bar opening OR 2) open at current bar close. Difference between them is the true range of the bar, the whole movement of one bar.

To increase your backtest reliability use the scanner. I think you have done that already, so be sure to cover the time period with lower timeframes.

Re: Can someone check this strategy?

I will test it during 1 month using a demo and check the results. If the backtest is confirmed this would be a wonderful strategy. I wonder how much money people here are making with their own strategies...

Re: Can someone check this strategy?

I make a few pips, I do not have anything that does not need continual monitoring and updating so I do not publish them. I still have not discovered a strategy that will work for long periods untouched.

I am experimenting with systems saved as indicators and then using logical operators to see if I can get more trades. No definitive results so far.

I hope your testing works out well.

Good luck.!!

My 'secret' goal is to push EA Studio until I can net 3000 pips per day....